• Please review our updated Terms and Rules here

KayPro 10 Intermittent Issues with Booting

Do I understand correctly: re-seating the CPU did actually "fix" something? You no longer see the problem where RESET does not clear the display? By "back to the old symptoms" you mean that lockup? It sounds like there may be more ICs that need to be re-seated. Perhaps try the DRAM chips first. Of course, a massive re-seating effort requires care, to avoid damaging any ICs.
 
I'm stumped. I've reseated everything.

Same symptoms persist. Reset. Does not clear the display when the system locks up.
 
Well, from the symptoms it sounds like something happens to prevent the CPU from writing to the video RAM. From what you've said, it sounds like it might be a thermal-triggered issue. What we can't tell is whether the CPU effectively stops work altogether, or whether it is something along the path that the CPU uses to write to video RAM. That path is via an I/O port and is also predicated on reading (another I/O port) the video status to wait for retrace (flicker avoidance). So, if the CPU can still execute ROM during this problem, it may be the CRT controller chip or the logic which converts the I/O write into a save to video RAM.
 
I think you have it here. The CPU can execute ram up to the point where it starts talking to the controller then it locks up.

Sometimes if I'm lucky after a while on the garbled display, the CPU will slowly have it clear (line by line) and it'll boot or lock up halfway displaying the ROM version info. Short while later it'll lock up again regardless.

This is the WAIT pin on the CPU in yellow and the Chip select on the CRT controller
 

Attachments

  • SDS00003.jpg
    SDS00003.jpg
    83.7 KB · Views: 2
Interesting... if sometimes it "hobbles along" and eventually clears the screen, then it might point to something going wrong when reading the "status" - which I think the the horizontal retrace/blanking status. I (or you) would have to check the 6545 datasheet to see what status is being checked, and whether that status bit is derived from external signals or from internal CRT timing. I suppose, if you have an extra 6545 lying around you could swap it out. I should have some more time tomorrow to think about this, and I can look at the ROM code and see what is actually being read as status.
 
The 6545 also runs HOT, almost too hot to touch. I don't know if this is by design or not. I ordered what I hope is a spare but is there any way I can jump it out so the CPU isn't waiting on it (despite not having a monitor)
 
QuillOmega0 said:
The 6545 also runs HOT, almost too hot to touch. I don't know if this is by design or not. I ordered what I hope is a spare but is there any way I can jump it out so the CPU isn't waiting on it (despite not having a monitor)
Click to expand...
Probably not. I'll have to dig out my ROM listings and the datasheet to check, but I believe the "status check" is reading one of the 6545 registers, in which case I can't think of any easy way force the bit.
 
I ordered a 6545 which I hope is going to be good cause those are hard to come by.
If this doesn't fix it I don't know what to do beyond pulling the board apart and doing a rebuild from scratch
 
Confirmed this was the fix.

Now the WD1002 card is causing a fit. Cold it seems to be okay but the second it gets warm to any degree it'll refuse to write to the virtual MFM drive when it had no issue before
 
You must log in or register to reply here.
Back
Top