Fixing a TRS-80 Model 100 LCD Issue
I recently pulled out some of my old computers to dust them off and check them out. I have a sweet Tandy TRS-80 Model 100 that always worked fine, but I never really used it a whole lot. So after typing some BASIC code into it and running it for a few minutes, I noticed the LCD was flickering off and on. Eventually it stopped showing up altogether.
I figured this was probably the contrast control potentiometer needing a cleaning, but it was smooth as a whistle. I did a bit of searching to help me narrow down the issues that could cause this… and didn’t turn up much. So hopefully this post will help others that might have the same problem I did.
TL,DR; I had some dried up electrolytic capacitors (C49, C54) that needed replacing. That’s it, a simple -5V (VEE) power supply issue.
I found the Model 100 Service Manual here at the awexome club100.org website. Service Manuals back in the day were great!!! They have all of the schematics, PCB layout images, BOM, theory of operation, and troubleshooting sections. Everything you wanted to know and more.
After studying the schematics for a bit, I figured out that there are two main power rails that control the LCD, VDD (+5V) and VEE (-5V). The LCD bias voltages are generated with VEE. I quickly measured VEE with respect to GND and found it was -0.75V when the LCD was off, and -3V or more when the LCD was on. It should normally regulate to -5V. It was fluctuating up and down quite a bit, with no rhyme or reason. Here you can see it was measuring -1.429V.
The troubleshooting section for the power supplies, said that if VEE (page 5-4) was not -5V, check D15, C85 and D14. In practice, you should check D15, C85, R97, D14, C86, C47, C49, C54, and C56.
I had checked everything by the transformer section and it was fine, but something appeared to be loading the circuit down stream from there. I was having a bit of a time trying to find what other components were attached to VEE since the schematic is not searchable for “VEE”. It’s just a scan! After visually scanning the schematic, I noticed a bank of capacitors to suspect, but where were they on the PCB?
I fired up Photoshop and exported some Black and White PNG images of the Model 100’s PCB layout from the PDF version of the Service Manual. After a bunch of manipulation, I had created a RED top layer and BLUE bottom layer fairly well lined up. From there I could trace the VEE connections around the board easily. Here’s an overall view of where VEE goes (click for a really large image):
Here’s a close up of where the problem was, the yellow highlight all of the way on the left goes to C49 and C54. After desoldering them, it was clear their electrolytes had leaked out. One measured 3uF and the other wouldn’t even give me a reading. They are supposed to be 10uF.
If you’d like to play around with the Photoshop file to troubleshoot your own issues, please have a look at the file here.
Here they are after replacing with 10uF 50V caps.
Here’s a nice picture of the inside of the Model 100 (click to make it huge!!)
After putting everything back together and cleaning it up a bit, here it is running BASIC again…. albeit super duper slowly! Can you believe this thing cost $1400 back in the day? This is supposedly one of the last products Bill Gates programmed a fair amount of code for. It’s still a lot of fun to bang in some code and type RUN 🙂