Atari/Atari Games Memos and Status Reports 1989 Jed Margolin ___________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 12 January 1989 Driver MultiSync ---------------- Leo is working on the MultiSync Board. I have submitted a CER for programmed part numbers. Driver ADSP II -------------- I am ready with the inputs. We need to know which 1M ROMs to use. Self-Test --------- I am working on the last routine. Parts to Manufacturing --------------------- Manufacturing still owes us: (8) TWIG ICs (180) VRAM sockets Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 20 January 1989 Driver MultiSync ---------------- We have ordered Rev 2 boards. Driver ADSP II -------------- I have given Leon the inputs. Self-Test --------- I have added a monitor test pattern at Rick Meyette's request. I have added the menu selection to disable broken controls. Manual ------ I have reviewed Rosalind's latest draft and made appropriate comments. Parts to Manufacturing --------------------- I have received the eight TWIG ICs. They still owe us 180 VRAM sockets. Self-Test --------- Norm Maeder is planning on having the Motor Amp PCB tested at two of the Tech Stations which are also used for repairing all the game boards. He would like to be able to test all the boards without having to change the Test ROMs. Note that testing the boards includes running the game program. If I give him a special set of ROMs that allows him to set the motor force to 250 but which do not run the game (or the other tests) then they will have to change ROMs to test the Motor Amp. If I give him a special set of ROMs that does allow him to set the motor force to 250 and does run the game (and the other tests) then: 1. There is the chance that the ROMs will mistakenly get shipped in a game. 2. If Max changes the program he will also have to burn a special set of ROM 0 EPROMs so they will continue to be able to set the motor force to 250. Now that Max has taken control of ROM 0 I cannot update the test programs on my own and still have them run the game. Tell me what you want me to do. Jed _____________________________________________________________________________ To: Norm Maeder Fr: Jed Margolin Re: Driver Main Test Procedure Dr: 25 January 1989 On the Test Procedure for the Driver Main PCB please add, starting with step 4: Step 4: Select MONITOR TEST PATTERNS. ---> The screen should show the Color Bars. Verify that the colors are present. Press Right Coin to bring up the MONITOR BRIGHTNESS screen. Ignore this screen. Press Right Coin again to bring up the Gray Scale Screen. The screen starts with a white line on the left of the screen. It then goes to black and gradually increases until it is again full white on the right of the screen. There will probably be a faint line down the middle of the screen and two fainter ones (one at one-quarter screen and one at three-quarters screen). These are normal. ---> What we are looking for are dots that are obviously out of place. Dots that are out of place will typically repeat every eight lines. ============================================================================== Note to Norm: These 'dots out of place' are VRAM defects. In about 20 prototype and preprod boards (640 VRAMs) I have only seen one VRAM with this problem. I do not know if it will be a problem in production. This problem cannot be detected by a test program. It must be checked visually. If it is determined that a VRAM is defective it must be located and replaced. I have written a program that produces a line on the screen and indicates which VRAM is producing that line. The Left and Right Coin switches cause the line to move left or right, again and indicating which VRAM is producing that line. This is an undocumented test program that is run from Diagnostics mode. It is undocumented because I do not know how useful it will be. (I only have one VRAM with this problem.) If you find Driver main boards with these VRAM problems please let me know. Erik and I will try to figure out if my test program can locate these VRAMs with Dots-Out-Of-Place. Jed _____________________________________________________________________________ To: Norm Maeder Fr: Jed Margolin Re: Driver Main Test Procedure Dt: 30 January 1989 I thought of another thing we should add to the test procedure. In the new Step 4, after verifying that there are no dots out of place, ---> Go on to the Scroll Test Screen. This shows diagonal lines and a hex number (0-F). (There is a screen that just shows diagonal lines. Keep going until you find the diagonal lines with the number.) ---> The diagonal pattern should move smoothly left and right as the hex number changes. (The hex number is the value put into the scroll register at 140M to scroll the display.) Thanks, Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 30 January 1989 Driver MultiSync ---------------- We have ordered Rev 2 boards. I have converted the Driver Test program to run on the MultiSync Board at medium speed. I am waiting for a standard speed monitor which I need in order to convert the program to run at standard speed. Driver ADSP II -------------- I have received plots but have not had time to check them. Manual ------ I have reviewed Rosalind's last draft and made appropriate corrections. Manufacturing ------------- I have reviewed Jim Buchanan's proposed test procedure, made appropriate comments, and returned it to him. I have asked Norm Maeder to add to the test procedure: 1. Check the Monitor Gray Scale Test to look for bits-out-of-place (There are some VRAM shift register defects that cannot be detected in software.) 2. Check the Monitor Scroll Register Test. Self-Test --------- I have added the ability for Manufacturing to run the Steering Wheel force up to 250, while normally it will only go up to 100. I have changed the Line Voltage Calibration program so it can be set to 220V. The calibration factor now changes as long as one of the Coin Switches is held down, instead of requiring a pulse for each change. Going from 120 to 220 takes about 30 seconds. _____________________________________________________________________________ _____________________________________________________________________________ 1 of 2 Driver Main Test Procedure Jed Margolin 2 February 1989 DOTS-OUT-OF-PLACE ----------------- On the Test Procedure for the Driver Main PCB please add, starting with step 4: Step 4: Select MONITOR TEST PATTERNS. ---> The screen should show the Color Bars. Verify that the colors are present. Press Right Coin to bring up the MONITOR BRIGHTNESS screen. Ignore this screen. Press Right Coin again to bring up the Gray Scale Screen. The screen starts with a white line on the left of the screen. It then goes to black and gradually increases until it is again full white on the right of the screen. There will probably be a faint line down the middle of the screen and two fainter ones (one at one-quarter screen and one at three-quarters screen). These are normal. ---> What we are looking for are dots that are obviously out of place. Dots that are out of place will typically repeat every eight lines. ============================================================================= Note: These 'dots out of place' are VRAM defects. In about 20 prototype and preprod boards (640 VRAMs) I have only seen one VRAM with this problem. I do not know if it will be a problem in production. This problem cannot be detected by a test program. It must be checked visually. If it is determined that a VRAM is defective it must be located and replaced. I have written a program that produces a line on the screen and indicates which VRAM is producing that line. The Left and Right Coin switches cause the line to move left or right, again and indicating which VRAM is producing that line. This is an undocumented test program that is run from Diagnostics mode. It is undocumented because I do not know how useful it will be. (I only have one VRAM with this problem.) If you find Driver main boards with these VRAM problems please let me know. Erik and I will try to figure out if my test program can locate these VRAMs with Dots-Out-Of-Place. Driver Main Test Procedure 2 of 2 SCROLL REGISTER TEST -------------------- In the new Step 4, after verifying that there are no dots out of place, ---> Go on to the Scroll Test Screen. This shows diagonal lines and a hex number (0-F). (There is a screen that just shows diagonal lines. Keep going until you find the diagonal lines with the number.) ---> The diagonal pattern should move smoothly left and right as the hex number changes. (The hex number is the value put into the scroll register at 140M to scroll the display.) ============================================================================== STEERING WHEEL TESTS (Motor Amp PCB) The normal force limits are +/- 100. To properly test the Motor Amp PCB the limits need to be changed to +/- 250. This is done by grounding connector J4, pin 4. ============================================================================== SLAPSTIK TEST Insert test ROMs into 200Y and 210Y. (Test ROMs with known checksums will be supplied.) Go to Main Test Menu and select 'Special Functions'. In the Special Functions Menu select 'Main Board Checksums' The Checksum listed next to ROM 7 H should be the same as the checksum of the ROM in 200Y. The Checksum listed next to ROM 7 L should be the same as the checksum of the ROM in 210Y. _____________________________________________________________________________ To: Rick Moncrief 1 of 2 Fr: Jed Margolin Re: Hardware Status Dt: 4 February 1989 Self-Test --------- I have added the on-screen display of the program version number for both the Main Board and the Sound Board. The Self-Test programs for the Main and Sound Boards are ready to be released. Driver MultiSync ---------------- We have received the Rev 2 boards; one is almost stuffed. (It was delayed because Manufacturing was shut down for Inventory and Karen couldn't get parts from them.) I have converted the Driver Self-Test to run on the MultiSync at medium speed and am almost done converting it to run at standard speed. I will need to convert the steering wheel tests to use the Optical Steering Wheel reader. Driver ADSP II -------------- I have shown Leon a schematic (and therefore board) error but have still not had time to check the plots. 2 of 2 Electronic Coin Mechs --------------------- 1. Powering an electronic coin mech from the existing transformer is marginal. The ideal solution is for Ireland to use a 5 volt switcher that also has a +12 volt output. If that cannot be done for Hard Drivin' they should use a separate 12 volt supply. The electronic coin mech requires 11 - 15 volts at 220 ma continuous, 700 ma peak (assuming no sorter). A linear supply with a 12 V transformer, four 1N5401 diodes, a 3300 uF capacitor, a 12 volt 3-terminal regulator with heat sink, a 10 uf capacitor and a few 0.1uF capacitors should do the job. 2. In reading the manual I could find nothing to suggest that it could be set up to produce more than one pulse per coin. A question that only the manufacturer can answer is: If two channels are both programmed for the same coin will both outputs be pulsed, and if so, will they pulse at the same time? That might take care of having a two Swiss franc coin produce two pulses but would not help a five Swiss franc coin produce five pulses. How about this: For a 2 franc coin to produce 1 credit: Set Coin Mode to: 1 coin/1 credit Set Right Mech Multiplier to: 1 coin counts as 1 coin. For a 5 franc coin to produce 3 credits: Set Coin Mode to: 1 coin/1 credit Set Left Mech Multiplier to: 1 coin counts as 2 coins Set Bonus Adder to: 2 coins gives 1 extra coin Unfortunately, if they were both set up sumulanteously: two 2 franc coins will also give 3 credits. oops. There is a hardware solution: a circuit that outputs 3 pulses for every 1 input. The real solution is: Add the capability to the SOFTWARE, either in the game or in the coin mech. In the game this would require adding to the Left Mech Multiplier: 1 coin counts as 3 coins. {5 franc coin} The other settings would be: Coin Mode: 1 coin/1 credit Right Mech Multiplier: 1 coin counts as 1 coin {2 franc coin} Bonus Adder: No Bonus Adder Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 4 February 1989 Self-Test --------- The Self-Test programs for the Main and Sound Boards have been released. Driver MultiSync ---------------- We have received the Rev 2 boards; one is almost stuffed. (It was delayed because Manufacturing was shut down for Inventory and Karen couldn't get parts from them.) I have converted the Driver Self-Test to run on the MultiSync at medium speed and am almost done converting it to run at standard speed. I will need to convert the steering wheel tests to use the Optical Steering Wheel reader. Driver ADSP II -------------- I have shown Leon a schematic (and therefore board) error but have still not had time to check the plots. Optical Steering Wheel ---------------------- I have given Art the circuits for the Single and Dual Optical Readers. Manufacturing ------------- I was told that the Driver Main Boards were being held up because of parts shortages caused by parts having recently been scrapped. I called around but was unable to verify it. Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 13 February 1989 Manufacturing ------------- 2000 pf capacitor on the Motor Amp Board: ---------------------------------------- The capacitor originally specified (2000 pF, 5%, NPO, 100V axial) was not available. Candy found some 2000 pF, 5%, 500V Mica radial capacitors which were determined to be acceptable. The only other ones she could find were 1800 pF, 5%, NPO, 100V axial. The axial capacitor is within acceptable parameters and has the advantage of being auto-insertable. I advised her to take that into account along with the cost and to use her own judgement. The 4066 gets real hot: ---------------------- Jim Buchanan and his group have noticed that the 4066 at 205B on Driver Main gets hot. When we looked at it we discovered that the 4066 is being supplied with Ground to +5V instead of -5V to +12V which is what I specified in my inputs to PC and which, indeed, appear on the schematic. The PC Board does not match the schematic. Since this circuit is essential for certain kinds of kits we could do later, I agreed with Jim that the boards should all be modified. Since there is a good chance that the 4066s that have been powered up have been damaged I have also asked Jim to have them replaced on the Driver Main Boards that have been powered up. (He thinks about 50 boards have been powered up.) I will do a test program to test the DUART circuit and add it to the Test Menu (as soon as I can). Art has been notified of this PC Design problem. Driver MultiSync ---------------- I have received a Rev 2 MultiSync Board. I have converted the Driver Self-Test to run on the MultiSync at standard speed and am almost done converting it to run the compact Steering Wheel. Driver ADSP II -------------- It is on hold until I am able to give it proper attention. Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 15 February 1989 Driver MultiSync ---------------- 1. The Rev 2 board is up and working without mods, except as explained in item #4. 2. I have converted Driver Self Test to run on MultiSync Standard speed, including steering wheel tests for thr Optical Steering Wheel. 3. I have signed the film for the Opto-Coupler boards for the Steering Wheel . The boards are 'Dual OSR' and 'Single OSR'. (OSR stands for Optical Steering wheel Reader.) 4. During a chance meeting in the hall with Gary Stempler he informed me that some time ago he had made some component changes in his SCUM circuit and that he had recently told Glenn to make these changes on the MultiSync Board. He also recommended that I change the SCUM clock buffer to an 'F' part because of some problem they had had in Blasteriods. I reminded him that when I first came to him to ask his help in putting the SCUM circuit on the board he told me to just copy the circuit from a game, which I did. When I tried to establish whether he knew about these changes when he directed me to copy the circuit from a game manual he got huffy and walked away. I now have a MultiSync Rev 2 board which we had planned to turn into Rev A on which the SCUM circuit may or may not work properly. If he had informed me of these changes in a timely manner it would not have been a problem. Also, his having instructed Glenn to make changes on the MultiSync Board without asking me first, is totally unacceptable. Manufacturing ------------- I am working on the test for the DUART circuit. Driver ADSP II -------------- Still on hold. Jed _____________________________________________________________________________ 19 February 1989 Rick and Erik, This is where we are so far, On the production board set we have: 1. The reason the program crashes is because the ADSP board has a flakey RAM. When the screen says 'GSP Timeout' it means the GSP timed out waiting for the ADSP board. 2. I found a bent lead under the board. The Color RAM passes the test, even at 5.5 Volts. I would like to see more boards with 'bad' Color RAMs. There are several new Diagnostic programs to test Color RAMs. I have documented them at the end of this memo. The new GSPTST correctly reports the locations of the Color RAMs that fail the test. 3. The new GSPTST Version 8.1 is available [Margolin.GSPTST]GSPTST_R8_1.VLDA . I have made some ROMs but they will not run the Game. (Max has to do his thing with them.) ------------------------------------------------------------------------------ Erik: 1. Replace R165. 2. On my Rev A Board: The Color RAMs (30W, 40W, 30U, 40U, 30S, 40S) have bravely given their lives in order to verify that the color RAM test correctly identifies the locations of Color RAMs that have failed the test. Please have Karen remove them, put sockets on the board, and install new RAMs. The old Color RAMs should be disposed of with full honors. Jed Tests that are not documented in the Manual ------------------------------------------- 1 1 1 1 1 1 1 1 Switches are All ON (In case people read the manual which says that Up is OFF. X X 1 1 1 0 0 0 Identify VRAM (ORs ones) Use INC and DEC to scroll the line Left and Right. X X 1 1 0 0 0 0 Identify VRAM (ANDs zeros) Use INC and DEC to scroll the line Left and Right. X X 0 0 0 0 1 0 Write, then Read Red/Green Color RAM, adr=155, data=5555 X X 0 0 0 1 1 0 Write, then Read Red/Green Color RAM, adr=2AA, data=AAAA X X 0 0 1 1 1 0 Write, then Read Blue Color RAM, adr=155, data=5555 X X 0 0 1 0 1 0 Write, then Read Blue Color RAM, adr=2AA, data=AAAA X X 0 0 1 1 0 0 Color RAM Test Loop _____________________________________________________________________________ Mary, Would you please find out the lead time for: 138000-001 Opto Sensor, Dual, Custom 138001-001 Opto Sensor, Single We are using them on the Driver Compact and the Dual unit is custom. Thanks, Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 24 February 1989 Manufacturing ------------- 1. I investigated a production board set that had been approved by Test. The screen had said 'GSP Timeout' but it was not a GSP problem. The reason the program crashed was because the ADSP board had a bad RAM. The Test Program reported the location of the bad RAM. We replaced it and the program ran. I found a bent lead under the board in the Color RAM section. The Color RAM passes the test, even at 5.5 Volts. I have advised Jim Buchanan to watch out for bent leads in this section. The Color RAM test had a bug that prevented it from reporting the RAM location if more than one Color RAM failed the test. I have fixed the problem in Version 8.1 . I have also added several Diagnostic programs to test Color RAMs. They will not be documented for public use. 2. I have added routines to test the DUART (internal loop) and DUART circuit (with a Test jumper at J2/J3). ------------------------------------------------------------------------------ 3. On Tuesday it was determined that there was some kind of problem causing Game Resets. We modified three board sets to generate an NMI instead of a WatchDog and ran them overnight in Attract Mode. By Wednesday morning all three games were hung. The problem appear to be caused by noise on the /GSPWAIT wait that occasionally causes the 68010 to miss /DTACK, causing it to hang. The Bus Error Timeout Counter is less susceptible to this noise and thus did not generate a Bus Error signal. We tried two different fixes. One was to use a different address validation strobe on the GSP; the other was a resistor/capacitor to reduce the noise, generating a 1 ns delay in the signal which is ok. On Thurday Morning both modified games were still running. The unmodified game had hung. The preferred fix is to clean up the signal rather than move the noise burst, so that is what we have given Manufacturing. (a 33 ohm resistor in series with /GSPWAIT at 190E with a 39 pF capacitor to ground.) Between Wednesday night and Thursday morning I figured out a better fix, which is use a spare section of the LS244 at 170P to clean up /GSPWAIT. The timing margin in the GSPWAIT circuit is 23 ns so the 18 ns of the LS244 is ok. We tried it Thursday night and gave it to Manufacturing on Friday. This is the preferred fix. I have given the information to Art; he will ECN the assembly drawing and start the artwork for the next rev board. Assisting in the investigation: Max, Erik, and Stephanie. ------------------------------------------------------------------------------ 4. We have released Version 8.1 software. Rick Owens has the appropriate documentation. I have given Art a copy of the ROM Release Sheet and he will ECN the parts lists. Since we will have several different versions because of languages and Left and Right Hand Drive he will maintain different separate parts lists rather than have everything on one parts list. Oops Board for the Gate Array ----------------------------- 1. The Gate Array has some differences from the original circuit. The original circuit would have to be modified and tested. 2. Because the Oops Board would be physically separated from the ground plane, ground noise might be an insurmountable problem. 3. The original circuit has 17 chips. 4. How would the Oops Board connect to the PLCC footprint? 5. Security is lost. 6. A more workable solution is to do a new Main. Unfortunately there are only 13 empty spaces. 7. Increasing the space between boards to accomodate the Oops Board will increase the length of the Expansion Bus Interconnect Cable with possibly deleterious results. As per your order I have given Art a copy of the original circuit and asked him to analyse it but not to do anything about it. Jed _____________________________________________________________________________ To: Rick Moncrief 1 of 1 Fr: Jed Margolin Re: Hardware Status Dt: 10 March 1989 German Version -------------- I have completed work on the German Version of Self-Test. MultiSync --------- Film and fab for MultiSync Rev A is ready to be signed off. We need to decide on the 40 MHz oscillator module before releasing the parts list. Driver Compact -------------- We have received the boards for the Single Optical Steering Wheel Reader and for the Dual Optical Steering Wheel Reader. Karen has built some up. We should test them electrically. The Dual Opto Sensor (138000-001) is listed as a custom part. I have asked Mary Burnias to find out the availability of this part as well as the Single Opto Sensor (138001-001). Stun Runner ----------- I have given them a Rev 2 MultiSync Board (Board #2) so they can verify that the SCUM circuit still works. I do not have a monitor cable so I was not able to test the board. When Glenn put it on his system it worked except for the DUART, which he said he will fix. Jed _____________________________________________________________________________ To: Rick Moncrief 1 of 1 Fr: Jed Margolin Re: Hardware Status Dt: 17 March 1989 Driver Main - Color RAM Problem ------------------------------- There is a problem with about 7% of the boards; the Color RAM fails Self-Test but the monitor test patterns are ok and the game is ok. The problem is a circuit whose timing is marginal. The quick fix is a resistor and capacitor. Erik has given this mod to Manufacturing and and has also faxed it to Ireland. I have given it to Art and asked him to ECN the Assembly drawing. Erik and I are working on a real fix for the next PC Board Rev. MultiSync --------- We have signed off Film and Fab for MultiSync Rev A. We need to add the Color RAM fix to it; it looks like we will be able to do it before Purchasing buys boards. We need to decide on the 40 MHz oscillator module before releasing the parts list. Driver Compact -------------- We have received the boards for the Single Optical Steering Wheel Reader and for the Dual Optical Steering Wheel Reader. Karen has built some up. We should test them electrically. Opto Sensors ------------ According to ManMan the stock status and cost are as follows: 138000-001 Dual Opto Sensor 238 $2.61 138001-001 Single Opto Sensor 790 $1.00 Melanie Martin has checked into the availability of the Opto Sensors. Accoring to Honeywell (who makes the sensors) they are available in 6-8 weeks, maybe sooner. There is stock of fewer than 200 in distribution. Stun Runner ----------- Glenn has repaired the DUART in the Rev 2 MultiSync Board (Board #2) and the board is now 100% functional, including the SCUM. He has also repaired the REV 1 board so that it now works. Altogether, they have three working MultiSync Main Boards. Jed _____________________________________________________________________________ _____________________________________________________________________________ To: Pat Brosnan Fr: Jed Margolin Dt: 21 March 1989 Re: Problem with Hard Drivin' board set in France "ADDRESS ERROR" indicates a software bug, possibly because it is trying to report a hardware problem with the ADSP board that was not simulated in testing that software version. Please find out: 1. Was this one of the games at ATE that was sold? 2. What version of software is it? The software version appears on the Main Test Menu on all production software. 3. Does the ADSP Test in Automated Self-Test report bad ADSP Memory? 4. Does it do it at the start of the ADSP Test or at the end? 5. Does the game run? If I know what software release it is I can simulate an ADSP memory error and determine if the problem is software. _____________________________________________________________________________ To: Pat Brosnan Fr: Jed Margolin Dt: 22 March 1989 Re: Problem with Hard Drivin' board set in France It sounds like a problem with the ADSP Board. There are actually two problems: 1. The game doesn't run. 2. Self-Test gets an Address Error when it tests the ADSP Memory. ----------------------------------------------------------------------------- Part 1 (In order to be rigorous about this thing): 1. The most recent software available in English Language, Left Hand Drive is Version 8.1 . Please confirm this is what they have. (The software version appears on the Main Test Menu.) 2. Does the Driver Main PC Board have the mod to the GSPWAIT circuit? This is shown on Drawing No. A044425-01 "ASSEMBLY, SUB, DRIVER MAIN PCB" The mod to the GSPWAIT circuit is the mod that starts: MODIFICATIONS TO ASSEMBLIES USING 044426-01 REV A,B and C PC BDS: 1. COMPONENT SIDE OF PC BOARD: FOLLOW THE TRACE FROM 190E PIN 1 TO THE FEEDTHROUGH. and so on. 3. Did the game ever run? ------------------------------------------------------------------------------ Part 2: 4. This sounds like a problem on the ADSP Board. Could you give them a new ADSP Board and send the old one to me? (Sending me the whole board set would be better.) The reason is that I don't think I have seen this problem before. When I test the memory routines I usually pull out the RAMs, one at a time. (I have a board set with all the RAMs in sockets.) This may be a problem that only causes this software reporting problem when the RAM errors are at certain addresses. 5. Rick has told me it is a chore to be sending boards through Customs and has suggested that the main Distributor in each country where we sell these games have a spare board set so that individual boards can be more easily swapped. What do you think? ------------------------------------------------------------------------------ Part 3: 6. We will probably do a flying game on Hard Drivin' hardware but it won't be much like the TomCat game you saw. And it certainly won't be called TomCat. I was really unpopular with management for insisting on doing the game and as a result now I only do hardware (and some test programs). Regards, _____________________________________________________________________________ To: Pat Brosnan Fr: Jed Margolin Dt: 23 March 1989 Re: Problem with Hard Drivin' board set in France 3. Audio problem: A. It is possible that in working on the ADSP board the Interconnect Cable has become flakey. Try a new one. B. If the Interconnect Cable is not the problem: If the Sound Wave ROM checksums are bad it could be because the ROM is bad or because there are circuit problems. One way to find out is to plug the ROMs into an empty socket on the Main Board and do ROM checksums in 'Special Functions'. Location 200 V is ROM 4 H and 210 V is ROM 4 L. Do not unplug existing ROMs. Also, the ROM 7 pair will not work because they are connected through Slapstic which is not loaded. This only works with 27512s. The Sound Program ROMs are 27256s which cannot be checked this way. 4. Spare Board Sets at Distributors: Rick suggests one board set per Customs boundary. 5. I am sending you some block diagrams. 6. Erik says he wants his case back first. Regards, _____________________________________________________________________________ To: Rick Moncrief 1 of 1 Fr: Jed Margolin Re: Hardware Status Dt: 23 March 1989 Driver Main - Color RAM Problem ------------------------------- I have given Art the real fix for the Color RAM problem, to update the board. MultiSync --------- We have signed off Film and Fab for MultiSync Rev A. We need to add the Color RAM fix to it; it looks like we will be able to do it before Purchasing buys boards. We need to decide on the 40 MHz oscillator module before releasing the parts list. Driver Compact -------------- We have received the boards for the Single Optical Steering Wheel Reader and for the Dual Optical Steering Wheel Reader. Karen has built some up. We should test them electrically. Jed _____________________________________________________________________________ To: Rick Moncrief 1 of 1 Fr: Jed Margolin Re: Hardware Status Dt: 27 March 1989 Driver Main - Color RAM Problem ------------------------------- We have gone out for Film for the new boards with the real fix for the Color RAM problem. MultiSync --------- Art had Gary do the board even though Joe had been scheduled to do it. I have received plots and schematics for the MultiSync board and sent the schematic back with several part value corrections. Driver Compact -------------- We have received the boards for the Single Optical Steering Wheel Reader and for the Dual Optical Steering Wheel Reader. Karen has built some up. We should test them electrically before they are released. Pat Brosnan ----------- I responded to several faxes from Pat Brosnan. It appears that there was a Hard Drivin' in France that: Was a production game; Had Software release 8.1 (English, Left Hand Drive) The game did not run, it said 'ADSP Timeout' in Attract Mode; The ADSP Test in Automatic Self-Test said the ADSP RAM was ok; In 'Special Functions ADSP' the RAM Test terminated in an address error at 810003 and did not display the board picture. Not all the information is consistent. Although I explained that I had not seen that failure before and that it would be really useful if they returned the board to me, they elected to replace some (all?) the 6264s on the ADSP Board, thus solving the problem. Now we will never know. His last fax was that now the Sound Board didn't work, that there were ROM checksum errors. He didn't say if the checksum errors were Sound Program ROM or Sound Wave ROM. I relayed Erik's suggestion that the Interface Cable may have been damaged while they were playing with the ADSP Board. I also suggested that if the ROM errors were Sound Wave ROMs, he could check them by putting them in ROM 4 sockets on the Main Board and running the ROM Checksum test. This would identify whether the ROMs were really bad or if there was a circuit problem on the Sound Board. He asked for information about how the hardware worked. I sent him a block diagram. I have not received a response. Jed _____________________________________________________________________________ _____________________________________________________________________________ To: Rick Moncrief 1 of 2 Fr: Jed Margolin Re: Reply to Jim Seward's Fax Dt: 28 March 1989 A. The split pad to enable coin counters to operate independently cannot be accessed when the board is assembled. Cutting a resistor on the audio board as is used on other games is preferable. Answer: The Driver Main board was the first board I did for this project; the APU board came later. I had originally planned to run the cable from the Main Board directly to the Coin Counters in order to simplify the harness. That is why the split pad is on the Main Board. I do not agree that the split pad cannot be cut when the board set is assembled. The ADSP Board and Sound Board do not come out to the edge of the Main Board and there is room to use an Exacto knife to cut the split pad. As long as it is done before the board set is installed in the game it should not be too difficult. I agree it would have been preferable to use a resistor on the APU Board. I will keep that in mind for future projects. ------------------------------------------------------------------------------ B. The Motor Amp PCB should have been designed to operate on either 110VAC or 220VAC without using a stepdown transformer, as was done with the TRIAC PCB on APB. Answer: The TRIAC PCB on APB turns an incandescent lamp on or off. (The lamp must be a 220V lamp when used on 220VAC.) The Motor Amp PCB uses Pulse Width Modulation and is considerably more complicated than a Triac Switch. The standard method for making a 110/220VAC compatible unit is to have a jumper so that at 220 VAC the circuit operates as a full wave bridge but at 110 VAC it operates as a voltage doubler. This means that the motor circuit would always operate at 320 VDC instead of 160 VDC as it does now. The circuit uses both N and P channel FETs. Even today, the P channel FETs that we would need are not available above 200V. Of course, the circuit could have been done totally differently, but it would cost more to build and would have taken more time to design and test. Also, in the TRIAC PCB used in APB there are only a few places where an unwary technician can contact the line, and when operated at 110VAC that is the voltage. In the Motor Amp PCB, even when connected to 110VAC, there would always be 320 VDC on the board, thereby increasing the danger. If we had found it necesary to enclose the Motor Amp because of this, it would have increased the cost. ------------------------------------------------------------------------------ 2 of 2 C. I have not made any provision to power Electronic Coin Mechs. Answer: You are right. The problem was FIRST brought to my attention in late January of this year, long after the hardware was designed and released. At that time I was given copies of the correspondence on the subject, most of which was between Pat McCarthy and yourself. Although you might have reasonably expected Pat to keep the other engineers informed on the matter, such was not the case. My recommendation is that games intended for use with Electronic Coin Mechs come with 5V Power Supplies with a 12V output. ------------------------------------------------------------------------------ D. The Component Bracket cannot be secured when two coin counters are fitted. I am using the same Component Bracket that I used on Star Wars five and a half years ago. I assumed that since no one had complained in that time, it was ok. I have passed your complaint on to Jerry Lichac. ------------------------------------------------------------------------------ E. You seem to be saying that I should not have used the Coin Switches to run Self-Test. I needed three accessible, momentary switches to have Self-Test be of the most amount of use for the factory and for the Operators. The Key Switch is mostly used to execute commands. The Abort switch alone would have made Self-Test clumsy to use. ------------------------------------------------------------------------------ My Turn: The hardware for Driver Compact has either already been released or is too late to change. If you had made your concerns known to me earlier I could have done something about them. (some of them) _____________________________________________________________________________ To: Rick Moncrief 1 of 1 Fr: Jed Margolin Re: Hardware Status Dt: 29 March 1989 (interim) MultiSync --------- Have received the Parts List. Film and Fab are ready to be signed. Jim Seward ---------- I have replied to Jim Seward's fax wherein he whined about a bunch of stuff. ADSP Boards ----------- Problem: If they order ADSP-2100s in PQFP for Stun Runner, we cannot use them in Driver Compact. If they order ADSP-2100s in PGA for Driver Compact we cannot use them in Stun Runner unless we do another board (PGA with extra 27512s) If we ordered 2100s in the PGA would Analog Devices let us change some of the order to PQFPs? The only way to be completely flexible is to do an ADSP Board with a PGA and six 27512s. (This requires adding two 27512s to the existing ADSP Board.) Schedule: Driver Compact STUN -------------- --------- ADSP Board ADSP II Board ADSP-2100 in PGA ADSP-2100 in PQFP 4 x 27C512 6 x 27C512 Finished Goods May 22 Finished Goods July 14 ADSP II ------- Film and Fab April 13 Order Proto Boards April 14 -- } 3 weeks Build and Test Proto Boards May 5 -- } 1 week Order for Production May 12 -- } 8 weeks Production Build July 14 -- We get one shot at a prototype board. If there are problems with it we would probably have to correct it and buy production boards without another prototype. ADSP with PGA and six 27512s would have to meet the same schedule. Jed _____________________________________________________________________________ { To PCB Department, Date Unknown } For the Driver Compact parts list: Let's entitle the Top Parts list 'H.D. Compact PCB' The Sub Parts List should remain 'MultiSync PCB' Thanks, Jed _____________________________________________________________________________ To: Rick Moncrief 1 of 2 Fr: Jed Margolin Re: Hardware Status Dt: 31 March 1989 Driver Compact -------------- Have signed off Driver MultiSync. Purchasing has ordered boards. We have received the boards for the Single Optical Steering Wheel Reader and for the Dual Optical Steering Wheel Reader. Karen has built some up. We should test them electrically before they are released. The monitors all look bad at fast Sync. Stun Runner ----------- I gave Glenn a schematic for a circuit to detect video from the VCR. He built it and after some futzing we got it to work. ADSP Boards ----------- Problem: If we order ADSP-2100s in Plastic Quad Flat Pack (PQFP) for Stun Runner, we cannot use them in Driver Compact which uses them in the Pin Grid Array (PGA). If we order ADSP-2100s in PGA for Driver Compact we cannot use them in Stun Runner unless we do another board (PGA with extra 27512s) Solution #1: Do an ADSP Board with a PGA and six 27512s. (This requires adding two 27512s to the existing ADSP Board.) Solution #2: Analog Devices has agreed that, if we order the ADSP-2100 in the PQFP, they will allow us to change the order to PGAs if we give them three weeks notice. I like solution #2. Memo from Tom Smith is reproduced: ------------------------------------------------------------------------------ From: MIKE::SMITH 31-MAR-1989 16:32 To: KIM::MARGOLIN Subj: RE: ADSP-2100s Jed, We are going to spec the 500 units for the compact hard drivin. We all feel confident that it will be a real success. If it should go the other way, Mary feels that we will not have any problem in returning them for full value and/or exchange for the quad pack. ---------------------------------------------------------------------------- Schedule: 2 of 2 Driver Compact STUN -------------- --------- ADSP Board ADSP II Board ADSP-2100 in PGA ADSP-2100 in PQFP 4 x 27C512 6 x 27C512 Finished Goods May 22 Finished Goods July 14 ADSP II ------- Film and Fab April 13 Order Proto Boards April 14 -- } 3 weeks Build and Test Proto Boards May 5 -- } 1 week Order for Production May 12 -- } 8 weeks Production Build July 14 -- We get one shot at a prototype board. If there are problems with it we would probably have to correct it and buy production boards without another prototype. ADSP with PGA and six 27512s would have to meet the same schedule. Jed _____________________________________________________________________________ To: Rick Moncrief 1 of 1 Fr: Jed Margolin Re: Hardware Status Dt: 10 April 1989 Driver Compact -------------- The Dual Optical Steering Wheel Reader PCB is ok. The Single Optical Steering Wheel Reader needed the documentation changed to make sure the OptoCoupler is installed correctly. Hard Drivin Manual, 2nd Edition ------------------------------- It seems that the file containing Chapter One was lost and had to be regenerated from an earlier version. She forgot to check the Specifications Table. She will do an errata sheet. ADSP Boards ----------- Have received plots to check. Stun Runner ----------- In order to use existing coin software: 1. The Coin Switches will be connected to the MultiSync Board instead of to the SA II. 2. One end of the Coin Counter will be connected to the Coin Counter driver on the MultiSync Board. The other end will be connected the CC1+ on SA II to pick up +14V through a resistor. Carole will find out if the Counter Counter has an internal protection diode. If it doesn't, she will put one at the Coin Counter Connector. Carole will add a power ground from SA II to the MultiSync Board to augment the ground in the SCUM Cable. The Stun Runner current requirement from the +5 Volt supply is 7.5 Amps. Rick Meyette assures me that tne +5 Volt 10 Amp supply is rated to supply 10 Amps up to 50 degrees Celsius. Therefore, Stun Runner can use the +5 Volt 10 Amp supply. I have obtained a series number for Stun Runner Programmed Parts. I have received a memo on diagnostic test results from RPM. I need to know: 1. What is the protocol for communicating to SA II? 2. How do I call S_STAT? 3. Is there a period after which, with no reponse, I can assume the SA II is broken? 4. The Status bits indicate three ROMs (at 4000, 8000, and C000). The SA II schematic shows only one program ROM. If this ROM is good will all three ROM status bits be set or just one? If one, then which one? 5. The situation for RAMs is similar. 6. Does the SA II program deliver ROM checksums? 7. Does the SA II test the ROMs connected to the MSM6295? Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Motor Amp Dt: 9 April 1989 Motor Amp Problem ----------------- Board #1 -------- 1. The problem is sensitive to the line voltage. Threshhold Force -250, line is 120VAC unloaded -> 108 loaded Threshhold Force -180, line is 105VAC unloaded -> 100 loaded 2. The problem is mostly with Negative Force, while turning the motor positive. Some do it at maximum Negative Force without turning the motor. One board does it at Positive Force while turning the motor negative. 2. Swapping Motor Amp Drivers makes no difference. 3. No Difference: 1000 uF across CR4 1000 uF across CR2 1000 uF across CR2 to ground 1000 uF across CR3 1000 uF across CR2 and CR3 0.01 uF Capacitor across the motor 10 uF across R36 makes it worse Capacitors were attached with clip leads. 4. Average Line current goes down when it happens. With Current Probe and oscilloscope at primary - it does not appear that there are any enormous peaks, such as would happen if two FETs in the same leg were turned on at the same time. 5. When the problem happens, the +20V goes below +10V, the +15V goes to 0V. This turns off the motor, the +20V returns, the motor turns on, the +20V goes away, the motor turns off, and so on. Supplying the +20V Zener independently almost fixes it. When supplied independently, the +20V circuit requires 30ma (motor off) before going into zener limiting. Theory: Current pulses on the positive and/or negative Rails are getting into the +20V zener circuit. Probably also the U15V Section. With: +20 supplied independently, J4 unplugged, Q2 off (R43 shorted), Q6 shorted with a jumper, It stopped doing it completely. Unfortunately, when I cranked it up I hosed the board. (Board #1) Board #2 -------- With: +20V supplied internally, J4 unplugged, Q2 off (R43 shorted), Q6 shorted with a jumper, The negative force section died after a short period of vibration. (The positive side still worked.) Board #3 (a good amp) --------------------- With: 0 Force and Primary 120VAC (unloaded), current thru R36 is 40 ma. -250 Force ( Primary drops to 108 VAC), current thru R36 is 26 ma. Board #4 (a bad amp) -------------------- Current thru R36 spikes up to 100 ma (and down to zero). With another 3K resistor in parallel with R36 the problem improved considerably but was not totally eliminated. Current thru R36 ---------------- Board #4 (Bad) 25 ma motor off, 32 ma with -250 force Board #3 (Good) 25 ma motor off, 25 ma with -250 force Board #2 (Bad) 25 ma motor off, 25 ma with -250 force Voltage on +Rail Primary 120VAC (unloaded) ---------------- Board #3 (Good) 160VDC with 0 Force, 110VDC with -250 Force Board #4 (Bad) same (voltage ripples due to motor chattering) Current sufficiency thru R36 ---------------------------- Voltage across R36 = Vin - 15(CR3) - 5.1(CR2) - 15(CR4) = Vin - 24.9 = Vin = 30V Primary power of 120VAC produces +Rail of +150(unloaded) and +100(loaded) The current thru R36 is (160-30)/3K = 43 ma unloaded (110-30)/3K = 26 ma loaded The power thru R36 is (160-30)**2/3K = 5.7 Watts unloaded (110-30)**2/3K = 2.1 Watts loaded ------------------------------------------------------------------------------ When supplied independently, the +20V circuit requires 30ma (motor off) before going into zener limiting. ------------------------------------------------------------------------------ Adding another 3K Resistor in parallel with R36 would produce, at maximum high line 132VAC, worse case conditions: VDC = 132*1.414-30 = 186-30=156V Current is 156/1.5K = 124 ma unloaded CR3 Power = 15*124 = 1.86 W (5W zener) CR2 Power = 5.1*124 = 0.63 W (1W zener) CR4 Power = 15*124 = 1.86 W (5W zener) ============================================================================== Connecting a wire jumper across zero Ohm resistor R41 made no difference. Grounding the Control Ground (CGND) at +15V zener CR4 instead of at protection diode CR14 made it WORSE. ============================================================================== Connecting a jumper from CR18 to the -Rail made it WORSE. ============================================================================== A 10 uF Capacitor from 15VTP to Ground made no difference. ****************************************************************************** To: Rick Moncrief Fr: Jed Margolin Re: Motor Amp problem Dt: 10 April 1989 1. The problem is sensitive to the line voltage. Threshhold Force -250, line is 120VAC unloaded -> 108 loaded Threshhold Force -180, line is 105VAC unloaded -> 100 loaded 2. Swapping Motor Amp Drivers makes no difference. 3. Average Line current goes down when it happens. With Current Probe and oscilloscope at primary - it does not appear that there are any enormous peaks, such as would happen if two FETs in the same leg were turned on at the same time. 4. Voltage on +Rail was the same on both good and bad Motor Amps. 5. Supplying the +20V Zener independently almost fixes it. When supplied independently, the +20V circuit requires 30ma (motor off) before going into zener limiting. With: +20 supplied independently, J4 unplugged, Q2 off (R43 shorted), Q6 shorted with a jumper, It stopped doing it completely. Unfortunately, when I cranked it up I hosed the board. (Board #1) The current thru R36 appears to be insufficient for reliable circuit operation. 6. With the motor force set to zero, using a variac to set the line voltage, the +20V circuit goes out of regulation with a +Rail below about 105V. This agrees with the +Rail voltage at full line and full motor load when it screws up. ============================================================================== With another 3K resistor in parallel with R36 the problem improved considerably but was not totally eliminated. Adding another 3K Resistor in parallel with R36 would produce, at maximum high line 132VAC, worse case conditions: VDC = 132*1.414-30 = 186-30=156V Current is 156/1.5K = 124 ma unloaded CR3 Power = 15*124 = 1.86 W (5W zener) CR2 Power = 5.1*124 = 0.63 W (1W zener) CR4 Power = 15*124 = 1.86 W (5W zener) ============================================================================== Theory: 1. The current through the zeners is insuffient for reliable operation. 2. Current pulses on the positive and/or negative Rails are getting into the +20V zener circuit. Probably also the U15V Section. 3. Circuit parameters and board layout allow the FETs to break into oscillation at high currents. ============================================================================== Connecting a wire jumper across zero Ohm resistor R41 made no difference. Grounding the Control Ground (CGND) at +15V zener CR4 instead of at protection diode CR14 made it WORSE. ============================================================================== Connecting a jumper from CR18 to the -Rail made it WORSE. ============================================================================== A 10 uF tantalum capacitor from 15VTP to Ground made no difference. ============================================================================== The capacitance of a 'scope probe (10pF) across CR18 fixes it on my worst board. This is in addition to the resistor. _____________________________________________________________________________ From: KIM::MARGOLIN 10-APR-1989 21:08 To: MAEDER,MONCRIEF,MOORE,MARGOLIN Subj: Motor Amp Mods ..... Make sure the board is completely discharged before doing these mods..... I have talked with Rick and he agrees on the following Mods: 1. Add another 3K Ohm 10 Watt in parallel with the existing one. It is important that this new resistor assembly not come in contact with the Case. 2. On top of the board, add a 100 pF, 5%, 500 Volt (or higher) Mica capacitor across CR18. This should be located close to CR18. It would be ok to use the feedthrough located about 0.7" from the CR18 anode for one end of the capacitor. The other end of the capacitor should go directly to where the CR18 cathode terminal connects to the board. I believe this mod can be done to the board without removing it from the heatsink. I do not have a part number for this capacitor. Sorry. If necessary you can use 100 pF, 10%, 500 Volt (or higher) ceramic. We should start these mods on the existing rejected boards. After we have done a few we will probably want to add the mods to all new boards as well. I will try to come in today in time to go over these mods with you but have sent this VAX Mail in case I don't make it. Regards, Jed _____________________________________________________________________________ From: KIM::MARGOLIN 10-APR-1989 21:08 To: MAEDER,MONCRIEF,MOORE,MARGOLIN Subj: Motor Amp Mods ..... Make sure the board is completely discharged before doing these mods.... I have talked with Rick and he agrees on the following Mods: 1. Add another 3K Ohm 10 Watt in parallel with the existing one. It is important that this new resistor assembly not come in contact with the Case. 2. On top of the board, add a 100 pF, 5%, 500 Volt (or higher) Mica capacitor across CR18. This should be located close to CR18. It would be ok to use the feedthrough located about 0.7" from the CR18 anode for one end of the capacitor. The other end of the capacitor should go directly to where the CR18 cathode terminal connects to the board. I believe this mod can be done to the board without removing it from the heatsink. I do not have a part number for this capacitor. Sorry. If necessary you can use 100 pF, 10%, 500 Volt (or higher) ceramic. We should start these mods on the existing rejected boards. After we have done a few we will probably want to add the mods to all new boards as well. I will try to come in today in time to go over these mods with you but have sent this VAX Mail in case I don't make it. Regards, Jed _____________________________________________________________________________ 11 April 1989 Erik, I have a fix for the Motor Amp problem and have given it to Norm. A copy is attached. Ralph has rescheduled the Compact drop test for Thursday. (He has the unit.) Please see if you can bring up one, preferably two, Compact Motor Amps. I have some good and bad Strain Guage Amplifiers to look at. Rick says, "Things went well at the Show. There were favorable comments about both the Sitdown and the Compact..... . There was some concern about the security of the cash box." Also, The flicker is bad in a dark environment. And, There was noticeable beating between the line frequency and the vertical rate. Rick thinks it might have something to do with the games having been powered through a 50' extension cord. He wants us to power a game through a 50' extension cord. Rick will be back April 20 Max and Stephanie will be back Monday, April 18. I expect to be in at my usual time today. Jed ____________________________________________________________________________ _____________________________________________________________________________ To: Norm Maeder Fr: Jed Margolin Re: THE NEW AND IMPROVED HARD DRIVIN' MOTOR AMP MODS Dt: 4/12/89 ..... Make sure the board is completely discharged before doing these mods.... 1. Add the 3K Ohm 10 Watt in parallel with the existing one as in the previous mod. 2. Add a ferrite bead (141003-001) in series with the Q8 Gate. Observe standard static control procedures when unsoldering the gate lead, installing the ferrite bead, and resoldering the gate lead. 3. The 100 pF capacitor used in the previous mod is no longer used. I will initiate an ECN after it has been verified that these new mods solve the problem. Jed Margolin Erik Durfey _____________________________________________________________________________ { Date Unknown } Monitor Problem: When I heard that the video output wasn't right, I assumed it was too low. Rick Meyette says the problem is that the Stun Runner board produces Too Much video output voltage for the Sanyo Monitor. The way to fix it is to install the 3.3K resistors at R97, R108, and R112. Heretofore they were not loaded in order to boost the video output. With these resistors loaded, the maximum video output should be 3.0 volts. If it isn't we will adjust the value of the resistors until it is. There may be a problem with threshhold brighness values. To wit: Wells-Gardner Sanyo ------------- ----- Max Video 3.5 Volts (255 DAC) 3.0 Volts (255 DAC) Gray 1.5 Volts (85 DAC) 1.5 Volts (102 DAC) Threshhold 1.0 Volts (42 DAC) 1.0 Volts (51 DAC) Min Video 0.5 Volts (0 DAC) 0.5 Volts (0 DAC) The Sanyo cannot be adjusted to work with Wells-Gardner video levels. Meyette say the Wells can be adjusted to work with Sanyo levels. I suggest we use Sanyo levels. Let's plan on adding the resistors to a board and connecting it to a Sanyo monitor. It may be necessary to adjust the gray levels in software. Jed _____________________________________________________________________________ { Date Unknown } TIs request for a waiver is for the following parameters: Parmeter From To Datasheet Waiver -------- ---- -- --------- ---------- TPLH D Q 6.0 ns max 8.0 ns max TPHL D Q 6.0 ns max 7.0 ns max TPLH C Q 11.5 ns max 13.0 ns max I have looked at the Driver Main circuit where the 74AS573s are used and I believe these degraded parts will work, BUT ALL BETS ARE OFF IF OTHER PARAMETERS ARE ALSO DEGRADED, ESPECIALLY: tsu Setup time, data before enable C low transition th Hold time, data after enable C low transition Also, I object to their insistence that these parts be non-cancellable and non-returnable. If these parts don't work properly in the circuit we ought to be able to return the parts not used and cancel the parts not received. Regards, Jed _____________________________________________________________________________ { Date Unknown } When Ralph and the guys removed the front piece to clean the monitor on one of the H.D Compacts they are getting ready for shipment, they noticed that the monitor bracket was bent. Nobody noticed it in the game. We were primarily concerned that it reach Ireland in one piece. The options were: 1. Replace the monitor which would require mechanical time as well as electrical time to reverse the yoke wires and perform monitor setup. Erik isn't here today so I would have to do it. (It would take me longer to do it than it would take him, and apparently the truck is waiting.) 2. Remove the monitor and ship it in a separate box. 3. Not send the game today. Ralph was pretty sure that the monitor bracket will hold together so we decided to send the game as is. The reason for this memo is so: 1. You know what we did. 2. You will alert Ireland that the monitor bracket is bent. I assume we are sending it to Ireland so they can convert it to 220VAC operation. Regards, Jed ____________________________________________________________________________ Atari Confidential Hard Drivin' Status Report Date: April 13, 1989 Reported by Margolin This week's top story, filed by correspondent Rick Moncrief, concerns the Atari Games Distributers' Show. Moncrief reports: "Things went well at the Show. There were favorable comments about both the Sitdown and the Compact..... . There was some concern about the security of the cash box." In local news, the problem with the Hard Drivin' Motor Amp has been addressed and, with Erik Durfey's help, hopefully solved on this, the second attempt in as many days. A petition for an ECN will be filed after it has been verified that these new mods solve the problem. -30- _____________________________________________________________________________ To: Rick Moncrief 1 of 2 Fr: Jed Margolin Re: Hardware Status Dt: 19 April 1989 Sitdown Motor Amplifier Problem ------------------------------- The problem seems to be fixed with the resistor added to R36 and a ferrite bead on the Q8 Gate. ECNs are in progress. RAM Tests --------- I got an ADSP board from Al Vernon that does the same thing as the ADSP board in France, which is, the ADSP RAM test geta a 68010 Address Error Exception. The problem is due to one of the sections of the RAM test. (We got the RAM Test from Mike Albaugh.) The RAM Test usually is stopped after it finds a single error; we log the error and keep on going. This particular section screws everything up when it is made to continue after finding an error. This section of the RAM test does not detect missing parts (which is how I tested the RAM tests). I propose we remove this section of the RAM test for the ADSP RAM tests. The RAM tests for GSP Color RAM, GSP Simple, GSP Complete, and MSP Complete appear to not use this test. The code for the GSP Verify and MSP Verify tests should be checked and changed if necessary. The code requires changing ROM 0 High and ROM 0 Low. We should release new versions for all four languages so that all future games will have useable RAM tests and so that Customer Service can send an update to those who need it. ADSP II ------- I have reviewed the plots and the parts list and sent them back for regrooving. Driver Compact -------------- The movement on the right hand side of the screen is related to the line frequency. It goes away at 53 Hz which, by an amazing coincidence, is also the Vertical rate. Unplugging the power to the flourescent lamp makes the movement go away. The problem may be a ballast transformer located too near the CRT yoke. The flicker, even in a darkened room, is not objectionable. In both units that came back from the show the cable to the motor amp had become disconnected. After consulting with Rick Meyette, Stevie decided to use a ferrite bead with a smaller diameter so it will stay on the cable near the Main board. Therefore it will not hang down, so it will not swing back and forth and pull the other end out of the Motor Amp. 2 of 2 Texas Instruments request for waiver ------------------------------------ Erwin asked me to look into a request TI made to Mary to accept 74AS573s that did not meet specifications. The following was my reply. =============================================================================== TIs request for a waiver is for the following parameters: Parmeter From To Datasheet Waiver -------- ---- -- --------- ---------- TPLH D Q 6.0 ns max 8.0 ns max TPHL D Q 6.0 ns max 7.0 ns max TPLH C Q 11.5 ns max 13.0 ns max I have looked at the Driver Main circuit where the 74AS573s are used and I believe these degraded parts will work, BUT ALL BETS ARE OFF IF OTHER PARAMETERS ARE ALSO DEGRADED, ESPECIALLY: tsu Setup time, data before enable C low transition th Hold time, data after enable C low transition Also, I object to their insistence that these parts be non-cancellable and non- returnable. If these parts don't work properly in the circuit we ought to be able to return the parts not used and cancel the parts not received. Regards, Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 28 April 1989 Driver ------ We have released new ROM versions with the ADSP RAM Test repaired. I have documented the error messages that are used in Self-Test. Max has documented the error messages used in the game. Customer Service was working with a Rev A Main Board schematic. I gave them the newest set (Rev G) as well as: The list of error messages (Both Max's and mine); A write-up of the ZRAM; A complete list of program versions with locations and checksums. Driver Compact ------------- We were supposed to get some MultiSync Boards in by April 10 for a quick look but it didn't happen. We are scheduled to receive 420 production boards on May 4, which is next Thursday. ADSP II ------- We are ordering prototype boards. Stun Runner ----------- I have given to Glenn The Mod to eliminate the infrequent GSP lockup. Star Wars --------- About two years ago Atari Games licensed Star Wars for a consumer home game to Domark in Europe. Lately, they have also licensed it to Broderbund. According to Becky Depew the current bonus plan only goes back to Marble Madness. Star Wars was developed before then. Although she did not come out and say it, the implication was clear that the game is considered to be exempt from the plan. Also, according to Becky Depew, They (the Executive Committee) will make a decision about it sometime in May. My position is that the fact that the current bonus plan started with Marble Madness is immaterial; Star Wars bonus ought to be paid under the Bonus plan that was in effect at the time the game was developed. I have supplied Becky with a copy of the Bonus plan that was in effect at the time Star Wars was developed. I would like to know what, if any, support I will receive from Engineering Management in this matter. I would also like to know how much of the license fee is kept by Tengen and how much gets passed on to Coin Op for bonus purposes. Jed Possible Stun Runner kit for Pole Position ------------------------------------------ Prices are from John Ray's document distributed at the meeting. A kit for Stun Runner would require: Material: $1150.52 Labor: $ 22.00 Overhead: $ 127.60 -------- Fully Burdened cost $1300.12 Distributer's cost (40% margin) $1820.17 ============================================================================== PCBs: Multisync PCB with VRAMs and EPROMs $568.52 ADSP II PCB with OTPROMs $256.60 SA II PCB with EPROM, OTPROM $ 84.20 MultiSync/ADSP flat cable $ 5.00 ------- $914.32 Power Supply (10A): $ 52.00 Flight Controller: $110.00 Flight Controller sub panel: $ 9.00 (This is a total guess) Attract Plex: $ 10.00 Attract Decal: $ 5.50 Side panel decals: $ 7.20 Main Harness: $ 30.00 Control Panel Harness: $ 7.00 Labels and Manuals: $ 5.50 ------- Material Total: $1150.52 PCB Labor Hours 2.20 * $10/hour $22.00 Overhead Fixed 2.20 * $42.00 $ 92.40 Variable 2.20 * $16.00 $ 35.20 ------- $127.60 _____________________________________________________________________________ To: Pat Brosnan Norm Maeder Mark Sherman Elaine Shirley Bob Stewart Al Vernon Fr: Jed Margolin, Rick Moncrief Re: New ROMs Dt: 28 April 1989 We are releasing new ROMs for all four versions: North American, British, German, and Japanese. Changes were made to Self-Test; there are no changes to the game play. 1. The ADSP RAM Test should now work properly. There was a problem that sometimes resulted in a 68010 Processor Error instead of showing the RAM Test on screen. 2. On the ADSP 'Scope Loop screen I have added a test to toggle the BCON signal. This is to help find bad 74LS259s on the ADSP Board. In this test, /BCON at 7K pin 7 should be high for 2 us and low for 11 us. New boards should have the new ROMs, there is no need to change ROMs on existing boards. Customer Service should probably send new ROMs to any customers who need them to fix a game. Note: Only ROMs at 200R and 210R are changed. ECNs are in progress. We are also coming out with a list of error messages and what they mean. Regards, Jed Rick _____________________________________________________________________________ To: Al Vernon and Mark Sherman Fr: Jed Margolin Re: Writing to the ZRAM Dt: 28 April 1989 The ZRAM (Zero-Power RAM) stores information for High Scores, Game Statistics and Control Setup adjustments. It also contains the Timekeeper Clock/Calender. The ZRAM package contains a Lithium Battery so it retains the information (and the Timekeeper continues to run) even with the power off. The ZRAM can be read at all times without doing anything special. ZRAM Writes must be enabled by writing to $604008 and $60401A. You must write to both. (It doesn't matter what data you write.) Once the ZRAM Writes are enabled they stay enabled until they are disabled. ZRAM Writes are disabled by writing to either $60400A or $604018. You should write to both to make sure the ZRAM Writes are safely disabled and immune to program crashes or power down glitches. (It doesn't matter what data you write.) The ZRAM address range is $FFFF 4000 to $FFFF 4FFE but $FFFF 4FE2 - $FFFF 4FFE is used by the Timekeeper and should not be written to. Writing to it can screw up the clock. Therefore you should only write to: $FFFF 4000 - $FFFF 4FE0 You should disable ZRAM Writes before powering down the board. Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 19 May 1989 ADSP II ------- The Rev 1 prototype board works completely. There are a few changes I want to make for Rev A. Originally, the board ran all the test programs but did not run the game. The problem turned out to be that the graphics ROMs were in the wrong sockets. The ROMs were in the wrong sockets because the production parts are mislabeled and then swapped so they end up in the correct sockets. Driver ADSP Checksum ROM Parts List PN ------------ -------- --- ------------- 10L A701 1L 136052-1104 10K DC02 0L 136052-1103<--- Note 10J CB03 1H 136052-1102<--- Note 10H 2C04 0H 136052-1101 Note: the parts we have from Manufacturing have -1102 and -1103 labels swapped. Stun Runner ----------- I have given them the ADSP II Board and have asked Karen to make some more. I have given Ed an improved Self Test program with the changes he requested. I am working on the schematic for the ADSP Graphics RAM Board to accomodate the extra memory space. Driver Compact ------------- I have reviewed the draft of the H.D. Compact manual and made corrections to it. We received two MultiSync PC Boards on May 14. Gwen is on vacation and Karen and Emmitt are busy. Also, Leon refuses to provide her with a stuffing diagram and Karen does not have time to do one. Besides, it is PC's job to do stuffing diagrams. Electronic Coin Mechs in Europe ------------------------------- I recommend that Components approve versions of the +5V switching supplies that come with a 12V output. According to Downend's memo the current required is 1.6 Amps. The data sheet I have calls for 11 - 15 volts at 220 ma continuous, 700 ma peak (assuming no sorter). Jed Primary Secondary x 1.414 - 0.7(diode) 12V Regulator ------- -------- ------- ------------ --------------- 120 VAC 12 VAC 16.7 VDC 16.0 VDC 16.0 - 12.5 = 3.5 VDC 105 VAC 10.5 VAC 14.8 VDC 14.1 VDC 14.1 - 12.5 = 1.6 VDC 120 VAC 14 VAC 19.8 VDC 19.0 VDC 19.0 - 12.5 = 6.5 VDC 105 VAC 12.2 VAC 17.3 VDC 16.6 VDC 16.6 - 12.5 = 4.1 VDC 132 VAC 15.4 VAC 21.8 VDC 21.0 VDC 21.0 - 12.5 = 8.5 VDC A 7815 12 Volt regulator is 11.5 V to 12.5 V. Dropout Voltage at 1 Amp @25 degrees Celsius is 2.0 V. Recommended input voltage is 14.6 VDC. 7812 is not recommended for currents greater than 1 Amp. DC Voltages do not take into account Ripple. A 14 VAC winding at low line will produce 16.62 VDC not including ripple. Assume enough capacitance to reduce ripple to 1 V which gives 15.62 VDC. This would be enough to power a 7812. Assuming a normal current of 220 Ma the dissipation will be 1.9 Watt high line. At 0.7 Amp it will be 6.0 Watt. A 12 VAC winding will not work reliably for this application. _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 30 May 1989 Driver Compact Self-Test ------------------------ The Self-Test program is waiting for 'Play Sounds' so it can be released. Ditto on British and German versions. MultiSync --------- The MultiSync PC Board has not been received. Norm thinks maybe tomorrow. To recap: Karen refused to stuff the board without a stuffing diagram and Leon refused to do a stuffing diagram. If and when we do receive a board from production it will be the start of the run, not a preliminary board. This is not what was suposed to happen and it really stinks. Stun Runner ----------- I have completed the schmatics for the ADSP II Graphics RAM. Presumably it is now being built. ADSP II board #2 was received DOA. Erik is working on it. I have contacted 3M and am getting samples of their PQFP socket. Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 2 June 1989 Driver Main ----------- Production gave Erik two non-working Rev E boards and told him that practically all of the Rev E boards refused to work. We quickly figured out that the problem was that the 50 MHz oscillator modules that run the 34010-MSP were bad. We fixed one and verified that the oscillator on the other board was bad. Erik returned the boards with an appropriate note. In the process I have discovered that production is not using the new ROMs that we released last week. They are still using the original Version 8.1 . Driver Compact Manual --------------------- I have marked up the MultiSync schematics for the manual as well as a memory map appropriate for publication and given them to Rosalind. I have proofed the final draft and passed it on. Driver Compact Self-Test ------------------------ The Self-Test program is ready to be released in North American, British, and German versions. MultiSync --------- We have received a MultiSync PC Board from Production. It appears to work. I have given Art the inputs for the changes to the parts list. The last two digits of the part number will also be the last two digits of the checksum. For example, the checksum for the ROM with the part number 136068-1102 is xx02. Rick Meyette did not want to let us do this but was persuaded to do so at the meeting. Stun Runner ----------- ADSP II board #2 was received DOA. the problem was a trace short that would have been picked up in production by the board tester. I received samples of the PQFP socket made by 3M. They don't work with the footprint for the AMP socket. (It is a mirror image.) I have informed the rep of this and will call 3M directly next week. Too bad, because the 3M socket is a better socket. It has locking tabs to hold the IC in. Jed _____________________________________________________________________________ To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 8 June 1989 Driver Compact -------------- We have released the improved version (3.1) with the bugs fixed. We have provided Production with the Test Procedure for the hardware. Sound Board ----------- We have released the improved version (3.2) with the bugs fixed. Driver ADSP ----------- We have ECNed the parts list to use OTPs. Other ----- I written a reply to the "Fan Mail" we received about XY. 8 Bit A/D Converters -------------------- National ADC0809CCN Total Unadjusted Error is +/- 1 LSB for 0 to 70 degrees Celsius, +/- 1.25 for TMin to TMax which is -40 to +85 Celcius. I expect that our temperature range is 0 to 70 degrees C. Total Unadjusted Error includes offset, full scale, linearity, and multiplexer errors. We are not concerned about offset and full scale errors, only differential linearity. Unfortunately, they do not separately characterize differential linearity. However, the DAC section uses a 256R Ladder (instead of R2R) which guarantees monotonicity. Therefore I would expect the differential linearity to be less than 1 LSB. (Presumably the multiplexer errors affect only absolute error, not differential errors.) By comparison the ADC0808CCN specifies: +/- 0.5 LSB at 25 degrees Celsius, +/- 1 for TMin to TMax which is -40 to +85 Celcius. Notice the specsmanship. The ADC0808 is 0.5 LSB at 25 degrees C while the ADC0809 is 1 LSB from 0 to 70 degrees C. The above comments also apply to the Texas Instruments ADC0808 and ADC0809 except: They use capacitors in their ladder instead of resistors. It is a 256C ladder and they also guarantee monotonicity. They do specify differential linearity separately, which (at 25 degrees C) is: ADC0808 is +/- 0.25 LSB ADC0809 is +/- 0.5 LSB I tried to fill in a chart comparing how the 8 Bit and 12 Bit A/Ds read the steering wheel pot but it was impossible to get the mechanical resolution needed for such a task. I can write a program to do it in software if you want. Let me know if you do. I have submitted a CER for the ADC0808 anyway. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 16 June 1989 H.D. Compact ------------ I have signed off the parts lists for H.D. Compact British and H.D. Compact German. ADSP II ------- Most of the prototype PC Boards have manufacturing errors: traces shorted to each other, to +5V, etc. Erik has managed to bring up four boards so far. The ADSP Graphics RAM board works and has been turned over to the Stun team. I have signed the ECNs to change from Driver ADSP to ADSP II on both Hard Drivin' and H.D. Sitdown on the next buy. AMP makes an extraction tool for the PQFP (821958-1) that costs $98. We probably won't buy very many. I have given Art the inputs to turn Rev 1 into Rev A. They are: ------------------------------------------------------------------------------ ADSP II REV 1 TO REV A J. Margolin 6/13/89 ---------------------- 1. PQFP Silkscreen needs to be chamfered. [see attached sheet] 2. Add PQFP assembly instructions to the assembly drawing. [see attached sheet] 3. Add 1K pullup resistors to /BUFF1 and /BUFF2. [see attached sheet] 4. Add traces as shown on attached sheet but do not load the IC. [see attached sheet] 5. Create a -02 version (parts list) so we can use it on the H. D. Compact. 6. ECN both Hard Drivin' and H. D. Compact to use ADSP II, effective with the next buy. ------------------------------------------------------------------------------ 8 Bit A/D Converters -------------------- The ADC0808 has been added to the AVL with its own number: 137651-001 . Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Review of Patent Applications Dt: 23 June 1989 The drawings are of execrable quality. Gearshift For A Vehicle Simulator .... Page 1, line 11: 'field' should probably be 'feel' . Rear Entry Booth ... Contains an additional set of drawings from the Gearshift application. The text does not contain line reference numbers (5, 10, 15, ...). Control Device ... ok Driving Simulator with Moving Painted Dashboard The text does not contain line reference numbers (5, 10, 15, ...). Page 12, line 16: '4 mHz' should be '48 MHz'. (Also note that the 'M' in MHz means 'Mega' (million) and must be capitalized. A small 'm' in 'mHz' would mean 'milli-Hertz', or one-thousandth of a Hertz.) Page 14, line 9: 'lode' should be 'load' Page 14, lines 17 thru 21: This explanation of memory-to-shift register transfers is totally wrong. The signal he calls 'memory-to-shift register-transfer' (120) is clearly shown on my schematics as '/TFR-SE' and is the signal that makes all the serial enables low for the purpose of allowing bulk screen clears. This signal has nothing to do with memory-to-shift register-transfers during the pixel scanning process other than that the signal must not be low. Page 15, lines 4 thru 6 and Figure 6: the writer does not understand the function of gates 112 thru 119. The signal on line 120 does not act as a gate, it acts to force all serial enables low, thus permitting the simultaneous shift register transfers needed to perform bulk clearing. This signal has nothing to do with pixel scanning other than that the signal must not be low. Fig 5 appears to show the VRAM as '4401'. It should be '4461'. _____________________________________________________________________________ 1 of 2 To: Rick Moncrief Fr: Jed Margolin Re: Data Retrieval System Dt: 26 June 1989 Max and I have looked at the material supplied by Pat Brosnan. We have some questions for him when he comes back July 15 but it looks like something we might be able to do. This is what it looks like: 1. They use an electronic coin mech such as the MARS or the SENTINEL. 2. There is a box called the ECHO box (Electronic Coin Handling Optimiser) connected to the elctronic coin mech's parallel interface. The ECHO box provides software control of the electronic coin mech and does the following: a. Allows the operator to set the coin modes with much more flexibility than our game software does; a. Controls the electronic mech's lock-outs to ensure that money is only accepted when the system is operational; b. Does not count coins if the door is open; c. Allows the operator to selectively disable the acceptance of certain coins (apparently slugs are a problem over there); d. Has an alpha-numeric interface; e. Has diagnostic routines to help locate faults with the electronic coin mech; AND f. Has an RS-232 output to the Datapak 2. 3. The Datapak 2 is the electronic data capture system. " It provides an automatic means of capturing and transferring information from site to a central location for management and security purposes." It has a realtime clock, an internal battery backup, and it connects to sensors to determine if a door is open. 4. The information in the Datapak 2 is collected by the Datapak 2 Auditor which is a self-powered, handheld unit that plugs into the Datapak 2. Data Dump takes about 10 seconds. 2 of 2 What it records: Name of site Serial number of machine Current Collection details comprising: Name of Machine Time and date of collection Number of individual coins in Number of individual coins out Number of games played Entry log giving each occurrence of: Time and date of entry Time and date of exit Identity of intruder Time log showing each occurrence of: Time and date machine switched on Time and date machine switched off Games played in above period The previous collection data The accumulated recording of individual coins in and out How to combine information from ECHO and the game for Datapak 2: The game has two RS-232 ports. Connect one port to ECHO and the other port to Datapak2. (The +12V, -12V, and Power Ground signals between ECHO and Datapak 2 must be connected.) Normally, the game will simply take the data received from ECHO and retransmit it to Datapak 2. Similarly, it will take data received from Datapak 2 and retransmit it to ECHO. When it needs to send data to Datapak 2 it can determine when it would be a good time to do so. If necessary it can temporarily store the transmissions from ECHO whilst it is talking to Datapak 2, and transmit it afterwards. The data is: 1200 Baud Half Duplex 1 start bit 8 data bits (LS Transmitted first) Odd parity bit 1 stop bit Our DUART supports these parameters. The first question is: What information do they want from game stats that is not already provided by ECHO/Datapak 2? The second question is: Where are they currently getting the -12V needed by the Datapak2? Jed _____________________________________________________________________________ _____________________________________________________________________________ 1 of 2 To: John Ray Fr: Jed Margolin Re: STUN RUNNER Parts that are common with Compact Hard Drivin' Dt: 6/27/89 ADSP II ------- We have ECNed Hard Drivin' and Compact Hard Drivin' to use ADSP II on the next buy. Therefore, ADSP II is common to Hard Drivin', Compact Hard Drivin', and Stun Runner. (Stun Runner will presumably make use of the two extra ROM sockets.) MultiSync Board --------------- 1. Stun Runner uses a 40 MHz oscillator module at XOSC4 instead of the 32 MHz unit that Compact Hard Drivin' uses. Oscillator modules frequently require several weeks' lead time. 2. To reduce costs the Timekeeper will be replaced with another ZeroPower RAM. 3. Slapstik 17 was reserved for Hard Drivin' but not used. I assume it is available for Stun Runner. 4. There are several parts that Compact Hard Drivin' uses that Stun Runner does not. The following are not used on Stun Runner 5 137540-150 1 IC, 48T02-15, RAM 200E 8 137543-001 1 IC, 68681 200A 46 137580-001 1 IC, 4066B 205B 47 137546-003 4 IC, 4464, 64K X 4, DRAM 5K,15K,25K,35K 83 136068-1135 1 IC, PROM, 82S123 195U 84 137403-001 1 IC, MC1488 210A 85 137263-001 1 IC, MC1489AL 210B 89 137614-001 1 IC, AD711KN 15B 136 144000-011 1 XTAL, 3.6864, STANDUP XTAL1 138 144008-005 1 XTAL, 50 MHZ, OSCILLATOR XOSC3 miscellaneous other ICs, connectors, resistors, and capacitors 2 of 2 5. There are a few parts that Stun Runner uses that Compact Hard Drivin' does not. They are: 137526-001 1 IC, SCUM 175W 137484-001 1 IC, AS08 175T 137134-001 1 IC, 74LS245 175R 144008-004 1 Oscillator, 40 MHz XOSC4 137412-117 1 Slapstic 17 200K 6. Parts with changes in quantities: 137538-002 IC, 34010-50 Uses One instead of Two 179237-068 SOCKET, 68 PIN Uses Two instead of Three 137442-150 IC, 48Z02-15, RAM Uses Two instead of One I need Ed to tell me that the special board I had made up for him works, so I can give Art the schematic, so we can get an official Stun Runner parts list. Jed _____________________________________________________________________________ ATARI CONFIDENTIAL HARD DRIVIN' APPLIED RESEARCH Date: 29 June 1989 Number 478xx Reported by J. Danforth Margolin while Rick Moncrief is on a well deserved vacation. Hard Drivin' ----------- Erik has tested the new seat post weldment and pronounced it satisfactory. Compact Hard Drivin' ------------------- Now that we are using shielded speakers (in order to use them up, thereby saving money) the motor assembly, which was a tight fit, anyway, doesn't quite fit anymore. The motor assembly will be ECNed at great cost in order to continue to use up the old speakers. Everything else is copacetic. ADSP II ------- ADSP II has been released. ADSP II is a cost reduced version of the Driver ADSP board with the capability of using two additional graphics ROMs. It was developed for Stun Runner but is compatible with Hard Drivin' and Compact Hard Drivin'. We are phasing it in with the current buy. Stun Runner ----------- I have had a MultiSync board regrooved to reflect how it will be stuffed for Stun Runner. Ed has given the board his blessing and I have given the inputs to Art. (The bare board does not change, only how it is stuffed.) I have given Ed two more ADSP II boards. Max has been working on improving vector mode in the STUN RUNNER GSP code; Stephanie, while recuperating from her operation, has been making some modifications to the ADSP code to reduce the control to display lag. Data Retrieval for Compact Hard Drivin' -------------------------------------- Max and I have looked at the material supplied by Pat Brosnan. We have some questions for him when he comes back July 15 but it looks like something we can do. The main question is, what information do they want the game to deliver that the ECHO/Datapak 2 system does not already have? Home Computer version of Hard Drivin' ------------------------------------- Max has created a detailed map of the Hard Drivin' universe for Steve Calfee, and is working on supplying other needed information. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 30 June 1989 TI 34010 -------- Last week Max and I had a very nice conversation with Carrel Killebrew and Karl Guttag (the architect of the 34010). Stun Runner ----------- I have had a MultiSync board regrooved to reflect how it will be stuffed for Stun Runner. Ed has given the board his blessing and I have given the inputs to Art. (The bare board does not change, only how it is stuffed.) I have given Ed two more ADSP II boards. Erik has obtained and connected a Stun Sound Board so I can figure out how to use Sound Self Test. ADSP II ------- ADSP II has been released. I have signed off the parts list for the Driver version of ADSP II. The version for Stun Runner is ready anytime. Patent Applications ------------------- I have reviewed the patent applications and placed my comments in a separate memo. Data Retrieval for Compact Hard Drivin' -------------------------------------- Max and I have looked at the material supplied by Pat Brosnan. We have some questions for him when he comes back July 15 but it looks like something we can do. I have generated a separate memo on the subject. Jed _____________________________________________________________________________ 1 of 2 To: Pat Brosnan, Atari Games Ireland Ltd. Fr: Jed Margolin, Atari Games Corp USA Dt: 6 July 1989 Re: Echo/Datapak 2; Bad GSP and PSP ICs; Sound Board fixes. Echo/Datapak 2 -------------- I looked at the information you sent Rick and am sending you a copy of my analysis. If my understanding of the Echo/Datapak 2 system is correct, a. What information do you want the game to provide that the Echo/Datapak 2 system does not already have? b. Where do you get the -12 Volts for Datapak 2? If it is used only for RS-232 can you use -5 Volts? (We do on our RS-232.) c. No additional black box would be required to have the game add information to the Datapak 2. The game has Two RS-232 ports. Port 1 would be connected to Echo, and Port 2 would be connected to Datapak 2. The game would receive data from Echo on Port 1 and retransmit to Datapak 2 on Port 2. Similarly, the game would receive information from Datapak 2 on Port 2 and retransmit it to Echo on Port 1. The game would add its own information to Datapak 2 as required. This would require some programming (and replacing a pair of ROMs) but it looks like something we can do. We would need a Echo/Datapak 2 system as well as an Auditor in order test the code. Bad GSP and PSP ICs ------------------- You are right; the 34010-50 and 34012-50 ICs you sent us are bad. Self-Test does not do everything. Here's the deal: a. The GSP Verify Test is a complete memory test that is run by the GSP, which means that the Verify program lives in VRAM. Any bad VRAM or problem with the VRAM circuit can prevent the GSP Verify program from running. The 68010 loads the GSP Verify program into GSP VRAM and tells it to run. The 68010 periodically checks with the GSP to see if the program is done. If it is not done within a certain amount of time, the 68010 says, "times up", and reports that all the VRAMs are bad. It is unlikely that all of the VRAMs are actually bad. The Verify test is the most difficult test for it to pass; it Verifies that the VRAM is ok. It is generally not very useful for anything else. b. The GSP Complete Test is similar to the Verify test in that it is a complete test but it is run by the 68010 through the GSP interface. This means that the test will continue to run (and log errors) even with a bad VRAM or problem with the VRAM circuit. It takes a lot longer to run than the Verify test because the 68010 is a lot slower than the GSP. 2 of 2 c. The GSP Simple test is also run by the 68010 through the GSP interface. It performs fewer tests than the complete test which is why it takes less time to run. Again, it will continue to run (and log errors) even with a bad VRAM or problem with the VRAM circuit. d. If a test reports that all the VRAMs are bad, this is a valuable clue since it is unlikely that all the VRAMs are actually bad. The problem is most likely a bad address bit (LAD bus, 1MA bus, or 2MA bus) caused by a short on the bus, a bad 34010, or a bad 34012. It could also be caused by a bad bus resistor (the 33 ohm bus resistors on 1MA and 2 MA as well as the signals going to 150U). I have seen at least one board with a hairline crack in a bus resistor. If a test reports that an entire row is bad, this is another valuable clue since it is unlikely that all the VRAMs in an entire row are actually bad. The problem is most likely a bad data bit caused by a short on the bus (LAD, VD), a bad 34010, or a bad 34012. e. If the 48MHz oscillator is bad the 34010 will fail to respond to the 68010 and will generate a BERR (Bus Error). f. "Complete" as in "Complete Test" means "as complete as is possible if you want the results this week" . Sound Board Fixes ----------------- I have seen Sound Boards that pass Self-Test but do not run game sounds during the game (or in the Play Sounds screen). The problem is in the COM RAM circuit and can usually be fixed by Changing R89 from 1K Ohms to 470 Ohms and/or Changing 80S from a 74LS04 to a 74F04. Regards, Jed _____________________________________________________________________________ 1 of 2 To: Rick Moncrief Fr: Jed Margolin Re: Data Retrieval System Dt: 26 June 1989 Max and I have looked at the material supplied by Pat Brosnan. We have some questions for him when he comes back July 15 but it looks like something we might be able to do. This is what it looks like: 1. They use an electronic coin mech such as the MARS or the SENTINEL. 2. There is a box called the ECHO box (Electronic Coin Handling Optimiser) connected to the elctronic coin mech's parallel interface. The ECHO box provides software control of the electronic coin mech and does the following: a. Allows the operator to set the coin modes with much more flexibility than our game software does; a. Controls the electronic mech's lock-outs to ensure that money is only accepted when the system is operational; b. Does not count coins if the door is open; c. Allows the operator to selectively disable the acceptance of certain coins (apparently slugs are a problem over there); d. Has an alpha-numeric interface; e. Has diagnostic routines to help locate faults with the electronic coin mech; AND f. Has an RS-232 output to the Datapak 2. 3. The Datapak 2 is the electronic data capture system. " It provides an automatic means of capturing and transferring information from site to a central location for management and security purposes." It has a realtime clock, an internal battery backup, and it connects to sensors to determine if a door is open. 4. The information in the Datapak 2 is collected by the Datapak 2 Auditor which is a self-powered, handheld unit that plugs into the Datapak 2. Data Dump takes about 10 seconds. 2 of 2 What it records: Name of site Serial number of machine Current Collection details comprising: Name of Machine Time and date of collection Number of individual coins in Number of individual coins out Number of games played Entry log giving each occurrence of: Time and date of entry Time and date of exit Identity of intruder Time log showing each occurrence of: Time and date machine switched on Time and date machine switched off Games played in above period The previous collection data The accumulated recording of individual coins in and out How to combine information from ECHO and the game for Datapak 2: The game has two RS-232 ports. Connect one port to ECHO and the other port to Datapak2. (The +12V, -12V, and Power Ground signals between ECHO and Datapak 2 must be connected.) Normally, the game will simply take the data received from ECHO and retransmit it to Datapak 2. Similarly, it will take data received from Datapak 2 and retransmit it to ECHO. When it needs to send data to Datapak 2 it can determine when it would be a good time to do so. If necessary it can temporarily store the transmissions from ECHO whilst it is talking to Datapak 2, and transmit it afterwards. The data is: 1200 Baud Half Duplex 1 start bit 8 data bits (LS Transmitted first) Odd parity bit 1 stop bit Our DUART supports these parameters. The first question is: What information do they want from game stats that is not already provided by ECHO/Datapak 2? The second question is: Where are they currently getting the -12V needed by the Datapak2? Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 7 July 1989 Stun Runner ----------- Last week I gave Art the inputs for Stun MultiSync, which requires marking the schematics to show what is loaded on the board and doing the parts list. It does not involve any PC layout changes. We have been bumped from the schedule by a job they are doing for Tengen. I have not been given a reschedule date. ADSP II ------- The parts list for the Stun Runner version of ADSP II is ready to be signed off. This assumes they want all six graphics ROMs stuffed. Atari Ireland ------------- I have sent Pat Brosnan a memo about: a. Having the game send additional to the Echo/Datapak 2 system; b. The bad 34010 and 34012 ICs he sent us (they were bad); c. A detailed explanation of what the GSP memory tests mean; d. The Sound Board fix if the Sound Board passes Self-Test but does not run game sounds. Jed _____________________________________________________________________________ 1 of 2 To: Pat Brosnan, Atari Games Ireland Ltd Fr: Jed Margolin, Atari Games Corp Dt: 14 July 1989 Re: Hard Drivin General Points and Queries Rick has asked me to relay his comments to you and to add some of my own. 1. STC Ten Turn Pots: We would like the price and the specifications on them. (The pots may degrade the steering.) Yes, please send us some failed pots and some new STC ones. 2. Problems with strain guages: We need more specific information on "problems". Different Strain Guages used on the Compact: The substrate was changed by Purchasing so they could get the part. All the important parameters should be similar. 3. Prototype Compacts with pot problems in the Accelerator/Brake Assembly: You can swap the assemblies with the Production Assembly but you may also need to update the software to the Production Version. 4. I will refax my memo about the Data Retrieval system. 5. OK on waiting for information from Brent Leisure. 6. I am sending the printout on all EPROMs used in Hard Drivin'. The new Sound update adds a few more sounds for the Compact Driver. The new Sound update can be used with the Hard Drivin' sitdown, but the old Sound program (for the sitdown) cannot be used with the Compact because it does not have the extra sounds. The Sound PCB with the new program is completely interchangeable between sitdown and Compact. The ADSP PCB is also completely interchangeable between sitdown and Compact. By the way, we are phasing in a new ADSP board (the ADSP II) that will also be completely interchangeable between sitdown and Compact. The reason we are phasing it in has something to do with it being used in Stun Runner and how parts are ordered. They are scheduled to be used in the last 300 Compacts so you might end up with some. I don't know if Mark Sherman is seeing the same problems with pots and strain guages. He had gone by the time I was doing this (Friday evening). New Information: I am not sure what you have so I don't know if there is anything else I should send you. Training on the Boards: I am not planning on any. Generally I prefer to write stuff down so I can think about it first. I also prefer to give written documentation to people so they can think about it, too. Thanks for the compliment. <--- {Tell Management} Regards, Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 14 July 1989 Stun Runner ----------- Two weeks ago I gave Art the inputs for Stun MultiSync, which requires marking the schematics to show what is loaded on the board and doing the parts list. It does not involve any PC layout changes. We have been bumped from the schedule by a job they are doing for Tengen. I have not been given a reschedule date. I received a list of part numbers from Ed which someone gave to him, telling him the parts were long lead time items. I generated a list of descriptions and verified that Stun Runner did, in fact, use them, except for a 40 pin connector that was attributed to SA II. I also generated a list of MANMAN Inventory from the part numbers, and many of the parts were either ordered or already in inventory. The exceptions were the ADSP 2100 in Plastic Quad Flat Pack, the PQFP Housing, and the PQFP Lid. I called Lori Robbins and she said we will be going to the ADSP II for the last 300 Compact Drivers and that the POs hadn't been entered in the system yet. ADSP II ------- I have signed off the parts list for the Stun Runner version of ADSP II. Sound Board ----------- I finally looked into the COM RAM problem. There are three options: A. Change a 74LS04 to a 74F04 and Change a 1K resistor to 470 Ohms. B. Do two cuts and two jumpers, and Change a 1K resistor to 470 Ohms. (No need to change the 74LS04.) C. Do four Cuts and four jumpers. This is the optimum solution but should be done only if we are going to buy new boards. We have already ordered the Sound Boards for the current run of the Compact Driver. The question is whether they are going to release more. Pat Brosnan ----------- I have responded to Pat Brosnan's fax of 14 July and sent him a bunch of stuff. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 21 July 1989 Stun Runner ----------- Stun MultiSync has been released. Art will do a Stun Runner MultiSync schematic showing which parts are not loaded for Stun Runner. ADSP II ------- I have given Stun Runner two more (working) ADSP II boards, for a total of six. We have: 2 working boards; 1 board with shorted traces (by the vendor); 1 blank board. Out of the 9 stuffed boards there were about four with shorted traces by the vendor. And each one was different. Perhaps the boards were not properly cleaned at some step in the process. Sound Board ----------- I have signed a deviation to be followed by an ECN to: Change a 74LS04 to a 74F04 and Change a 1K resistor to 470 Ohms. Pat Brosnan ----------- I have not yet received a reply to the memo I sent. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 26 July 1989 ADSP II ------- We have started using the ADSP II in the production of Compact Hard Drivin'. Upon investigating the report of PQFPs with bent leads: Don Wrightnour showed me a package of 110 PQFPs. On the top section (55 pieces) most had bent leads, some very badly bent. Don said he had not been able to find out how they had gotten that way but that they probably hadn't come that way from Analog Devices. The PQFP Housings and Lids have not come in. It is my fault because I hadn't alerted them that the part needed special handling. Some parts have apparently been put in standard plastic wrapping. That is apparently my fault, too, because I did not alert them that the part is CMOS and therefore static sensitive. (Not to mention the physical damage caused by throwing them into a bag.) Apparently they will either try to unbend the leads when the lids come in or they will send them back to Analog Devices for repair. I have told Don that the PQFPs must stay in the shipping package until they are put into the PQFP lid. If they need a special container to hold the lidded PQFP then so be it. Everything seems to be my fault because I did not tell anyone that this board was different from the Driver ADSP. 1. I spent a lot of time making certain that Purchasing coordinated the buy of the PC Board, PQFP, and PQFP socket assembly. 2. Perhaps either the ADSP II documentation was not published in a timely manner or it was and Manufacturing didn't read it. This problem will almost certainly be discussed in various staff meetings. I suggest we offer to design the PQFP part out and the PGA back in if Manufacturing is unable to handle the part properly. (Note, for Compact Hard Drivin' we can go back to Driver ADSP; for Stun Runner we would need to redo the board because of the extra Graphics ROM.) Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 28 July 1989 Driver ADSP ----------- I am working on test routines to find out why some boards pass Self-Test but do not run the game. ADSP II ------- We have started using the ADSP II in the production of Compact Hard Drivin'. The problem with handling the PQFPs was discussed in a previous memo. Stun Runner ----------- I have signed off the schematic for the Stun Runner MultiSync Main. Joe needed programmed part numbers for the SA II board so I gave him some. The schematic called out 27512-300 which we do not have because they are NMOS and 300 ns, but Byron approved the use of 250 ns CMOS. Soon they need to decide whether to use EPROM or OTP on the SA II. EPROM Prices ------------ Standard Current Cost Cost QOH -------- ------- ------ 137448-250 27C512-250 EPROM $6.50 $4.50 52,238 137454-250 27C512-250 OTP $5.10 $5.10 46,559 Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 4 August 1989 Strain Guage PWM Reference -------------------------- I am working on it as my first priority and am making progress. ADSP II ------- PQFP: They are about to start using the PQFP. Don reported that the PQFPs were loose in the housing. I explained that the PQFP had to be put in the Lid first; then the Lid assembly goes into the housing. PGA: A problem has developed recently where apparently the PGAs are coming in with one of the pins bent at about a 10 degree angle. They seem to be mostly the same pin each time. When it is put in the socket the pin is bent flat. Attempting to unbend the pin snaps it off. Margot came in and collected some samples to send back to the factory. Sam is inspecting each one before she hands it out (and fixing the pin). I hope Analog Devices fixes it on their end soon. 2101 ---- Tom Clark is getting me some WW Sockets. Driver ADSP ----------- Test routines are on hold while I work on the strain gauge centering circuit/ software. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 16 August 1989 Strain Gauge PWM Reference -------------------------- I have rewritten the Binary Rate Multiplier software so that the worst case has gone from 128 us to 40 us. ADSP II ------- PQFP: The first ADSP IIs have been assembled, and they work. PGA: The 2100s with the bent pins were sent back. Apparently, Analog Devices has determined that it was their error and has replaced them. 2101 ---- Tom Clark is still getting me some WW Sockets. Driver ADSP ----------- I am making progress on the addional memory tests. Stun Runner ----------- The resistors that establish the video levels (R97, R108, R112) were left off the parts list. I have done a deviation to put them on. An ECN will be done after the Pre-Prods are built. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 25 August 1989 2101 ---- Tom Clark is still getting me some WW Sockets. Driver ADSP ----------- I am making progress on the additional memory tests and am working on the complete memory tests performed by the 2100. Stun Runner ----------- Various omissions are being corrected in the latest rounds of ECNs. I have asked Jeff Bell for a series number for the Stun Runner OTPs. 136070 will be used for EPROMs. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 25 August 1989 Driver ADSP ----------- I have completed the following tests; most can be performed by both the 68010 and the 2100: Output Buffers: Test that the Output Buffers can be switched. Shift various bits through data and address line and display the bits onscreen. Test buffer with data equals address. Test buffer with pseudo-random data. Data Memory: Shift various bits through data and address line and display the bits onscreen. Test data memory with data equals address. Test data memory with pseudo-random data. 2100 does standard complete tests and reports the results. Remaining: Program Memory Tests; Redraw the board onscreen to look like ADSP II. Stun Runner ----------- Various omissions are being corrected in the latest rounds of ECNs. I have asked Jeff Bell for a series number for the Stun Runner OTPs. 136070 will be used for EPROMs. 2101 ---- Tom Clark is still getting me some WW Sockets. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 5 September 1989 ADSP II ------- I have completed the following tests: Output Buffers Data Memory: Bus tests and standard complete test performed by 2100 Program Memory: standard complete test performed by 2100 Remaining: Program Memory bus Tests; Redraw the board onscreen to look like ADSP II. Stun Runner ----------- Ed has given me the final configuration of EPROMs and OTPs for initial production. Rick Meyette has given me the EPROM/OTP numbering scheme where we use get to do the checksum/part number thing while using only one programmed part series. Art is working on the ECNs. 2101 ---- Tom Clark is still getting me some WW Sockets. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 8 September 1989 ADSP II ------- I have completed the additional ADSP Tests. I have added the tests to the Compact Driver format and will give a pair of ROMs to Production and Customer Service. There are no plans to release a new version to retrofit games in the field. Also, Self-Test is now too big to coexit with the game program in ROM 0. Stun Runner ----------- I have signed off the final production parts lists for the PC Boards. I have given Ed and Andy the final Self-Test which includes the new ADSP tests. 2101 ---- Tom Clark is still getting me some WW Sockets. Jed 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 15 September 1989 ADSP II ------- I have given Test ROMs to Production and Customer Service. (They do not run the game, just Self-Test.) I have also given them documentation for the tests. Stun Runner ----------- They have started Final Assembly on Stun Runner. There was a problem with the game playing the wrong sounds that was due to their not having received the latest ECN in a more timely fashion. The ECN was to change what ROMs went where. I have reviewed the final draft of the manual and made corrections. 2101 ---- Tom Clark has gotten me the WW Sockets for the 2101. Hyundai 8Kx8 SRAMs ------------------ Tom Clark has disqualified Hyundai 8Kx8 RAMs from the AVL. Even so, we will be using sockets for the 8Kx8 SRAMs on the ADSP Board and also on the MultiSync Board. Tom Clark and I met with Steven Hayes of I Squared and W.S. (Douglas) Chee of Hyundai. They said they knew about the pattern sensitivity problem with their RAMs. They claim the low power version of the part is probably ok and they will either replace our inventory with said part or just take 'em back. ADSP-2101 --------- Analog Devices is changing the signal pinout of the PGA so that it will be compatible with the footprint of the socketed PLCC. Cellular Phones --------------- I have started getting information on cellular phones. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 22 September 1989 Atari Ireland ------------- I have sent to Pat Brosnan: Documentation for the new ADSP Tests; Special Test ROMs for Compact Hard Drivin'; A 34010 Data Manual; Stun Runner ----------- They are producing boards and have started Final Assembly today. Hyundai 8Kx8 SRAMs ------------------ Tom Clark, Mary Burnias, and I had a meeting with the Hyundai Senior Manager of the Reliability Engineering Department (Mr. Seung-Il Kim), the Regional Sales Manager (Mr. Patrick Bonham), the Applications Engineering Manager (Mr. W.S. Douglas Chee), and the I-Squared Manufacturers rep (Mr. Steven Hayes). The results: 1. I agreed to provide a timing diagram for my RAM circuit. I actually did one better by taking pictures of the 'scope display. I have given the write-up to Tom to give to them. 2. Tom has agreed to test 100 samples of their low power part. They will be tested on my ADSP Board with my ADSP tests. (Jacob has had my board modified with zero-insertion force sockets.) Tom will also have some tested in Production. If they are all good presumably we will swap the parts we have for the new parts. If any one part fails he will reject the entire lot and Hyundai will buy back our inventory of their parts. ADSP-2101 --------- Analog Devices is changing the signal pinout of the PGA so that it will be compatible with the footprint of the socketed PLCC. Cellular Phones --------------- My initial survey is complete. Street Driver Amp ----------------- I have received from Karen an APU modified into a single supply bridge amplifier. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 29 September 1989 Stun Runner ----------- Carole will ECN the harness so the Left and Right Coins Switches are correct. Street Driver Amp ----------------- Bridge Amp Performance: Sine Wave Load Power Output Current from +12VDC Supply The Bridge Amp will produce 4 Watts (Sine Wave Power) into 8 Ohms and 6 Watts into 4 Ohms. The standard amp run from +/- 14 VDC produces about 6 Watts into 8 Ohms. I am ready for PC. AR340 ----- I have generated the cost figures for a low cost 34010 game system and am waiting for TI to get back to me so I can find out what the ICs will cost next year. Transputer ---------- I have received data books from Bill Woods. SGS Thomson now owns INMOS and has announced a price reduction on Transputers in order to increase market share. In quantities of 1K the T800 will cost $135.00 . Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 6 October 1989 Stun Runner ----------- I have not been notified of any problems so I will assume everything is copacetic. Street Driver Amp ----------------- I have added a connector to supply +12V for the Electronic Coin Mechs in Europe. I have asked Tom Clark to find a knob for the volume control pot. I have given Leon the inputs. Leon has done the schematic and is working on the board. Transputer ---------- I am still reading the user's manual. Cycle-Tron ---------- I have made a new Opto-State PROM so that it will act as a tachometer with only one input (Q0), instead of requiring two quadrature signals. It will give half the number of counts but will allow the use of a Single OSR Board ($8.17) instead of a Dual OSR Board ($9.87). Parametric Horizon ------------------ I expect to start designing a test circuit soon. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 6 October 1989 Stun Runner ----------- I have not been notified of any problems so I will assume everything is copacetic. Street Driver Amp ----------------- I have added a connector to supply +12V for the Electronic Coin Mechs in Europe. I have asked Tom Clark to find a knob for the volume control pot. I have given Leon the inputs. Leon has done the schematic and is working on the board. AR340 ----- I have gotten revised cost figures for ICs from Jeff at TI. AR340 Board, Version 1 (Fully Loaded 68000-based 34010) $ 218.82 $ 203.95 (with Gate Array) $ 203.82 (with 40 MHz 34010) $ 188.95 (with 40 MHz 34010 and Gate Array) AR340 Board, Version 2 (Partially Loaded 68000-based 34010) $ 194.91 $ 180.04 (with Gate Array) $ 179.91 (with 40 MHz 34010) $ 165.04 (with 40 MHz 34010 and Gate Array) AR340 Board, Version 3 (34010 only) $ 128.95 $ 123.95 (with Gate Array) $ 113.95 (with 40 MHz 34010) $ 108.95 (with 40 MHz 34010 and Gate Array) Transputer ---------- I am still reading the user's manual. Cycle-Tron ---------- I have made a new Opto-State PROM so that it will act as a tachometer with only one input (Q0), instead of requiring two quadrature signals. It will give half the number of counts but will allow the use of a Single OSR Board ($8.17) instead of a Dual OSR Board ($9.87). Parametric Horizon ------------------ I expect to start designing a test circuit soon. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: 13 October 1989 Stun Runner ----------- I have not been notified of any problems so I will assume everything is copacetic. Street Driver Amp ----------------- We have gotten film and have presumably sent out for boards. AR340 ----- I have finished the schematics of the Main Board and have started on the 2101 board. PC will not be ready for them until Nov 6. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: November 3, 1989 Stun Runner ----------- I have not been notified of any problems so I will assume everything is copacetic. Street Driver Amp ----------------- We have received the 15 boards. Karen has built two that we will test presently. Tom has found a knob for the volume control: Cosmo #39-7902-020-00-3-10 Setup up charge is $125. Cost is $0.98 . Minimum order is 5,000. (The Pot shaft is 3/16 and this is apparently nonstandard.) $0.98 sounds like a lot for a knob. (The pot is only $0.64 .) AR340 ----- I have finished the schematics of the Main Board and the 2101 board which will go by the name "DS II". I will give them to PC on Nov 6. Rich Comeau of Motorola is getting me some 35 ns SRAMs for the DS II board. Jed _____________________________________________________________________________ 1 of 2 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: November 10, 1989 Street Driver Amp ----------------- The amplifier seems to work. With +12V power it delivers: 18Vp-p no load 15Vp-p at 8 ohms, = 7.5Vp = 5.3Vrms. power = 5.3*5.3 / 8 = 3.5 Watts Supply current is 0.7 Amps. 13Vp-p at 4 ohms, = 6.5p = 4.6Vrms. power = 4.6*4.6 / 4 = 5.2 Watts Supply current is 1.2 Amps. At 4 ohms each amplifier is effectively seeing a 2 ohm load and has a peak current of 1.6 Amps. The amplifiers are rated at 3.5 Peak Amps max. The standard amplifier runs on +14V and -14V, is single ended, and is used with an 8 Ohm speaker. It delivers 22Vp-p into 8 Ohms (7.5 Watts). Since there are two channels this is a total of 15 Watts. Bridge Amp Referenced to 15 Watts ---------- ---------------------- 1 x Eight Ohm Speaker 3.5 Watts -6.3 dB 1 x Four Ohm Speaker 5.5 Watts -4.3 dB 2 x Eight Ohm Speakers 5.5 Watts -4.3 dB 1 x Eight Ohm Speaker/channel x 2 channels 7 Watts -3.3 dB 1 x Four Ohm Speaker/channel x 2 channels 11 Watts -1.3 dB 2 x Eight Ohm Speakers/channel x 2 channels 11 Watts -1.3 dB The DS II Board has two channels and so is capable of delivering 11 Watts into 2 four Ohm speakers. Multiple speakers are generally more efficient acoustically than single speakers. Acoustic Output in Driver Sitdown playing the 320 tune: Bridge Amp 100-106 dB Standard Amp 104-110 dB This agrees with the power differences between the two amplifiers. The Bridge Amp delivers 5.5 Watts into 2 eight Ohm speakers. The standard amp delivers 7.5 Watts/channel into eight Ohms; with two channels there are 15 Watts of audio. 2 of 2 Whether a single Bridge Amp producing 5.5 Watts into a single four Ohm speaker is loud enough depends on the acoustic design of the cabinet. In the Sitdown it produces a peak level of 106 dB which is certainly sufficient. AR340 ----- I have finished the schematics of the AR340 board, the DS II Board, and the DS II RAM board. Joe is currently working on the DS II board. Rich Comeau of Motorola is getting me some 35 ns 8K x 8 SRAMs for the DS II board. AT&T DSP32C ----------- Nilesh Amin of AT&T and Terry Speizer of Electec Sales (the rep) came in and gave a presentation on the DSP32C. The DSP32C does 12.5 MIPs and has a single cycle floating point multiplier that has a mode where it can do four floating point multiplies per cycle. The DSP32C costs about $90 (1K) for the 164 pin PQFP. Theye weren't sure if that was for the 40MHz or 50 MHz part. They offer software tools that run under MSDOS. The assembler and linker and simulator costs $500, the C compiler costs $1000. They also have a simple ICE that connects to the PC for $270; the Pod costs $540. Terry will get me more information and three more sets of data books. It has 1.5K of internal RAM. The downside is that is uses fast external memory: 25 ns. This device looks cheaper and faster than the Transputer. Jed _____________________________________________________________________________ 1 of 4 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: November 17, 1989 Driver Sound ------------ I have worked out the mods so Manufacturing can use 27256s for the Wave Table ROMs. It will be transparent to the software; however, checksums will now be for pairs and the reported locations will be wrong. We should send out a sheet explaining the new locations. Reported location Locations with 27256s ----------------- -------------------- 65A 65A, 55A 55A 45A, 30A 45A 20A, 5A 30A 65C, 55C 20A 45C, 30C 5A 20C, 5C A bad checksum means that one or both 27256s in the pair are bad. We should have Karen mod a board. (Someone will have to make the ROMs.) DS II ----- Joe was doing a great job on the board until we got dumped by KLAX. DS II RAM --------- Joe started on the schematics while he was doing DS II. AT&T DSP32C ----------- I am ready to start drawing the schematics for the MSP Board. Street Driver Inputs -------------------- Gas J8 - 7 A/D IN3 Brake J8 - 5 A/D IN4 Steering J9 - 6 A/D IN0 J9 - 5 A/D IN1 (not currently used) Jed 2 of 4 DSP32C ------ The 40 MHz part will require memory speeds of either: 25 ns with 1 wait state or 45 ns with 2 wait states Using 45 ns RAM permits the use of banked 8Kx8 SRAMs. A normal instruction cycle is 4 clocks (100 ns). This is apparently 2 clocks for reading the instruction and 2 clocks for reading or writing memory. Note that if successive memory accesses are to the same physical memory a conflict wait cycle is inserted. [from "WE DSP32C Digital Signal" page 2-16] Internal RAM and External RAM are different physical memories. It is not clear whether Internal RAM 0, RAM 1, and RAM2 memory are considered physically different. >>> An instruction fetched from external memory that accesses internal RAM is 6 cycles. At 40 MHz this is 150 ns which is: 6.7 MIPS (also at least 6.7 MFLOPS) ------------------------------------------------------------------------------ DSP32C Memory Calculations -------------------------- 40 MHz T = 25 ns N = Wait States N = Wait States ----------------- 0 1 2 ASN (Address Strobe) T - 5 + N x T 20 ns 45 ns 70 ns Data Setup 10 ns (Assuming this is the same as for the 50 MHz part) Memory access 10 ns 35 ns 60 ns Memory Byte Select must be strobed with ASN Assume 6 ns for a 74AS139 4 ns 29 ns 54 ns Memory speed - 25 ns 45 ns ------------------------------------------------------------------------------ 3 of 4 DSP32C Memory Calculations -------------------------- 50 MHz T = 20 ns N = Wait States N = Wait States ----------------- 0 1 2 ASN (Address Strobe) T - 5 + N x T 15 ns 35 ns 55 ns Data Setup 10 ns Memory access 5 ns 25 ns 45 ns Memory Byte Select must be strobed with ASN Assume 6 ns for a 74AS139 -1 ns 19 ns 39 ns Memory speed - 15 ns 35 ns ------------------------------------------------------------------------------ 4 of 4 Changing Driver Sound Board Wave ROMs from 27512s to 27256s ----------------------------------------------------------- Method 1 (Top of board): 1. Cut and lift 55E pin 7. 2. Connect a wire jumper from 45C pin 1 to 55C pin 28. 3. Cut and lift 100C pins 12, 15, 16, and 19. 4. Connect a wire jumper from 55E pin 7 to 95A pin 1. 5. Connect a wire jumper from 100C pin 19 to 95A pin 2. 6. Connect a wire jumper from 100C pin 16 to 95A pin 3. 7. Connect a wire jumper from 100C pin 15 to 95A pin 5. ------------------------------------------------------------------------------ Method 2 (Neater but requires cutting traces on the bottom of the board): 1. On the bottom of the board, Cut the trace going to 55E pin 7. 2. On the bottom of the board, Cut the traces going to 100C pins 12, 15, 16, and 19. On the top of the board: 3. Connect a wire jumper from 45C pin 1 to 55C pin 28. 4. Connect a wire jumper from 55E pin 7 to 95A pin 1. 5. Connect a wire jumper from 100C pin 19 to 95A pin 2. 6. Connect a wire jumper from 100C pin 16 to 95A pin 3. 7. Connect a wire jumper from 100C pin 15 to 95A pin 5. ------------------------------------------------------------------------------ Method 3 (Maybe): 1. Install a programmed PAL at 100C in place of the 74LS374. 2. Cut and lift 55E pin 7. 3. Connect a wire jumper from 45C pin 1 to 55C pin 28. 4. Connect a wire jumper from 55E pin 7 to 95A pin 1. This method should be done only on new boards so that the PAL at 100C can be installed before wave solder. _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: December 1, 1989 27256s ------ I have apprised Jim Freitas and Norm Maeder of the difficulties in modifying the MultiSync Main board and Driver Sound Board to use 27256s and of the problems that would be caused when we ship the Hard Drivin' 90 kit. They agreed it wasn't a good idea for 200 games. I told them about the DS II Board's use of a 27256 and they also agree we should hang on to our stock of these parts. DS II ----- Joe was doing a great job on the board until we got dumped by KLAX. We are still dumped by KLAX. Estimated film and fab for DS II is now 12/18/89. DS II RAM --------- Joe started on the schematics while he was doing DS II. AT&T DSP32C ----------- I have given Leon the inputs for the board. I have given Erik the inputs for the wirewrap. I am adding the RAM and ROM to the DSP32 board as requested. TI 320C30 (floating point) -------------------------- Unfortunately, the chip is currently too expensive. Engineering samples are $500. If we ordered 1K in January for delivery in March, Jeff thinks he could get them to us for $200. Jeff estimates that in mid March they will be $120-$150 and in 1Q '91 maybe $90 - $100. To: Rich Moore Fr: Jed Margolin Dt: 1 December 1989 Re: xxxxx There is a possibility that someone is preparing to copy Hard Drivin'. There is no hard evidence, just something that caught my attention. According to a reliable source xxxxx has ordered 600 ADSP-2100s (8MHz PGA). The parts will be billed to 'xxxxx' and they are being sent to xxxxx. The ADSP_2100 is one of the few parts in Hard Drivin' that is sole sourced and available only from the manufacturer. Most of the parts in Hard Drivin' either have multiple sources or are available from distribution. Someone could buy these parts without attracting any attention. The two exceptions are the TWIG custom part from National Semiconductor and the ADSP-2100 from Analog Devices. The ADSP-2100 is used in Hard Drivin', Compact Hard Drivin', and Stun Runner. We started with the 8 MHz part in the Pin Grid Array (PGA) package, which was the only version available when we designed the Hard Drivin' hardware. Since then, Analog Devices has come out with faster versions (12 MHz) in less expensive packages like the Plastic Quad Flat Pack (PQFP) that we use in the ADSP II board that was phased in during the Compact run in August and was also used in Stun Runner. Analog Devices also makes an improved part (ADSP-2100A) and a cost reduced version (ADSP-2101) that we will be using the Street Driver. These last two are not interchangeable with the original ADSP-2100 that we used. Someone designing a new product would probably use one of the newer parts, not the 8MHz PGA that xxxxx has ordered. It is, of course, very possible that they are buying these parts for a legitimate purpose. I recommend that we not do anything, just be alert for counterfeit Hard Drivin' games. xxxxx would know how we found out (actually it was xxxxxxxxxxxxxx) and I don't want her to get hurt. Enclosed is a copy of xxxxx's Fictitious Business Name Statement filed 12/23/87 in Santa Clara County. xxxxx is currently located at xxxxxxxxxxxxxxxxxxx. P.S. There are companies that are in the business of "peeling" custom chips like the TWIG so they can be copied. It is expensive and has probably kept the game from being copied so far. _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: December 8, 1989 DS II ----- I believe we have sent out for film and fab. DS II RAM --------- I have checked Joe's schematic and he has started the board. AT&T DSP32C ----------- We have received the WW from Emmette and I am in the process of checking it out before giving it to Max. Motorola is sending me a quote on the 45ns 32Kx8 SRAMs. I have received a checkprint from Leon. Other ----- I have verified that the documentation for Hard Drivin' sitdown calls out the ADSP II board. On Wednesday I spent about 5 hours supporting the 34010 Overlay Processor that I did for System 3D. I spent about 1 hour with Sam Lee and about 4 hours generating documentation for him to use. On Thursday I spent about 1 hour with Sam Lee. On Friday, Max and I spent some time with Ed Logg discussing aerodynamic simulation. Jed _____________________________________________________________________________ 1 of 2 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: December 18, 1989 We have film and fab and are ready to buy the following boards: DS II (has been sent to the vendor) DS II RAM AT&T DSP32C Melanie's last day before vacation is Wednesday. Really Low Cost ADSP-2101 (They don't have a part number yet) -------------------------------------------------------- According to Damon Chu of Analog Devices it will have: 1. 1024 words of program RAM (24 bits); 2. 256 words of Data RAM (16 bits); 3. Only one serial port; and it will: 4. Be pin compatible with regular 2101 (and have access to external memory); 5. Run at 100 ns (10 MHz); 6. Be sampled in April 1990; 7. Cost $10 (quantities 1 - 1M) AT&T DSP32C ----------- After getting a checkprint from Leon I solicited input from Manufacturing (Norm, Tim, Don, and Kurt) because the DSP32C is the largest (164 leads) fine pitch (25 mil) SMD we have used to date. Tim, Norm, Don, and I met with Art and Leon and had a good discussion of what we needed to do for this device, after which Leon made some layout changes. One of the more interesting facts that came out of this discussion was that we need to have the board done with the solder mask over bare copper (not tinned). The reason is that in attaching the SMD, the tinned solder plate has a tendency to melt, damaging the solder mask. It can also cause shorts between the teeny tiny traces. The only downside to having solder mask over bare copper is that if the boards are stored for a long time before they are built, the copper can corrode. Since we usually build the boards as soon as we get them this will not be a problem. 2 of 2 We will also need a separate mask for the vendor to apply the solder paste for attaching the SMD. Also, Don says that the SMD station that Customer Service has will not work with our 164 lead device. Mark Sherman agrees; his rework station would not be able to do an IC the size of ours. This means that if the DSP32C has to be removed for rework it will mean sending it out unless we get a new rework station, which is unlikely. (I hope these boards don't need a lot of rework.) I have received a quote from Motorola for 45ns 32Kx8 SRAMs. for 1990: $18.00 (1K) $17.00 (5K) CERs ---- I have submitted CERs for: ADSP-2101 DSP32C 12.000 MHx crystal for DS II Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: December 22, 1989 We have film and fab and are probably buying the following boards: DS II; DS II RAM; and DSP32 DSP32 WireWrap -------------- Max has fired up the board and it seems to work, so far. TWIG ---- According to Mark Sherman we have only sold 13 loose TWIGS. I told him about the problem and he will have the part flagged. Parts ----- I have received part numbers for ADSP-2101 and DSP32C. Tom has flagged the DSP32C as possibly being subject to export restrictions. Jan Bautch used to handle it but she left. I discussed it with Rick Meyette and Tom Clark and Rick agreed that they would handle it. Information from Margot via Max: 2101 - Cheapie -------------- Samples in April ('90), Production quantities in May ('90). 2101 - Regular -------------- 2101 has 6 week lead time PGA ADFP 2101KG-50X $161.00 (ouch) PLCC ADFP 2101KP-50X $161.00 (ouch) (I think these numbers should probably be ADSP, not ADFP) We only have one sample. We need at least ten more, with at least some of them in the PLCC. (The board can take either package.) ROM version available 4Q 90 for 5K/year $10K NRE + 15% price It looks like the internal ROM can be read out very easily by a program in external memory. Jed _____________________________________________________________________________ 1 of 1 To: Rick Moncrief Fr: Jed Margolin Re: Hardware Status Dt: January 3, 1990 DS II ----- We have received the DS II Board and it is being stuffed. Driver Sound Mods to accept 27256s ---------------------------------- The mods have been tested and seem to work. I will give them to Art. I will do a cross reference chart because the 27256s do not correspond to the locations reported by Self-Test. The 27256s are made from, and contain, exactly the same data as the 27512s. However, someone will need to officially release them. We will also have to figure out Programmed Part Numbers. DSP32 Board - 320E14 -------------------- I talked to Lisa Johnson from TI about programming the Security Bit (which they call the 'R Bit'). With the other EPROM versions (320E15 and 320E17) the EPROM socket adaptor contains an R Bit programmer. The EPROM socket adaptor for the 320E14 does not. I called Data I/0 and talked to Tom Jett. He confirmed that their programmers do not support programming the R Bit on the 320E14. He said he would put the request on a list and send it somehere. He agreed that the fact that these other DSP EPROMs that had been out for a while were not supported did not bode well for the 320E14. Here are the choices: 1. Use it without setting the R Bit (no security). 2. Wait for Data I/O to support the 320E14. 3. Wait for TI to do an EPROM Socket/R Bit Programmer for the 320E14. 4. Use a 320E15 or 320E17. 5. Don't use anything. 6. Do our own R Bit programmer. Purchase Reqs ------------- I have filled out purchase reqs for 2101s and DSP32Cs. I need to know the answer to the 320E14 question before filling one out for it. Jed _____________________________________________________________________________ { Date Unknown } The following things remain to be done in order for Stun Runner Self-Test to be complete: 1. Turn on checksums for the program ROMs that will be used; 2. Use final Sound ID number for the test sound; 3. Integrate the ROM with the Stun Runner program. (Self-Test uses about half of the ROM 0 pair.) I have specified Slapstic 17. According to MANMAN we have 2,745 Slapstic 17s in inventory and the part was assigned only to the Driver, which did not use it. If it is changed for any reason it will be necessary to change Self-Test which works through Slapstic to get ROM 7 checksums. Don Paauw is on vacation this week and therefore unavailable for comment if we are to release the Hardware this week. I have a list of the pulls and keys for Slapstic 17 if you want it. Jed _____________________________________________________________________________