TRS-80 Model 1 repair help - garbage screen & inconsistent CPU behavior

[Blast_Brothers]

I acquired a broken Model 1 (Rev. G, lowercase, Level II, piggybacked 16K RAM chips with a daughterboard, so 32K?) years ago and am only just now getting around to fixing it. After replacing Z5, Z6, and Z57 I'm getting video out, but it just shows a screen full of garbage and I've been unable to narrow the problem down further. This is my first "big boy" repair so I'm hopeful that I'm just missing something obvious.

The pattern of garbage is almost, but not entirely, consistent between power cycles. In addition, sometimes there's a flickery black bar rolling down the screen, and sometimes there isn't. This is correlated with differences in CPU activity; when the bar is present, the address lines show activity, and when it's not, they're just stuck high or low. The CPU is getting a clock signal, and all of the input pins (INT, NMI, wait, BRQ, reset) are high as expected regardless of this (outside of some more noise on pin 24 when the bar is present) so I'm not sure where the inconsistency is coming from. I assume the black bar is the CPU interrupting the video display, but it's unable to initialize the machine for whatever reason?

The keyboard cable had disintegrated, so I replaced it with a detachable IDE cable. I'm measuring that some adjacent pins on the connector (12/13 on the keyboard side, and 15/16 and 18/19 on the mainboard side) are shorted together. Looking at the schematic I don't think that's supposed to be the case, so I'm assuming that some solder bridges formed under the IDE pin headers when I soldered them on? But that feels like something to look into after I get the rest of the unit working.

I've tried swapping the CPU, removing the CPU, unplugging the keyboard, and taking out the ROM and RAM shunts (and most combinations of these), and nothing changes the pattern of garbage, other than the black line never appearing when the CPU is out, which makes sense. I don't have any other spare parts on hand so I figured I'd ask for advice before ordering replacement RAM or ROM chips.

 

[ldkraemer]

If you remove the Cable going to the ROMS, and Power UP what do you see Displayed?
You should see a screen filled with:
@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9@9

If you do not see that Post a Picture of what you are seeing.


Go to:

Wayback Machine

web.archive.org

Then go to
http://ts-inc.dyndns.org/Diagnose Repair TRS80 TRS-80 Model 1.html

This site used to have good troubleshooting information for the Model 1.

Larry

 

[Blast_Brothers]

If I remove the ROM cable, the image is the same as what I attached to the first post. No kind of repeating pattern, or a change at all.

 

[ldkraemer]

The next step I'd try is build yourself a NOP tester from a SPARE Z80 CPU. Just bend all the D{0..7}
Pins out far enough they will not go into the Z80 IC Socket. Jumper each one to the GND Pin of the
Z80 IC. Remove your OEM Z80 and insert the NOP tester. You can power up and check A0..A15
at the CPU and chase A0 across the Motherboard using an O'scope. You will have to set the
Direction you are testing (READ/WRITE) accordingly.

You can also check the Control Pins (INPUT/OUTPUT) of the CPU for the correct state.

BUT, you ALWAYS Start by Checking the +5 VDC and +12 VDC Power Supply along with the -5 VDC
Power Supply.


Larry

 

[Blast_Brothers]

The power lines are fine. I'll have to get back to you on the NOP generator.

 

[ldkraemer]

Fine is not a percentage of tolerance specified in the TRS-80 Model 1's Technical Manual.

How much AC Ripple is riding on the +5 VDC Line?

Is Q2 warm or VERY HOT when you touch it?

If you lay your Index finger across the top of the RAM IC's (one at a time) does any of them
burn your finger? They are very prone to short internally to the -5 VDC Supply.

Attached are the actual pages with the exact allowed percentage, and order of adjustment.

Larry

 

[Blast_Brothers]

I forgot to mention that I already measured the voltages and adjusted the pots to match the manual. The 12V line is 11.99 V, +5V is 4.996 V with some ripple (Vp-p is 812 mV), and -5V is -5.039 V.

The RAM chips are all equally warm. None of them are significantly hotter than other chips on the board, like the CPU.

I'll try the NOP tester next.