Light Outrun Restoration

Does anyone recognize this chassis or can tell me where the part number might be? I can only find a serial number. From my searches it is probably a Nanao MS7-?? but I want to confirm before I try to find a cap kit.

1000049740.jpg

The tube has a part number, but that is the tube, not the chassis.


1000049737.jpg
 
I confirmed the part number when I pulled the PCB to recap it.

1759782327924.png

So after recap, I pulled all the socketed IC's, cleaned them with a dremel/wire wheel, and sprayed deoxit on the sockets.

Only 1 casualty... If there is a more elegant way to do this, let me know...

1759782447686.png


With the monitor recapped and clean PCB, Harness, power supply, pedals, and steering, here goes what we call at work, "the smoke test".

Bench.jpg

No smoke, success... not much of a game though.

20251005_204158.jpg

Rebooting for kicks, brought up a messy looking service screen acting like the steering was stuck to one side. Time for the RAM test.

20251005_163914.jpg

Sometimes the RAM test flakes out and appears to reboot.

20251005_211917.jpg


Sometimes, it doesn't flake out, says it will restart, but the countdown goes to zero and then cycles through random characters.

Looks like 4 problems. IC130, IC62, and IC64 seem easy to find on the board, but not sure what MSBx and LSBx are. Most/Least signifcant byte? The first column says "ROAD" which matches up with the "usage" in the docs having this note....

*) Road RAM is swapped by IC#7, but has no defined startup state.
You should measure to get a definitive answer there.

Is this some sort of page flipping? What should I measure?



20251005_203949.jpg



Few more questions...

- Does AB point to issues with the the addressing logic, not the RAM? Or possibly both?
- Are those error codes or addresses where the fault occured? IC130 looks like an address in that RAM, but the others are not in the address range of those chips.

IC# Type Size Usage Address
IC130 TMM2063P-10 8KB x 8 Main work Region 0: 0x60000
IC21 or 39* TMM2115BP-10 2KB x 8 Road Region 5: 0x80000
IC20 or 38* TMM2115BP-10 2KB x 8 Road Region 5: 0x80001
IC21 or 39* TMM2115BP-10 2KB x 8 Road Region 5: 0x80000
IC20 or 38* TMM2115BP-10 2KB x 8 Road Region 5: 0x80001
IC64 HM65256 32KB x 8 Tile Region 1: 0x00000
IC62 HM65256 32KB x 8 Tile Region 1: 0x00001
 
AB is address bus, DB is Data Bus, DEV (IIRC) reports an error when it reads the RAM.

I would worry about the IC130 first, that would prevent anything from happening. AB is shared across all the main RAM, so I'd suspect the specific IC is bad. You can try piggybacking with compatible RAM.

Once you fix the main ram, proceed to the Road/Tile ram, IIRC I don't think these would prevent the game from booting.

Stepwise is the way to go…

p
 
Last edited:
Thx, i dont have a spare RAM, but I thought it was worthwhile check continuity between IC130 and IC115. All the address lines were around 600 ohms which didnt seem bad. I didnt have the schematic open but im guessing a chip select is used for the lsb of the address. The data lines were all around 10K between them. Not sure what is connecting them but it is consistent and IC115 isn't failing. Also used a logic probe on the address lines to see if there was a notable difference as well as look for dead data lines. Nothing obvious, so ill order some RAM.

Does piggybacking only work if there is a dead data output? I was trying to think how it would work with a dead address input. I would think the data would be a logical or of 2 different addresses.
 
Stepwise is the way to go…

Ummm... I must have taken a step in the wrong direction....

I ordered this replacement on Mouser. The pic of the IC looked correct, but when I got them, they were the wider 0.6" instead of the 0.3"...


Yes, I know the correct side parts are available at APAR, but Mouser doesn't have a minimum order. If I were smarter and read the datasheet instead of looking at the picture, I would have ordered the correct part at APAR. But since I didn't, the cheaper route than returning them was to buy a simple conversion board.

1762094142369.jpeg

When socketing the IC130, I was proud my novice soldering skills I hadn't lifted any pads or traces, until I got to pin 28. I didn't lift the pad, but a leg broke off in the plated through hole. I don't have a hot air station, and it would not come out with sucktion, so I filled the hole with solder and patiently heated up a lead cut from a capacitor until I was able to push it out. I'm curious if there are any other tricks to use in this situation.

Double checking my work, I caught that I had broken a trace connected to pin 22. It was simple enough to run a wire wrap to an adjacent RAM that shares that signal. I cleaned up the area of the break to make sure there was no loose copper and moved on with installing the socket.

1762095122557.png

New socket and RAM, I powered it up to a black screen, that had previously been the test ROM screen, so I immediately powered it down. Checked the voltage, then went back to checking my work. I found I missed a small fragment of the broken trace, and it appears to have been shorting pin 28 (Vdd) to pin 27 (WE*). I thought for sure I checked throurouly, but apparently not enough.

I removed the copper fragment and powered it up to a black screen again. While I was thinking about how to approach this, I saw some signed of life on the screen.


Then after a while, I had a full screen of garbage. I had to dial in the horizontal sync, but finally saw this.


Looking at what damage could have been done, I see that pin 27 (WE*) is an output of this tri-state buffer (Pin 14 - UWR*). I imagine shoring this signal to ground could damage that output.

1762096056163.png

I thought the RAM test did not require the use of RAM, but I see this chip drives the CE* of the EPROMs and 4 other chip selects that go out to other pages in the schematic. I'm going to see if there is evidence of other ports of this buffer IC are smoked. If so, I'll order some and move on from there.

Let me know if that screen behavior points to any other obvious problems, or a problem in another part of the schematic. The other thought I had was possibly I cracked a plated through hole when putting in the socket and a signal is not getting to the trace on the top of the board. That might be tricky to investigate without removing the socket because it covers up anything underneath it. If anyone has a pic of the unpopulated PCB around IC130 or is familar with that part of the PCB, let me know.
 
I went over C123 with a logic probe.

Pins 1/19 are ground as expected. Vcc is 5V.


In
In-State
Out
Out-State
Signal
Connectio
2
Low
18
Low
RDC*
17
High
3
High
CS3*
4
Low
16
Low
RD*
15
High
5
High
CS2*
6
High
14
High
UWR*
13
High
7
High
CS1*
8
High
12
High
LWR*
11
Pulsing
9
Pulsing
CS0*
a.) G* (inverted) on IC128 (Data bus buffer to graphics board?
b.) Logical signal in CE* of RAM (IC130 and 3 others)
c.) Logical Signal in OE* of ROM0/1/2


The inputs seem consistent with the outputs, but I would expect more activity. I'll have to dig around a little more and see if there is anything I can see that doesn't make sense.
 
The game was running the memory test fine before so I would triple check the work replacing the ram. If need be, remove the socket. This may not be necessary however, looking at the schematics and using a DMM on continuity should identify broken traces or traces soldered together (where there should not be, e.g. adjacent pins).

You really need to get the game back to running the memory test successfully before proceeding further….

p
 
I removed the socket and it worked, put a new socket back in and it didn't... so I removed it again. Something is definitely jacked up. I made a spreadsheet to go through and exhaustively check each device has address/data/enables connected to the buffer of that signal. Also confirm no 2 signals are shorted.

II was thinking.. The RAM that I put in for IC130 is not the exact same part as the lower 8 bits of data. Do I need matching RAM even if the 2 parts are supposedly interchangeable.

I socketed the IC115, assuming I might need to put in an identical part as IC130 and cracked the axial filter cap (that looks like a resistor). I measured 24 nF. Does that seem right? And can I substitute a ceramic part? Or pull one of the caps from one of the unpopulated sockets? Pulling one might be a b*tch because every one seems to have 1 lead connected to a massive ground plane.
 
The ram part appears to be compatible, there is something going on with the socket, have you tried continuity to the pins without any ram install, just with a new socket?

Also, the capacitor is 25nF, I have used tantalums before. I am sure ceramics will also work.

p
 
Last edited:
Well one step in the right direction. I removed the sockets, put new sockets in, checked all the data/address lines, fixed 2 traces that I must have broken removing socketing the first RAM. I socketed IC 114, 115, 129. and 130. This is with the sockets empty.

If I put anything in IC 114/129, I get the exact same errors. If I put anything in IC 115/130, the game doesn't boot.

If DB=Data Bus, I'll recheck the data lines again. I checked for continuity on everything, but I didn't check for shorts yet.

1762403078867.png
 
I have nothing to add here other than to commend you on your methodical and dedicated troubleshooting! Working through some demons for sure with this one. Keep it up!
 
Took apart the steering assembly, wiped it down and upgraded the gears. Was happy to button it up with no missing parts or extra parts.

View attachment 840353

Nice breezy day to pullout the ozone machine and work on the rodent smell. The cutout for the pedal assembly makes a nice duct for the ozone machine to fill the empty cabinet and vent out the top.

View attachment 840354
Nonchalantly does a teardown and "quick change" of the gears that took me a month from start --> finish lol. Nice work!
 
Thanks @ccie38296! Thinking about how it works (or should work) is a lot less frustrsting than randomly swapping chips. Not to mention my soldering skills are a high risk on this board as I have already demonstrated. 😂 I did just get a hot air station and ordered a PCB preheater, so maybe the right tools will up my game.

Thanks @GameyMcGame! It took me over a month start to finish if you count how many days went by with it sitting in my garage waiting for me to get to it. 😂
 
Well one step in the right direction. I removed the sockets, put new sockets in, checked all the data/address lines, fixed 2 traces that I must have broken removing socketing the first RAM. I socketed IC 114, 115, 129. and 130. This is with the sockets empty.

If I put anything in IC 114/129, I get the exact same errors. If I put anything in IC 115/130, the game doesn't boot.

If DB=Data Bus, I'll recheck the data lines again. I checked for continuity on everything, but I didn't check for shorts yet.

View attachment 859898
I would double check the work for IC 115/130. The old parts allowed the game to boot so this is something that I'd try and sort before proceeding.

p
 
Thanks @ccie38296! Thinking about how it works (or should work) is a lot less frustrsting than randomly swapping chips. Not to mention my soldering skills are a high risk on this board as I have already demonstrated. 😂 I did just get a hot air station and ordered a PCB preheater, so maybe the right tools will up my game.

Thanks @GameyMcGame! It took me over a month start to finish if you count how many days went by with it sitting in my garage waiting for me to get to it. 😂
I had to swap a chip on mine and absolutely discovered that desoldering a multilayer board is trickier than a single layer! I had one pin of the old chip that took me about an hour to get out! Installing the socket was somewhat easier but I was very worried I had mangled something by the time I was done.

Kudos and good luck on your journey! Looking forward to seeing your finished Outrun (I'm looking forward to seeing mine someday too, lol)
 
I removed the suspect socket, checked my work, and put everything back. I thought it might have been a problem where IC130 was not getting 5V. This was one of the traces I had fixed, so I carefully fixed and double checked this. Apparently that's not it because when IC130 is present, the game will not boot.

I double checked for shorts to ground and between signals going in and out of the RAM (which has shared data lines, and many shared address lines with ROM). Nothing.

Looking at the schematic, I was trying to think how IC130 could take the board down when installed. It looks like banks of memory are switched in and out with RAM0/1 (Connected to CE* of RAM) and ROM0/1/2 (connected to the OE* of the ROM). Below is a logic trace of RAM0/1 and ROM0. This isn't right, is it? If RAM0 and ROM0 are both low, both ICs are trying to drive the data bus, right? Or is there some other mechanism that prevents IC30 from driving the data lines when RAM0 is low?


1762707906356.png
 
I installed the RAM at IC130 with RAM0 (Pin20) disconnected and the system booted, so focusing on RAM0.

I'm pretty sure something is shorted to RAM0 or IC141 is bad. If the circuit was working right, there would be no way for RAM0 or RAM1 to both be low at the same time. Also, When CS0* is high, RAM0 should never be low. If I can't find a short visually, I'll pull IC141 and test it.

1762714888727.png
 
I confirmed that RAM0 had a short to one of the address lines. Now IC130 is good.

Still have problems with the other 3 RAM ICs in that area of the board. Looking similar traces of them on the logic analyzer; there is definitely something bad going on with the data lines.

1762798589295.png



At some point, I might just bring it to work to look at it under a microscope or invest in a digital one. I have these magnifiers, but they just aren't cutting it.


Anyone have any recommendations on a reasonably priced but highly effective digital microscope?
 
Back
Top Bottom