Holiday Retro Works

So I had some time off over the Christmas / New Year period, and planned ahead to have enough parts to work on a project or two.

CGA Output from a C128D. ©2021

CGA Converter for Commodore C128

First up, I’d read about a CGA to VGA adapter, designed to work with Commodore 128 systems. It was a relatively simple design, with a handful of passive components, and two simple 74 series logic chips. It was designed to work with the Gonbes 8200 series scalers, but I figured I could try it with my 15KHz monitor.

I decided I didn’t want to use my usual veroboard, as it tends to need a lot more planning to work well. Instead I went with experimenter’s board, which I hadn’t worked with before. Overall the experience was a positive one and I will be buying more for future projects.

I laid out the components in a pattern similar to the original schematic, tack soldered them in and started bridging between connections. I used bot the top and bottom side, using fly wires from legs for longer distances.
I swapped out the entire 5v generation section and instead tapped directly in for 5V with a USB input.
I used turn pin sockets to make development and debugging less work
I used pin headers for external cables, just to make things easier.

Experimenters Board. If you stick the chips in the sockets the right way round, they work better. ©2021

Can I mention, right now, how much I love my cheap Chinese DuPont crimper? For 2.54mm Terminal Header connectors, it’s just the fastest thing!.

Anyway, gushing aside, I methodically worked through all the pins on all the sockets, added all the passives, and using my multimeter in continuity mode, checked for shorted legs. This was particularly important in this case as the experimenter’s board is based on shorting between pins, so there’s lots of solder floating around.
Thankfully there was no issues, so I finished off the two video leads plus the power lead and carefully plugged in the two chips.

I spend a lot of time soldering up DE9 connectors. This one plugs into the C128. ©2021

I dragged out my C128D on the bench and set it up. Plugging in the leads to my 15KHz monitor and the C128D I turned on the system and… Nothing. I immediately unplugged the power, and checking the chips, one had got very hot. Never a good sign.

Quickly I realised my mistake. Despite telling myself to check it, I’d mounted one of the logic chips in backwards. I let it cool, tested it in my chip tester, swapped it around and plugged it all back in again.

With my hand over the power lead, I tentatively turned it on again and… I got a picture! It was working!

Life! Give my creation LIFE!! ©2021

It looked a little off, but I was expecting that. There were several level adjust pots and I’d adjusted none of them. Time to get some colour test software. After MUCH trying I finally found some, which revealed my colours were quite off. In fact, I seemed to be missing an entire colour channel. In particular, the green channel was
Pulling the board out I checked everything carefully but could see no errors.
Time to break out the Multimeter again.
Tracing the green line I was finally able to identify that there was a disconnection between two points. Visually, it looked fine, but while probing, the wire suddenly sprung loose. A touch of solder and some slightly singed fingertips, and we were ready to go again.

Now it worked fine. I was able to tweak the colours and intensity and get a nice signal out on my modern flat panel display. This adds a lot of versatility to my C128. I look forward to using it in future for demos and applications.

Pretty! Even prettier in the flesh too. ©2021

CreatiVision Fixes

The CreatiVision has been a bit grumpy as of late. It will boot fine, but after a few minutes, the picture will distort and the it makes an angry buzzing noise. It’s also back to using RF, a situation I am not happy with.

CreatiVision Guts. The RF Modulator us that shiny box on the right. ©2021

I decided it was time to fix the problems, starting with the RF. At the suggestion of Clockmeister, I used a teeny single transistor circuit to make a small signal amp. I had all the components lying around the house, and this time I did use Veroboard as I had a scrap left over from another project. I decided to use my new desk setup to also remove the RF Modulator as it was only getting in the way.
Out comes the desoldering gun and off comes the RF Modulator.

Surplus to requirements. ©2021

This also gave me lots of room to mount the PCB. Learning from past mistakes, I mounted it using “hook and loop” tape so that I could easily remove the board if needed.

Dark in there. Single transistor amp. ©2021

I got an old RCA triplet, stripped off the third cable and tacked the whole thing together. I tapped the video out from the spot the video went into the RF Modulator. Ditto for Audio, although I did tap a more universal ground point as I didn’t need the isolated ground that the RF Modulator used. I was also able to locate a suitable 5V power tap point nearby and tapped into it.

I ran the two RCA leads out through the cable management in the CreatiVision case and out the back. Before I went to button it all back together, I have it a test and yeah! First go. Picture AND sound.

Of course after a minute or two the noise and visual distortion were back. Now we had a good picture, it was time to do deeper fault finding.
The good folk on the Madrigal CreatiVision forums had suggested I start with voltages so I checked the levels at the power supply which were fine.

(A neat trick for checking voltages on a device with pins like this is to cut some shrinkwrap or pieces of straw to the length of the pins, drop them over the pins and you can then stick your multimeter probes in without having to worry about shorting nearby pins.)

These are the 5v and 12v regulators. These turned out not to be the problem. ©2021

Next I tested the voltages coming out of the regulators by checking the 3 voltage pins of the onboard RAM. The Creativision uses the 4116 RAM chips, so by tapping the four corners, it’s easy to read the voltages. I used clip leads rather than probes, to remove any chance of shorting, say, 12v or -5v into somewhere it really doesn’t want to be.
All three voltages were fine. Alrighty then… Recap maybe? We were all so sure it would be voltages.

At this point serendipity struck. While preparing to video the noise / distortion, I bumped the case by accident and the noise / distortion simply went away. It came back, I bumped the case and gone.

Anyone who has done any regular fault finding knows exactly what the problem is at this point. You’re all shouting at the screen “It’s a DRY JOINT!”.

I broke out my magnifying visor and a permanent marker and slowly and methodically scoured the board for dry joints. Whenever I found something suspicious, I’d mark the board with an arrow pointing to the potential dry joint, all the time desperately hoping the fault wasn’t under the soldered on RF shield covering the underside of the RAM.
There were probably a dozen spots that looked dubious, including an obvious candidate in the form of a bodge wire that sprung loose when I touched it.
I applied flux and some fresh solder and reflowed every single dodgy joint on the board. Once done, I tested the board and it ran fine for a significant period of time. I rinsed the board off with Isopropyl Alcohol (Which conveniently removed any permanent marker residue) and put it back in the case.

It’s been running fine since.

There’s still one final issue, and unfortunately there’s not an easy fix in this case. The membrane keyboard isn’t registering a single character, the space key, which means I can’t enter any BASIC programs for now. The fault appears to be in the actual membrane itself, which is folded and riveted into the controller. It’s not coming out without destructive removal, which I’m not prepared to do quite yet.

To that end, I’m trying to map out the schematic for the keyboard addon, with an aim to make a mechanical keyboard replacement, initially with simple tactile buttons and then once the PCB has been proven, I’ll do a mechanical keyboard replacement. It’s a slow process.

The Lyte

And now for something equally vintage but completely different.

Many years ago I had purchased from a housemate a strange audio-visual addon called, cryptically “The Lyte”. It took an RCA audio input and turned it into a visual representation by splitting the audio into high and low frequencies and then showing them on two VU meters that were spinning. You ended up with a very visual and impressive light show that pulsed and twitched in time with your music input. Certain tracks, in particular would end up with mesmerising circles and squares pulsing in time with the music. This earned it’s nickname among our group, The Orbital Mind Control Laser.

Here’s a good video by a third party demonstrating it in action.

Mine stopped working several years ago, much to my distress. While it would spin up, no matter what I did, no lights would show. At the time I had a half hearted attempt to fix it, before putting it aside. 8 years later, I figured my skills and equipment was so much better, I’d have another go at fixing it.

Another regulator. This one 10v. Also not faulty. ©2021

Step 1: Replace the power supply. The official power supply had been dropped before I even got the unit, and it had shattered the case. It limped on thanks to some duct tape for a few more years before I finally, officially, discarded it. It had been replaced by a nice, high power PSU but even that had gone missing somewhere in the moves.
A bit of googling and I found the name of a manufacturer (Thanks to, of all places, a pinball forum. These were apparently mounted on top of “Laser War” pinball machines) and from there an instruction manual. It used a standard 12v 800mA centre positive power supply. I grabbed one from my stockpile (What? You don’t have a stockpile of 12v PSUs just lying around?), cleaned up the wiring on it (It had been centre negative at some point in the past, but the polarity had been flipped with a hurried splice. I removed that, resoldered it and applied shrink-wrap for a neater appearance. I also relabelled it to make it clear it was no longer centre negative) and tried it. We had the motor happily spinning so that was good.

Step 2: Disassemble the unit. Turns out it was a bit tricky to disassemble. Both the power switch and the RCA jack were both attached to the back of the system in such a way that they needed desoldering. Thankfully my desoldering station was able to make short work of this. I then needed to remove a cardboard light shield and then a handful of screws. I was now looking at the “guts” of the unit.

Step 3: Try and fix it. This one was a lot more vague. As the motor was running I know power getting in was fine, so I assembled it on the bench in pieces, soldered in a temporary power switch and started looking for potential fault points. I figured it would have to be on the main board as the both of the VU strips were failing at the same time. I checked the voltages coming out of the voltage regulator, and it was a rock solid 10V, so I was able to eliminate that. I resoldered a few dry joints that looked suspicious and used some contact cleaner on the three connectors that went from the main board to the rotating VU strips. I also resoldered the RCA jack and tried it.

Of course it worked first time.

(At this point I’m very suspicious of the RCA jack.)

Spinning guts. ©2021

Anyway after all this it was a simple case of reassembling the whole thing. Spinny lights for the win!
I’m in the process of designing a small board to act as a preamp / splitter so that I can plug in a 3.5mm jack from either a mobile or a Bluetooth module to use as a source, and pump that to an amp and to The Lyte at the same time. At the moment, the output of either of those inputs is a little too low to use consistently with the unit.

So there you have it. Hope you all had a break as fun and as productive as mine.

Published by ilike8bits

I collect old computers and consoles

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: