Exidy Crossbow won't boot

akeym

Member

Donor 2012
Joined
Feb 14, 2012
Messages
56
Reaction score
0
Location
Oregon
Exidy Crossbow won't boot - FIXED

I'm getting what appears to be a "common" problem with my newly acquired Exidy 440 system. It will start up, but it will only display yellow/blue bars on the screen. It will make a painful attempt at playing some sort of sound, but it generally only spits out 60hz for a couple seconds (along with noise) and then stops. Starting in test mode doesn't make any difference.

As per some other threads here I checked for voltage drop issues at various points on the board. The incoming 5v line has an initial drop of about .2v on the main power connector (the connector has clearly gotten hot) but after that, it is the same at all points on the board. The +/-18v supply works, but has a decent bit of ripple and needs its caps replaced. The 5v supply has a lot of high-frequency noise on it.. probably should replace the big blue cap on that part of the supply as well. The 5v rail is currently at 5.18v.

I'm not sure what actually causes the blue/yellow bars - is this just how the palette ram is initialized or the proms? Or the line buffer? Does this indicate that the main CPU has not started?

I figured my best bet was to check on the clock generation circuitry.. Being a bit of a noob, I'm not sure the best way to test this. When I hook my o'scope up to the output from the 7404, I see a lot of echo in the signal and it's quite noisy. I tried to calculate the frequency it was oscillating at, and I get 4.7mhz (?!?!) which doesn't seem right at all - however, I could be doing something very wrong here.



TL;DR: What do the blue/yellow bars mean and how to I check the crystal oscillator properly?

Attached is a schematic of the oscillator circuit so you don't need to hunt it down in the full schematic.

Thanks for any help!
 

Attachments

  • crossbow-clock.jpg
    crossbow-clock.jpg
    11.6 KB · Views: 44
Last edited:
Well I tried to run the board with a known working 5v supply, in case the noise was causing the issue, with no luck. I also tried replacing the 74ls04 with a new one with no change in operability, and the clock still looks terrible. Would replacing the crystal with a 12.96mhz have a significant difference on the machine's working ability?

Does anybody know someone with a good knowledge of these systems I can contact for advice? ..What about the makers of the Collex'n?
 
Well, I replaced the no-longer-manufactured 12.9792mhz crystal with a programmable oscillator from digikey (I'll need to make a small daughter board to keep it in place, but for now it's just jumpered on, and the output from the 7404 was cut and the oscillator output put in its place.)

Clock looks clean, but the system still won't boot. Re-checked voltages at various chips (cpu, ram, rom) and all were at least 4.97vdc. On a whim, I decided to just pull the logic board's 6809 out and ended up getting the same thing on screen: blue and yellow bars.. This seems to confirm my suspicion that the pixel ram and proms are just initialized like this at power-up - so if the CPU isn't running, or the boot roms haven't initialized the ram, this is what you see.

This means the clock is good, since the single clock is also responsible for generating sync to the monitor, and the video output section is working fine. It's reading pixel ram, and putting it into the buffer, then generating a valid video signal (at least, that's my understanding of how it works).

I'll be looking over the schematics to see where I should start probing with the scope... Also tried resetting the CPU manually in case power-on-reset wasn't working, still nothing.

As always, any help is appreciated,

Back to the trenches!
 
So, for some reason the 6809 from the sound board was swapped with the 68B09E from the logic board. These aren't the same CPU, so clearly it wasn't going to work. However, after swapping them back, I've got a running logic board CPU, but still no change in outward symptoms. BS, BA indicate that the CPU is running, however I see no action on the R/W line. There's chatter on the address bus, E and Q are coming out, nothing is actually...happening. Program RAM is never written to either. I'm starting to think this 68B09E is borked.



Anybody have a PROM dump of the 74s288 (82s123) at position 3K on the logic board? The one MAME uses is 256*4, but crossbow (and cheyenne) use 32*8.. I don't think it's an issue at this point, but I figure I'd ask just to check.
 
I don't know if this helps, but mine does the same thing when first turned on. Then I turn it off, and back on again, it always boots up properly the second time - like it needed to warm up or something.
 
Verdict: Dead CPU. Replaced 68B09E with a known working one and it boots up. Still no audio, decided to just replace it's 6809 and it perks right up. Not sure what caused them to fail in the first place, possibly someone cranking the power supply up too high. I don't think the fact that the CPUs were swapped would cause them to "die," but I could be wrong. The Logic board CPU does run at a higher frequency than the audio board's CPU is rated for, so maybe it just burnt up?

I should clarify my first post regarding the cystal: When I first saw the machine, it gave me the blue and yellow bars. When I got it home, it only put out a blue and yellow non-sync'd mess on screen. I believe I somehow managed to damage the crystal in the moving process. If you're getting blue and yellow bars, it's not because your crystal is operating at the wrong frequency or broken.

It should be noted that the audio board works semi-independently of the logic board. It will play it's horns as a test even if the logic board CPU isn't running/working. However, it relies on a clock generated by the logic board - you cannot bench test the audio board in a running state without the clock signal from the logic board.

I now have a working Crossbow! Now to re-cap that monitor...
 
This remains over my head, but knowing the prosessors were bad may help me fix my spare boardset, as I was told similar symptoms are what this board exhibits....thanks for posting this..
 
Verdict: Dead CPU. Replaced 68B09E with a known working one and it boots up. Still no audio, decided to just replace it's 6809 and it perks right up. Not sure what caused them to fail in the first place, possibly someone cranking the power supply up too high. I don't think the fact that the CPUs were swapped would cause them to "die," but I could be wrong. The Logic board CPU does run at a higher frequency than the audio board's CPU is rated for, so maybe it just burnt up?

I should clarify my first post regarding the cystal: When I first saw the machine, it gave me the blue and yellow bars. When I got it home, it only put out a blue and yellow non-sync'd mess on screen. I believe I somehow managed to damage the crystal in the moving process. If you're getting blue and yellow bars, it's not because your crystal is operating at the wrong frequency or broken.

It should be noted that the audio board works semi-independently of the logic board. It will play it's horns as a test even if the logic board CPU isn't running/working. However, it relies on a clock generated by the logic board - you cannot bench test the audio board in a running state without the clock signal from the logic board.

I now have a working Crossbow! Now to re-cap that monitor...

Marking this for myself. Thank you for logging this.
 
Back
Top Bottom