*** Note: I formatted this on a PC. Not sure how it will show up on a phone ***
A repair shop that I've done some work for asked me if I would take a look at Sprint 2. I've only worked on a few bronze age games. One thing about them is you can see how the golden age games evolved from the earlier ones. Ex: All of the power conversion and regulation happens on these early boards. Atari creating the ARI and the ARII did a couple of things.. removed a common failure point (power) from the logic board and freed up real estate for more game logic. There are a number of things like this I noticed and a number of 'experience points' gained.
Fortunately the repair shop had 4 boards to send me.. A working reference board (it didn't immediately work btw). Their customer's board and 2 more they had that I could use as parts boards or preferably have me fix them too.
Sprint 2 is one of the first mass produced games to use a CPU. It has a 6502 processor. This was in 1976.. I was teaching myself 6502 assembly language on a Commodore 64 in the early 80's.. That is a long history for a processor by todays standards.
Working on these was a bit of a blur .. figuring out how it worked and some of the quirks slowed me down here and there..
First up was getting something on the screen - which requires a bench harness. I wasn't able to use my regular test rig (easily) so this is more of a one-off test harness than usual.

It started real simple - 5v+ connected from a bench supply and a composite video cable. I'll show the completed harness here, but this happened after getting the customer board working to the point where I needed ALL of the controls to continue testing..



I'd added connections for center tapped transformers to make sure the on-board power regulation worked. The bench supply got used for the 5v+ in the beginning, but I'd found a situation where I wasn't easily able to work around not supplying AC to the board. The AC setup works well enough and I had an old Pacman transformer that works great. My trackball is used for steering.. Horizontal for the white car and vertical for the black car. I can't drive 2 cars at once anyway, I just need to know they work.. All of the other controls took a few hours to get wired up and debugged. Good thing Amazon exists.. I was able to get some switch assortments to make up all of the bits I need to test with..


Trackball for steering - small CP for Coin 1,2 Track select, Test switch, P1/2 start, gas pedal (on/off on this game) and gears 1,2,3.. 4th gear is all switches off. The kit did not have a switch for that.. but if 1,2,3 work (all 1's) the 4 will work since the 0's work in the first case (hope that makes sense).. All game functions now accounted for..
Board #1 - Reference board
This one was sent as a 'working reference' but didn't immediately work - in the end it just behaved differently than the other boards (+a couple of glitches) I did not work on this one first.. I spent the majority of the time working on Board #2 before coming back to this one.
When I finally came back to it after getting the harness all sorted out, it locked up when a credit was added. The screen just froze and the credit LEDs did not light. It also locked up if I put it into free play.. this seemed like a software issue of course.. I retested the ROMs.. no problems. I even started reading source code.
While doing all of this - I followed the watchdog to see if it was frozen or firing.. I determined the watchdog circuit was not working.

Replacing the 7437@R9 fixed the watchdog and now instead of just freezing, the game would watchdog reset when a credit was added or I put it into free play - that at least made sense. Attract mode and even test worked correctly with all of the switches working as expected.. Somewhere along the way the source code made me want to look at this area..

I ended up comparing this to a working board. The input on @A5 Pin1 was ~4.3vdc on this board that locks up and ~1.3vdc on the working boards. The difference in the voltage caused the output on Pin2 to flip from high/low depending on the voltage. Keep in mind at this point I had bypassed the AC inputs and was using 5v bench power. Board 2 and Board 3 were working. I checked the resistors and replaced the diode and C1 - I'm not sure where/how the boards were different. I ended up wiring AC power (removing the 5v bench supply from the equation) and board stopped crashing when adding a credit. I'm not really sure why Board2,3 work since with AC power, the output @A5 pin2 is a 60Hz square wave when connected to AC but it is a steady 5v when connected to DC. The code definitely does not like D7 being low when adding a credit. I didn't really figure out what the purpose of that D7 line was..
During all of this I also found that the 4 ohm/10W resistor across the voltage regulator was flaky. The resistance would change anywhere from 14 ohms to 90 ohms just by hitting it.. That got replaced.
Board works!
A repair shop that I've done some work for asked me if I would take a look at Sprint 2. I've only worked on a few bronze age games. One thing about them is you can see how the golden age games evolved from the earlier ones. Ex: All of the power conversion and regulation happens on these early boards. Atari creating the ARI and the ARII did a couple of things.. removed a common failure point (power) from the logic board and freed up real estate for more game logic. There are a number of things like this I noticed and a number of 'experience points' gained.
Fortunately the repair shop had 4 boards to send me.. A working reference board (it didn't immediately work btw). Their customer's board and 2 more they had that I could use as parts boards or preferably have me fix them too.
Sprint 2 is one of the first mass produced games to use a CPU. It has a 6502 processor. This was in 1976.. I was teaching myself 6502 assembly language on a Commodore 64 in the early 80's.. That is a long history for a processor by todays standards.
Working on these was a bit of a blur .. figuring out how it worked and some of the quirks slowed me down here and there..
First up was getting something on the screen - which requires a bench harness. I wasn't able to use my regular test rig (easily) so this is more of a one-off test harness than usual.

It started real simple - 5v+ connected from a bench supply and a composite video cable. I'll show the completed harness here, but this happened after getting the customer board working to the point where I needed ALL of the controls to continue testing..



I'd added connections for center tapped transformers to make sure the on-board power regulation worked. The bench supply got used for the 5v+ in the beginning, but I'd found a situation where I wasn't easily able to work around not supplying AC to the board. The AC setup works well enough and I had an old Pacman transformer that works great. My trackball is used for steering.. Horizontal for the white car and vertical for the black car. I can't drive 2 cars at once anyway, I just need to know they work.. All of the other controls took a few hours to get wired up and debugged. Good thing Amazon exists.. I was able to get some switch assortments to make up all of the bits I need to test with..


Trackball for steering - small CP for Coin 1,2 Track select, Test switch, P1/2 start, gas pedal (on/off on this game) and gears 1,2,3.. 4th gear is all switches off. The kit did not have a switch for that.. but if 1,2,3 work (all 1's) the 4 will work since the 0's work in the first case (hope that makes sense).. All game functions now accounted for..
Board #1 - Reference board
This one was sent as a 'working reference' but didn't immediately work - in the end it just behaved differently than the other boards (+a couple of glitches) I did not work on this one first.. I spent the majority of the time working on Board #2 before coming back to this one.
When I finally came back to it after getting the harness all sorted out, it locked up when a credit was added. The screen just froze and the credit LEDs did not light. It also locked up if I put it into free play.. this seemed like a software issue of course.. I retested the ROMs.. no problems. I even started reading source code.
While doing all of this - I followed the watchdog to see if it was frozen or firing.. I determined the watchdog circuit was not working.

Replacing the 7437@R9 fixed the watchdog and now instead of just freezing, the game would watchdog reset when a credit was added or I put it into free play - that at least made sense. Attract mode and even test worked correctly with all of the switches working as expected.. Somewhere along the way the source code made me want to look at this area..

I ended up comparing this to a working board. The input on @A5 Pin1 was ~4.3vdc on this board that locks up and ~1.3vdc on the working boards. The difference in the voltage caused the output on Pin2 to flip from high/low depending on the voltage. Keep in mind at this point I had bypassed the AC inputs and was using 5v bench power. Board 2 and Board 3 were working. I checked the resistors and replaced the diode and C1 - I'm not sure where/how the boards were different. I ended up wiring AC power (removing the 5v bench supply from the equation) and board stopped crashing when adding a credit. I'm not really sure why Board2,3 work since with AC power, the output @A5 pin2 is a 60Hz square wave when connected to AC but it is a steady 5v when connected to DC. The code definitely does not like D7 being low when adding a credit. I didn't really figure out what the purpose of that D7 line was..
During all of this I also found that the 4 ohm/10W resistor across the voltage regulator was flaky. The resistance would change anywhere from 14 ohms to 90 ohms just by hitting it.. That got replaced.
Board works!
















