• Please review our updated Terms and Rules here

32k PET 2001-N - Video and running issues

D

dhoelzer

Veteran Member
Joined
Oct 20, 2012
Messages
523
Location
New York
Hey-

My life is still in turmoil, but I decided to have a go at my PET before moving it on to a new home (yes, some of you know who I'm looking at).

At this point, it is semi-conscious again... but...

I appear to have a strong clock, phi1, phi2, and there appears to be data appearing on the data lines of the 6502 (see the logic analyzer output). The display, however, is not happy looking. Additionally, while the display in the photo looks awesome (compared to complete darkness), it will randomly flicker to darkness or perhaps even go dark for long periods. Nothing is particularly hot on the board, I just burned a complete set of EPROMs (and verified them), and I've been over the board three times looking for solder joints, etc.

When the screen goes dark, the tube is still glowing, so I'm guessing that it's a video circuitry problem, but I'm stumped as how to tell if it's on the board or on the CRT board. (This is a 2001-N 32k..)

Any suggestions? I know we have some PET pros around here. Photos attached.
IMG_3086.jpgIMG_3087.jpgIMG_3088.jpg
 
Based on that photo of the screen, my first guess would be the video RAM.
It's mostly @ which is screen code zero which tells me the video circuits are reading mostly zeros from video ram.

There is some data on screen though which could just mean some addresses have some other bits stuck on.
Are the characters on screen the same every time you power on?
Does anything change if you type blindly ?

The screen flickering and going dark is likely something in the CRT analog board. I have one that does the same thing but have not gotten around to pulling it and re-flowing the solder joints.
If that doesn't work, I'll need to roll up my sleeves and start testing components.
 
Last edited:
Interesting. I replaced the VRAM about four years ago, so I feel fairly confident that they are good. The content is generally the same, though a few characters might change. The @ had me concerned that the processor wasn't running, which is why I had the logic analyzer hooked up. The good news is the character generator is definitely working.

Regarding the analog CRT board, I think I have the parts to make a composite adapter to attach to the board. I had been thinking about making that and hooking to a composite monitor for testing; asking here first was just me being lazy. :)

If you (or anyone else) have any other thoughts, please let me know. :)
 
In my experience, if the video ram is OK and the processor is not running, the video ram would display a more random selection of characters based on the random bit patterns in the ram at power on.
What's being displayed is much less random. If the video ram is socketed, it's worth a quick swap if you have spares.

If you feel certain that the video ram is good, then of course there could be some problem in the video logic causing it to read incorrect data from the ram, but in that case I'd expect even less randomness, like a single bit stuck on or off and every character shifted by the same value.
The character ROM and the shift register appear to be working fine.

I really need one of those composite adapters myself but have been too lazy to make one. Hoping I could find one on Ebay or one of the retro computing sellers online.
 
Last edited:
:) Ok. Well, more tomorrow. I spent three hours on it today, which got it back to where it is now from completely non-functional. I may start with the composite adapter so that I have a better view of the screen.

In your quote you say, "...and the processor is not running...", which is why I included the logic analyzer output. There is definitely data appearing on DA0-DA7 on the processor, along with good clock coming out of it. It *looks* like it's running.

I may have some spare VRAM, and I think I did socket those chips. If I did, I'll swap in some spares. If not, I think I'll start by wiring the analyzer up to the address bus and see if the CPU is fetching instructions.
 
What I meant to say was that IF the processor was not running, I'd expect more random characters on the screen, not large blocks of @.
That's why I think the vRam may be faulty and the processor is probably fine.

Yeah, I've been fooled by a bad CPU toggling address lines before. No reason to doubt it's working though.
You can try powering on with the CPU removed and see if the screen looks any different. Might confirm it's getting to clearing the screen.
 
Last edited:
Which logic board does this PET have? If it has a CRTC, I expect none of your problems to be with the monitor.

Either way, the ROM may be bad or not being read correctly; of course zero page and stack RAM are suspect, too.
 
KC9UDX said:
Which logic board does this PET have? If it has a CRTC, I expect none of your problems to be with the monitor.

Either way, the ROM may be bad or not being read correctly; of course zero page and stack RAM are suspect, too.
Click to expand...

This is not a CRTC PET. Also, as I said, I just burned fresh EPROMs for it; they are fine. :)
 
Hutch said:
What I meant to say was that IF the processor was not running, I'd expect more random characters on the screen, not large blocks of @.
That's why I think the vRam may be faulty and the processor is probably fine.

Yeah, I've been fooled by a bad CPU toggling address lines before. No reason to doubt it's working though.
You can try powering on with the CPU removed and see if the screen looks any different. Might confirm it's getting to clearing the screen.
Click to expand...


Hah... well, progress of a sort... :)

I do have spare 6502s.. I swapped out the 6502 and after some (very little) random stuff, the screen clears... and never comes back.

I'm going to look through my tubes for some more 2114s just in case I have them. At this point I have the 6502, character ROM, Kernal, edit, basic 1&2, and the... 6522? I forget.. I pulled the PIAs out. Without VRAM I do get a nice checkerboard (when the screen wants to cooperate). I didn't feel like shocking myself, but I'm going to have to pull that video board out.

I took a shot at breadboarding the video mixer with a hodgepodge of what I have on hand.. No joy. Not even a flicker. :) I'm not assuming it's the video circuit; it's more likely the components that I used and the rat's nest of wires.
 
When you say the screen is dark, with the brightness up, are there signs of a a raster scan?

You could monitor the vertical drive, horizontal drive and video data on the main board. Do they go away when the screen goes dark. If so there may be a problem with the main board video timing logic. If the signals stay, but the screen goes dark, then there may be a problem with the video board.
 
Dave_m thinks in exactly the same way that I do!

Before poking around with the monitor, make sure you have sensible H and V DRIVE signals from the main logic board (and something on the video signal as well).

Since you don't have the CRTC controller, you should be able to remove the CPU (so the random power-up pattern should never go away), scope the H, V and video signals and leave the PET run like that for a while. If the picture stays like that, it is more likely not to be the monitor to blame. If the monitor goes blank, scope the drive signals again. If they are still there, the monitor is to blame. If the signals are not correct, the logic board is to blame.

Dave
 
daver2 said:
Dave_m thinks in exactly the same way that I do!

Before poking around with the monitor, make sure you have sensible H and V DRIVE signals from the main logic board (and something on the video signal as well).

Since you don't have the CRTC controller, you should be able to remove the CPU (so the random power-up pattern should never go away), scope the H, V and video signals and leave the PET run like that for a while. If the picture stays like that, it is more likely not to be the monitor to blame. If the monitor goes blank, scope the drive signals again. If they are still there, the monitor is to blame. If the signals are not correct, the logic board is to blame.

Dave
Click to expand...

Yep. I always have h-drive and v-drive on the board. It sort of acts like the connector/wire is loose somewhere. Often it will go blank, sometimes it will drop down to a single narrow band.
 
Other Progress

Other Progress

After swapping out the 6502 and reseating everybody, I get a mostly clean screen with a free RAM report.. Even a flashing cursor.

It will not start reliably, however, which makes it seem as though the CPU isn't always starting. The last five times, for instance, it was just the same RAM report leftover in the chips with no screen clear. There are still some brokenish characters ($ and @ in a few places), so there might still be SRAM problems. The keyboard also seems to be non-responsive, which is a new symptom (it did work some time ago). Maybe a bad PIA? Not sure.

At this point, the monitor dropping it is the biggest hinderance to progress. I'm going to both wire up a reset switch so that I don't have to keep flipping the power and take out the board for the 9" monitor and give it a good going over. Before I picked this thing up back about 8 years ago, mice had made quite a nest for themselves. The greatest concentration of them was in the 9" CRT housing. It looks like they sharpened their teeth on the heat sink for *years*. In many ways, I'm surprised it works at all.
 
Monitor wise then we will have to follow the H and V drive through the chain. The fact that you say you get a narrow band sometimes is also a good indicator of vertical circuitry problems.

For the main logic board - you could try my diagnostic ROM that fits in the 2K EDIT ROM and see if that can diagnose a consistent issue.

Dave
 
KC9UDX said:
Horizontal or vertical band?

I'd put the old 6502 back in.
Click to expand...

Horizontal. It's not constant. When the image is up, it's up and rock solid; bright, clear. Then it will go out, sometimes completely, sometimes like something is loose, sometimes to just a horizontal band that seems to lack sync. More often, completely nothing for long stretches. It does seem happiest when it's cool.
 
daver2 said:
Monitor wise then we will have to follow the H and V drive through the chain. The fact that you say you get a narrow band sometimes is also a good indicator of vertical circuitry problems.

For the main logic board - you could try my diagnostic ROM that fits in the 2K EDIT ROM and see if that can diagnose a consistent issue.

Dave
Click to expand...

Hey Dave-

I thought I had posted asking for a link to the 2k Edit Rom diagnostic. I can't seem to locate it with Google. :)

Thanks
 
You must log in or register to reply here.
Back
Top