This time, I have a true curiosity for you. 🙂

It's a ZX Spectrum 128K. It was designed by Sinclair and their Spanish distributor Investrónica, and was a major upgrade of the ZX Spectrum 48K. At that time, the 48K model was rather outdated with its limited RAM and the simple sound beeper, and Sinclair had nothing in its hand to compete with the Commodore 64 which was gaining ground in more and more households.

The 128K model has 128K of RAM (which also allowed double buffering), an AY-3-8912 sound chip, an RGB monitor port, a serial port and an optional external numerical keypad. Hardware sprites are still missing though. On software side, it provides a heavily improved BASIC.

The shape of the prominent heat sink on the right side of the case gave the machine its nickname: "Toast Rack".

The model was first sold in Spain, as Sinclair UK still had a large number of unsold 48K models. In the end it could not save Sinclair from bankruptcy, but the 128K model was certainly very attractive for the new owner Amstrad. Today, the Toast Rack is a sought-after item for any serious Sinclair collector. (If you want to read more about the history of the ZX Spectrum 128K, there is a great post at 21twice.com!)

What makes this special model so curious is that it is an Arabic modification. 😀 There are stickers with Arabic letters on the keyboard, and at the front side there is a switch to select between the original 128K ROM and an Arabic version of the 48K ROM.

My first thought was that this was an elaborate DIY modification. But then I found a thread in a Sinclair forum. It says that Matsico, a Sinclair/Amstrad agency in Egypt, has produced these models. I could not find more information about it though, so I don't know if they were actually sold, or just made as a proof-of-concept or promotional gift.

What they all have in common, is the switch at the front, and an EPROM soldered on top of the ROM. The only known exception I found in a video by ByteDelight about a ZX Spectrum +3, where the ROM could be selected via a separate boot menu.

Restoration

A first diagnostics run showed that the board was working fine. The only issues were massive picture interferences, and an almost inaudible sound from the AY chip.

Both issues are known problems with that model. A blog article by Adam's Vintage Computer Restorations addresses them.

First of all, I replaced all electrolytic capacitors with premium ones. I'm doing that with all retro machines, irregarding their age and rareness. However I try my best to maintain the "retro optics", for example by using axial caps in that classic blue color.

To enhance the image quality, I used a 47µF electrolytic cap for C28 (original was 22µF), and replaced C7 and C8 with 1µF MLCCs. I could also rewire C126 as mentioned in the blog article, to enhance image quality even further, but I decided to postpone that.

To raise the volume of the AY sound chip (so it has a similar level as the beeper), I replaced R115 with a 1.65kΩ resistor.

The 7805 voltage regulator is rated at 1A, and is working at its load limit on the 128K. This is the reason for the big heat sink on the right side. I replaced it with an 78S05, which is a drop-in replacement that is rated at 2A and stays considerably cooler.

I was lucky here, because in the past, someone had already replaced the 7805 with a LM1085. It is rated at 3A, but has a different pin configuration. If I had replaced it blindly, it would have killed the machine. You should always be prepared for nasty surprises when restoring old machines the previous owner already tinkered with!

I also cleaned the case (although it was already in a very clean state). The previous owner had already replaced the keyboard membrane, but the extra keys were not working there, so I replaced it with a new membrane of a known-good brand.

The next diagnostics run showed that all tests were still green. Also the picture interferences were mostly gone (except of some minimal jailbars that I can live with), and the AY sound is much louder.

Depending on the position of the switch, the system either boots into the original ZX Spectrum 128K startup menu, or shows an Arabic boot prompt. In the Arabic version, the entire BASIC has been modified, with all texts in Arabic and written from right to left. Unfortunately I cannot read it.

A short test with the Dandanator module also showed that games are working fine. The mandatory part of the restoration is completed!

Freestyle Restoration

There were two more things I didn't like.

First was the ROM stack. The original solution switched the Vcc pins of the ROMs, so one of the chips was always powerless, but still connected to the address and data bus. To be honest, I wonder why this was working at all.

Anyway, I replaced it with a single 27C512 EPROM. On the bottom half of the memory, I burned the Arabic ROM (twice), and on the upper half, I burned the original 128K ROM. After that, I modified the switch to pull the A15 address line (pin 1) either to GND or Vcc. This way, the EPROM is always powered and the desired operating system is selected by an address line. I also upgraded the original Arabic ROM version 1 to the latest version 3.1 I could find on the web.

Of course I will keep the ROM stack. Mainly for licensing reasons, but also so that the original solution can be restored if desired.

I also didn't like the optics of the naked switch at the front, so I 3D-printed a small switch cap that also covers the screws.

There are a few more things that could be done:

• I could also rewire C126 (as mentioned in Adam's blog article above), to remove the sound signal from the RGB output.
• Due to a bug in the original PAL10H8, the system crashes just by reading the $7FFD port address. There is a fix that also removes a "rain" effect caused by refresh data on the bus.
• The original ULA can be replaced with a vLA128, as a replacement if broken, or if the precious original part should be conserved.
• Dave Curran reverse engineered the numeric keypad. An ambitious tinkerer could make a DIY keypad replica.

I recently got two Competition Pro Mini joysticks. The full-size Competition Pro was probably one of the most famous joysticks back in the 1980s and early 1990s. This model was said to be unbreakable, and was able to withstand even long and intensive gaming sessions. If one of the microswitches eventually failed, it was easy to get a new one from an electronics store and replace it just by using a screwdriver, no soldering required.

The Mini models came to the market in 1992, and were by far not that robust. My two examples had broken microswitches at the left direction. The microswitches are soldered to the PCB, and they are also out of production.

The closest available replacement is the Saia-Burgess F4T7UL and F4T7GPUL (the latter one with gold-plated contacts). Unfortunately it has different solder tails, so it cannot be used as a drop-in replacment.

I decided to build up a completely new PCB by InsaneDruid instead. It is a replica board with exactly the same size and fuctionality, but it is prepared to use the Saia-Burgess switches.

Auto-fire Model

The first thing I did was to completely disassemble the joystick. All plastic parts were cleaned in an ultrasonic bath. Meanwhile I noted the color order of the wires before cutting the cable from the old board.

The new board only needs very few components. Except of the switches, all of them are standard ones that can be found in any electronic store. The switches can be found at distributors like Mouser.

For the four directions, I first mounted the microswitches to the holder frame. This way it will be easier to perfectly align the switches to the PCB.

For the fire buttons, the middle solder tail needs to be cut off. Otherwise it will collide with a peg of the bottom case shell later.

Usually we would start with the flattest component, but in this case, I recommend to start with the four direction switches. Position them to the frame so the four screws are perfectly aligned with the corresponding PCB holes. Make sure that the two LED holes are aligned to the "up" direction. Then start soldering the switches to the PCB, using a generous amount of solder.

Additionally you can use wire brackets to secure the switches using the provided mounting holes. I was too lazy to do that though.

The remaining components are just soldered to the board. Finally the original cable is wired to the board, with the original order of wire colors.

After that, the joystick is ready for reassembly.

Standard Model

My other joystick is a standard model without auto-fire function. Again, I disassembled it. All plastic parts were cleaned, while I noted down the color order of the wires.

The cable of this model does not provide a +5V supply. For this reason, there is no need to populate the components for the auto-fire. We can also save the LEDs, as they won't light up.

The preparation is the same as for the other joystick. The four direction switches are mounted to the frame, and the two fire switches lose their middle solder tail. After that, the frame with the switches is aligned and soldered to the board, followed by the fire switches.

We do not need to populate the mode switch. However, we need to solder in a wire bridge to pins 5 and 7 (see photo), otherwise the fire buttons won't work at all.

The wires need a different order. On the original board, the "fire" signal is at the third position from the right (the orange wire on my joystick). On the new board, that signal is on the very right pad. The one left from it stays empty. For the remaining wires, the original order can be kept.

After that, this board is completed as well, and the joystick can be reassembled.

The joysticks look as good as new now, with their clean case and their brand new boards.

The ENIG plated PCBs are a true eye-catcher in their transparent cases.

After being on Twitter for 13 years, I decided that it is time to leave the platform for good. They say you should always leave on a high note. Maybe I have missed that moment already.

Today I have closed my account there. My handle was shred_ (with a trailing underscore).

You can follow me in the Fediverse: @shred@oldbytes.space

Thank you, blue bird! I have learned a lot from you and met a lot of great and interesting people. It was fun while it lasted.

PS: I have no plans to join other social media platforms. No need to send me invite codes. 😉

In the first part I successfully repaired an Amiga CD32 that got broken due to leaking capacitors and a botched restauration attempt. In this part I replace the laser pickup and calibrate the CD drive.

The old laser pickup of the CD32 might be worn out due to age and use. A common symptom is that the CD32 is unable to play CD-R media, or it is only capable of playing music CDs. There is no way to make the CD32 accept CD-RW media though, since they use a dye instead of pits that reflect too little light.

But before we start, read this:

CAUTION: The laser pickup is very sensitive to ESD. Use protective measures (such as an antistatic wrist band).

Make sure that the laser is always covered when the machine is turned on. Do not look into the laser beam.

I should also mention that I am not a trained technician. I have read manuals about how to calibrate CD drives, and it has worked for me. However, I don't claim that this is the best or most professional way to do a calibration.

You will need a soldering iron for the pickup replacement, and you will definitely need a scope for calibration. The drive might work without calibration after replacing the pickup, but the result will not be optimal.

Pickup Replacement

I started with disassembling the CD drive. I removed it from the case. Then I carefully disconnected the pickup and the motor unit, and removed the four screws that hold the pickup frame. There is a metal shield covering the pickup that needs to be removed as well.

The laser pickup unit is a Sony KSS210A. It is long out of production, but replicas are sold at online marketplaces for a few bucks. To remove the old pickup, I first removed the white cog wheel, then I pulled out the metal rod (it is secured by a plastic clip that can be pushed to the side). Since I was on it, I cleaned the old grease from rod and the cog wheels, and applied a bit of fresh silicone grease. After that, I mounted the new pickup and reassembled the CD drive just in the opposite order of disassembly.

After the new pickup unit has been connected to the controller, a solder blob on the pickup unit must be removed! It protects the laser from ESD, but will damage the drive controller if it is still there when powering on the drive.

If you want to keep the old pickup module as a backup, you can also apply a solder blob there before disconnecting it.

Preparation

For calibration, I opened the metal shield of the drive controller, and found a surprise underneath. There was a tiny board glued to the main PCB, and connected to some points with seven wires:

I first thought this could be some kind of mod to circumvent copy protection measures, but then again, the CD32 does not have a sophisticated copy protection scheme. Later I found the answer in a YouTube video: This modification immediately cuts the power from the laser and the spindle motor when the lid of the CD drive is opened. I could find many photos of the controller board without the modification, so I guess that it was a product safety requirement for selling the CD32 on the German or European market.

Okay, let's get back to the calibration. As a preparation, I first soldered wires to the VF, RFO, TEO-1, and FEO-1 test points. I recommend to use wires of different colors, it makes the calibration much easier. Unfortunately I only had red wire at hand, so I had to check each time which wire went where.

After that, I noted down the current settings of the four pots on the controller board, and of the pot on the laser module, using an ohmmeter. If I should mess up the calibration for some reason, I could always go back to these settings. (A photo of the pot positions is not sufficient, as very tiny changes can already make a huge difference.)

For the calibration, the drive needs to be connected to the mainboard again. The case top (with the LEDs, reset button etc) needs to be connected as well, since the CD32 won't attempt to read the CD unless the drive lid is closed. The laser pickup is moving during operation, and should have sufficient room for that.

To fix the CD to the spindle, I removed the spindle clamp from the inside of the lid, and used a bit of tape to keep the loose part fixed in the center of it. It is held to the spindle with a magnet, and ensures that the CD won't slip on the spindle.

Calibration

The calibration process is explained in this blog article by TSB. My attempts to explain it would be far worse. 😉

However, it turned out that on my drive, the process didn't work like that. After doing the first steps of the calibration, my drive was suddenly unable to spin up the CD for reading. I was lucky that I noted the pot positions (like recommended above), so I could revert to the original settings and start anew.

Then I first calibrated the TEB pot until there was approximately 0 mV between TEO-1 and VF. The drive was still working after that. However, after I calibrated FEB like documented, the drive stopped working, so I reverted that change again and moved on with calibrating the laser power.

CAUTION: Be very careful with the pot on the laser module and only turn it in very small increments. Otherwise the laser may be permanently damaged.

There is a drop of varnish on the pot from production that may require some force to break, so it might be a good idea to first turn the pot while the device is powered off, and then use an ohmmeter to return it to the factory setting that you previously noted.

To calibrate the laser power, I connected my scope to RFO and ground. Then I put a music CD on the spindle and started playing track 1. The scope should now show a so-called "eye pattern":

The tricky part is to turn the pot on the pickup module carefully while the CD is playing. I turned it very carefully until I reached a peak-to-peak voltage of about 900 mV. Take care never to exceed 1200 mV!

After that, I adjusted the FEB pot on the controller board until I reached a maximum amplitude on the eye pattern.

The last two pots, FEG and TEG, are calibrated by scoping the FEO-1 and TEO-1 test points against ground, respectively. The drive should play track 1 of an audio CD and should be in pause mode while calibrating.

I tried to find the sweet spot where the signal on the scope was as smooth as possible, and the correction noise from the optics was as silent as possible. There is a trade-off between these goals, and I found that the best results came from listening to the pickup noise and using my intuition.

The calibration is complete after that, and the CD32 can be assembled again.

One final tip: burn CD-Rs for your CD32 at the lowest speed supported by your recorder. This will increase the contrast of the data on the CD. Also, prefer CD-Rs that are not transparent when held up to the light.

I found this CD32 for a fair price, and bought it. The optical condition of the case is quite okay. It has some visible scratchmarks. The previous owner tried to fix them, but made it even worse. At that time, I wasn't aware yet that this would be the main theme of the whole restauration.

Together with the console, I got a PSU and an edutainment CD for learning math. The PSU wasn't the original one, but a simple power brick with a CD32 connector soldered on. The gamepad was missing, unfortunately, but I found a Honey Bee joypad as replacement a bit later.

Let's have a look inside the machine.

The State

The seller sold it as broken because it showed no picture. When I opened the case, the machine told me a completely different story. There was an attempt to recap the machine. It was abandoned after replacing the TH and the 100µF SMD caps, probably because the picture was gone after that.

I also found blotches of green varnish, presumably simple nail varnish. It was under the replaced SMD caps, but also on solder joints and some vias. The varnish made no sense at all, except of maybe cosmetical reasons.

And I found this:

I assume that when the picture was gone, the guy who tried the refurbishment assumed that the video encoder chip got damaged, but had no equipment at hand to unsolder an SMD chip, and attempted to cut it from the board pin by pin instead.

I found no further traces of mistreatment of the poor board. It's going to be enough of work to fix the current mess already.

To be honest, I am pretty upset about that. There is a difference between if the machine shows no picture after decades of storage, or because of a botched refurbishment attempt. The seller should have pointed out that fact.

Fixing the Mainboard

First I attempted to restore the picture by replacing the obviously broken video encoder chip. I also replaced an electrolyic cap next to it that looked suspicious. Unfortunately that did not bring back the video signal.

The question was now whether I was getting no picture because of further errors in the video area, or because the machine is not starting at all. To find out, I inserted a DiagROM and connected the CD32 to my PC. The DiagROM started and logged no errors to the console. So the good news was that the machine is basically working.

I then decided to remove everything from the previous restauration attempt, so I could start anew with a known state of the mainboard. I removed all the electrolytic caps, even those that had already been replaced, and cleaned off the green varnish with acetone and IPA.

There was a strange solder blob on the bottom side, covered with a layer of varnish. When I tried to clean it up, I smelled that revealing fishy smell of old electrolyte. I generously removed the SMD parts on both sides in that area, cleaned the board and checked the tracks and vias.

Unfortunately, I ripped off a few pads on the 100µF capacitors while doing so. I guess the leaked electrolyte and the thermal stress of two recappings was just too much for them.

Then I soldered in new components in that area, and fixed the broken pads with bodge wire. For two SMD capacitors, the board offered an alternative use of TH caps, which I thankfully accepted. The area is now looking quite ugly, but at least it should work again.

When I checked the tracks and vias at the other 100µF SMD caps, I found broken connections at C236 and C237. They are used for the luma or composite video signal, so the broken connections caused a black image.

I also found a broken via near C409, which carries the CSYNC signal. The missing connection causes a missing video sync signal at the outputs. I fixed it by opening the via and exposing the connected tracks on both sides, then soldering a thin wire to the tracks.

So there were more than enough reasons for this board to show no video picture.

The TH capacitors on the board are a bit special. For C408 and C811, the silk screen shows the positive end at the wrong side. Even Commodore soldered in the capacitors in the wrong orientation, and you will find many CD32 out there with bloated caps at that position. I decided to solder in SMD caps there instead, which can be soldered in like shown on the silkscreen.

After that, I checked the machine, and to my amazement, it was working again:

So the mainboard was repaired and refurbished. I checked all the video and audio connectors, and found a signal everywhere. The machine was also running stable.

I'm glad that the machine turned out to be repairable.

In the next part, I will replace the laser module and calibrate the CD drive.