TRS-80 Model 1 Video Problem

[rickjhorne]

Hi Guys, I am trying to resurrect my old Model 1.
when I turned it on the display shows graphic blocks at each location on the screen.
it does look like "MEM SIZE" is there hidden.

any Ideas for where I need to look? is this a video memory issue?

Rick



(Sorry the image is upside down)

 

[RedskullDC]

Hi Rick,

Looks like the left three

Is your Model 1 a USA or JAP model?

If it's a USA, check the LS174 at Z28, or the LS166 at Z10.
I'd test those before going and changing any RAMs.

Hopefully, it is not your character generator chip at Z29.

Regards,
Leslie

 

[ldkraemer]

If you look at the possible Graphics characters that the Model 1 can display, you can narrow down the
Graphics Character as one of about three. 149, 161, 165, 170. Of these I think it is 149 Decimal
which is 95 HEX, or 1001 0101 Binary.

"M" is 77 = 4D hex or 109 = 6D hex.

Since the Model 1 did not come out with Lowercase looks like M is 4D Hex = 0100 1101

0100 1101
1001 0101

^^x^ ^xxx These bits are different

I'd start with D7 of the Video RAM to see if it is HIGH.

Should be easy to see if it's always HIGH.


Larry

 

[Petrofsky]

You have a problem with character generation, with all pixels in the left two columns of a character erroneously always on (and the problem is restricted to the character portion of each character cell, i.e., the top 8 lines of each 12-line cell). The video RAM appears to be fine. If you poke some screen locations with graphics values (128-191), I expect they will all show up correctly, confirming there is no problem with the video RAM or the video data hex flip-flop in Z28.

The two left columns of a text character are generated on pins 16 and 15 of Z29 (the character generator) and connected to input pins 12 and 11 on the character shift register Z10 (a 74LS166). So the problem is likely: (1) the character generator is bad and is always outputting high on those pins; (2) the shift register is bad and is acting as if the input on those pins is always high; or (3) the chargen and shift register are both good, but the traces connecting them are disconnected or are shorted to a high voltage.

Btw, I tried looking at the circuit trace to see where a problem in the trace would most likely occur, but I failed. Most TRS-80 service manuals and technical reference manuals include images of the circuit traces, but the Model I manuals are a notable exception. I tried looking at the Sams Computerfacts, which has photos of the component-side trace, but the traces connecting Z29 and Z10 are on the back side. I googled around for photos of Model I boards, and found some good ones of the component side (like this one), but none of the back side. The internet has failed me. Someone please upload a good high-res photo sometime (or direct me to the one that I failed to find).

 

[Petrofsky]

I found a good picture of the Model I logic board backside, attached to a 2012 post at the Nightfall Crew site. Here's the area of interest:

Z29 is in the lower right, Z10 in the upper left. You can see the traces from Z29 pins 16 and 15 (the third and fourth pins from the top on the left side of Z29) to Z10 pins 12 and 11 (the fourth and third pins from the bottom on the left side of Z10).

I would start on the component side of the board, with the system off, and use an ohmmeter to check for connectivity from Z29.16 to Z10.12 and from Z29.15 to Z10.11, then check that those lines are not connected together and that neither of them are connected to Z10.9 (which is connected to R40, a pull-up resistor).

 

[rickjhorne]

Thanks Allen, this sounds encouraging.

I am currently recovering from ear surgery so I am not fully capable yet.

Earlier I did pull the character generator chip and I still saw the blocks.
I will check the traces later.
Thanks
Rick

 

[rickjhorne]

Hi Alan, Thanks you were spot on! I found a cold solder joint.

Now that I can read the screen I see I have a memory problem and a keyboard problem.

One thing at a time....

Thanks Everyone

Rick