I’ve had two weeks of leave, and we know what that means, right? A veritable frenzy of repairs and projects!
I’ve been beta testing a new SuperPAK Romset for Alan Laughton (Aka ChickenMan) as he has been developing an extensive set of ROMs for Suzy’s amazing SuperPAK coreboard. With literally hundreds of slots to store ROM based games in, Alan has been working up a frenzy of titles to fill those spots. I helped where I could by burning out EEPROMS and testing on my “vanilla” Standard Microbee base board.
We managed to find a couple of issues and Alan was always very prompt to fix things.
Awesome work, Alan!
I also finally got around to 3D printing Cherry MX keyswitch to Microbee keycap adapters.
While they took a bit of work to clean up, I was able to print these 15 at a time, and with some tweaking, 28 at a time. So time on the printer went down, but time with cleanup was still slow.
With some patience and time, I was able to finally get all the keys replaced on my Microbee Premium. How long did that take? Too long!
I still need to work out something with the stabiliser rod for the space bar. but for now, so much better!
Not my computer, but sent in for repairs by Eric.
Initial symptoms were that it had a black screen.
As the Atmos and the Oric 1 share all the same chips, I was quickly able to eliminate quite a few components. I knew it wasn’t the CPU (Which was socketed, but more on that later) and it wasn’t the ULA or ROM. Swapping these components showed no changes. More importantly, all worked fine in my Atmos.
Putting it on the bench, I went through the usual checks and spotted that there was no clock at the CPU. Things don’t work well without a clock. Tracing through the circuit, I realised there was no clock going into the ULA either. On the schematic, the clock was being generated by only a small handful of components. Notably, a 74LS04 inverter and a 12MHz crystal. I had both in stock, or on hand locally, so I replaced the 74LS04 initially. That gave me a clock, but it was running way too fast at somewhere approaching 30MHz, and was uneven. I replaced the crystal and was getting a nice solid 12MHz to the ULA and the divided clock signal from there.
Now I was getting stripes on screen. Better, as it meant the ULA was generating video. While this was going on, I also did some “quality of life” mods in the area to improve clock. This consisted of adding two resistors, cutting a track and adding a patchwire.
At this point, it looked like a RAM issue, and asking around suggested that was the next thing to investigate. I desoldered it all and tested it, one by one, using the Retro Chip Tester Pro at Artifactory. The RAM all passed. Hrm. OK.
I dropped the RAM back in. (There’s no room for sockets so I had to solder it back in). I then started trying to work through things on a more granular manner.
I was seeing activity on the RAM lines on the CPU, but I started to notice things that were “wrong”. There was a capacitor in backwards. All the electrolytics were axial in a board designed for radial. There had been some work done around the CPU socket.
I started checking all the rework, replacing the electrolytics with radials and triple-checking they were in the right way.
I then started concentrating on the CPU. The socket rework worried me. I cleaned it all up and toned out all the pins. While doing so I found a shorted data line and a broken address line. I fixed these all up, but still the same screen.
I tried both diagnostics ROMs at this point, neither got any further.
At this point, I got sidetracked. I thought the issue might be the single wipe socket the ULA was using, so I replaced that. It wasn’t that, but having a dual wipe socket in there does make me feel happier.
Finally, having checked the schematics, and consulted with the Oric Facebook group again, I reached the point where I was going to have to desolder and test the 74LS257 ICs.
Guess what? One was faulty!
Unfortunately I don’t have any on hand, so I have placed an order to get some. Stay tuned on this one.
In other Oric news, I also have a set of keycaps coming for the Atmos. Huge thanks to Steve Marshall for supplying these. Once they touch down, I’ll definitely be posting a glamour shot 🙂
I have also ordered a Cumana Reborn from Zaxon (Who I have already purchased stuff from, in the form of th DDI5 for my Amstrad). Can’t wait to play Oric disk games 😀
Let me just say, the Oric community may be small, but they make up for it in heart.
TRS-80 Model 1
Ah the venerable “trash 80” as it is often called. This was received after I started looking around for an Expansion Interface so I could try and adapt it to work with my System 80. (I’ve actually got a PCB ready to go when I get some time that’ll do just this. Adapt a System 80 to work with the Expansion Interface)
After asking around, I was very kindly offered one by Steven, who also sent me a TRS-80 Model 1 to go with it. Awesome! Steven did warn me they had been in storage for a while so I was curious to see what state it was all in.
Once they arrived, it was obvious that they’d need cleaning. They were still much cleaner than some of the systems I’ve worked on. 🙂
I cleaned the surfaces of both the Expansion Interface and the Model 1 (Henceforth EI and M1 to save a lot of additional typing) and cleaned the keycaps themselves in water with a denture cleaning tablet. They cleaned up nicely.
I only had one power supply on hand so I chose to start with the M1. An inspection of the inside showed the keyboard ribbon was beginning to delaminate, a known problem with these, but it otherwise seemed OK. I was also able to clean the inside of the case and the motherboard. All the caps seemed OK, but I might replace them later “to be sure”. Thankfully the motherboard itself hangs “upside down” so the component side of the board was quite clean.
I checked voltages from the supplied power supply and they seemed all within tolerance, so, what the heck. Probably time to proceed.
As I had the video cable from the System 80 already, all I needed to do was plug it in and…
Whattayaknow? It works!
(Before anyone asks, I also tested all three voltages, +5v, -5v and +12v and they all passed OK)
I was able to load some simple games successfully at this point too. Time to work on the Expansion Interface. 16k is OK and all, but a bit limiting in the long run.
At lis point I looked at various solutions for a second power supply. I did power up the EI and check voltages were OK, but only being able to power the EI or the M1 limited my ability to really do much with the system. After investigating what was out there, and what I had on hand, I decided I was going to just bypass the onboard power regulation, as I had a Meanwell RT-65A on hand. A modern, switchmode PSU capable of supplying +5v, -5v and +12v. Everything the EI needed.
After some work with cardboard, and then a sheet of acrylic, I was able to drop in the PSU. I then added an On/Off switch. I bonded the various ground and voltages around the board (Making sure the +5 and GND were distributed widely)
Now I was able to power up both.
Things weren’t perfect though. The M1 was randomly crashing and the keyboard was getting erratic.
Time to do more diagnostics. All the RAM in the EI was socketed so I tested it all in the Retro Chip Tester Pro, and it all passed. 4116 RAM tests in no time at all. It’s great that way.
It was at this point I noticed how corroded the edge connector was on the M1 that connected to the cable that flew across to the EI.
I cleaned up the contacts with an ink eraser, being careful not to strip too much of the surface. (I also added a tiny amount of deoxit to the edge connector.
Long story short, that fixed all the non keyboard instabilities.
Back to the keyboard. At this point, the ribbon had completely delaminated, and I was getting flashbacks to the System 80 all over again.
Right! Let’s head this one off at the pass, shall we?
Out came the ribbon with the Moo gun (And some shouting and more heat than I was comfortable with. I did lift a track but was able to salvage it. Old crusty connectors are the worst). In went right angle pin headers. I then crimped up a 20 way cable out of silicon wire and dupont crimps.
This fixed the keyboard issues. I felt I was starting to get somewhere.
Next was a cable to go from the floppy disk edge connector on the EI to a modern floppy disk drive connector, so I could use, you guessed it, another Gotek drive. I love those things.
After some false starts with disk images that weren’t compatible, as well as the discovery that the disks needed to be in HFE format, I was finally able to boot the TRS-80 from a disk image.
(BTW, there are tools that will convert most disk formats to HFE over on the HXC Floppy drive website)
I’ve been teaching myself some TRS 80 DOS since, but I haven’t yet got everything working the way I want. I plan to fire up an emulator and play around with things.
In the meantime, I built a dual Gotek rig so I could use two drives with the TRS-80.
The final piece of work on the TRS-80 (for now) was I swapped out the 5 pin DIN on the power lead as the original had become brittle and the plastic was breaking apart. I used a nice metal shelled one to replace it.
After testing the cache in this in (You guessed it) the Retro Chip Tester Pro, I have reverted this back to being a 486DX 40. It should be more than enough to play the types of games of the era I wanted to play. Maybe I’ll end up building a K6-500 system (Or equivalent) down the track, but for now this gives me excellent DOS performance for many games of the era.
Just some odds and ends.
I got a TRS-80 Colour Computer 3. It’s a real games machine. It came with a 512K RAM expansion, and I have added a 6309 to replace the 6809. Now I can run NitrOS-9.
A friend printed me a chip leg straightener on his FDM printer. (Hi Diego!)
This thing is awesome. I have used it a lot in the week since I got it.
I finally got around to mounting up my parts bins.
Now I can find things! No longer on the floor.
I also built a tiny shelf to go behind my wife’s laptop so she can store more things around her desk. She is very happy with it.
Finally a candid shot of “Work in progress”. Yes my desk gets stupidly messy 🙂