Since completing my GB A1000 build and playing around in Workbench with my Amiga Tank Mouse, I remembered how awful the old ball mice were to use.  They really need a mouse pad (who has one of those now?) and they’re horribly inaccurate.  What I really wanted was an optical mouse.  However, I wanted to keep the Amiga looking standard, which meant keeping the Tank Mouse.  What I really needed was the two combined..

So, off to Google to do some research.  I found a project on Aminet for a PIC-based PS/2 mouse controller.  It was designed to take a regular PS/2 mouse and convert it to serial protocol to plug straight into the Amiga.  Taking a look at the project details, in it’s most basic form using a PIC16F628 there were almost no external components other than a decoupling capacitor.  I figured there was a good chance of this fitting inside the Tank Mouse case, thereby keeping everything else looking original.

So here is the original Tank Mouse with its internals:

Tank Mouse original

and here’s the donor PS/2 mouse I chose, because the microswitches for the mouse buttons lined up with the Tank Mouse buttons (but more on that later)

PS2 Mouse original

I needed to hack the internals of the Tank Mouse about a little bit to get the PS/2 board to fit

Tank Mouse hacked

and trial-fitted the PS/2 board to make sure the optical lens and mouse buttons would line up ok.

Tank Mouse ps2 controller trial fit

Looks to line up ok

Tank Mouse underneath trial fit

I programmed and socketed the PIC and fitted it to some Veroboard along with the wiring from the PS/2 controller and the Amiga serial lead.  (Hot Glue may have been premature..)

Tank Mouse internals 1

At this point I though I’d cracked it.. however, the mouse hardly tracked at all.  It seemed that the optical lens was mounted too high in the Tank Mouse case for it to “see” the surface underneath it cleanly.  A redesign was needed.  I had to chop the plastic lens piece right down to embed it into the Amiga mouse as close to the base as possible.

Tank Mouse hacked optical lens

This seemed to work well.  However, now I’d lowered the height of the microswitches too, so the mouse buttons weren’t actuating properly.  My solution was to steal the button assembly from the original Tank Mouse and connect this into the PS/2 mouse controller instead of using its onboard microswitches.

Tank Mouse buttons retrofit

Another quick test (I’m glad I kept the connectors on the wiring harnesses..) and all was looking good.  Time to fix the new board into place (again!) using a hot glue gun.

Tank Mouse internals fitted

And finally a completed Optical Tank Mouse using the original connector cable and right-angled Amiga 1000 DB9 connector.

Tank Mouse alive

It tracks really well, I’m really pleased with it.  I’m considering adding some weight inside the mouse as it feels quite light compared to the original Tank Mouse but I’ll see how I get on with it.

Thanks to Denis Sotchenko for his PS/2 mouse controller PIC firmware which I used in this project.

I had an “incident” with my original Thorens plinth and as it turns out the Super plinth is different to the standard TD-160 so almost impossible to find second-hand.  So, I added a new plinth to my wishlist.

Well, one Birthday later and (courtesy of my Dad) here’s a new plinth in Oak from Russ Collinson at Layers of Beauty.


I drilled few pilot holes on the rear to mount the hinges for the lid and also tidied up the cables underneath to stop the possibility of them interfering with the suspension.  I used the original Thorens Super heavy base.

I’m really pleased with the plinth – the build quality and finish is excellent.  It looks great on the shelf, now I just need a new set of Proac speakers in matching Oak finish..


I loved the style of the Amiga 1000 – the desktop case with the keyboard “garage” underneath where it can slide away out of sight.  However it’s running pretty old tech now – even by Amiga standards.  They only have 512k RAM as standard and no expansion slots other than the same edge connector an A500 has – but upside down, just to make it even more difficult to fit expansions.  So, when I came across the GB-A1000 motherboard (designed by Georg Braun, hence the name) I immediately decided to try and source one.

The GB-A1000 is a motherboard using the orginal Amiga chipset but with 8MB RAM, a Flicker-Fixer, an IDE controller for Hard Disks and a 68030 processor (plus a 68882 FPU) to take the Amiga 1000 well into A3000/A4000 territory.

I came across an unbuilt kit on Amibay and set about tracking down all the various components required to build a working board.  It was the largest project I’d ever taken on with lots of surface-mount soldering required – quite a challenge.  These are some pictures from the motherboard build process.  I started this in 2013 and just about finished by the end of 2015!

Motherboard received – wow, that’s quite large..

GBA1000 Bare Board

I built up these Video Hybrid kits first to get me used to the SMT soldering before working on the mainboard

Video Hybrid 2

Start by soldering all of the small surface-mount components first

GBA100 Board birdfeed

and then.. well nothing for 2 years according to my image archive.  I built the board up quite slowly, had a few issues sourcing parts here and there, just worked in small chunks and then left it alone for a while.  Apparently I didn’t take any pictures either – until..

Tada! Ready for the first power-up.

GBA1000 ready for power up

Will it work first time? ..

GBA1000 first power up

Of course it will!  Just a quick power-up at this point to make sure it all appears to work before mounting the board inside the A1000 case.  I also programmed a newer Workbench 3.1 and the IDE ROM and added a lithium battery (along with a small diode to prevent charging).  The 2.5″ IDE connector for the onboard hard disk was particularly tricky to find – eventually I had to hack down another 2.5″ connector to make it fit the board properly.  Looks ok though.

GBA1000 44ide connector

I’m using that with a 32GB Compact Flash card via an IDE converter.

When I mounted the board into the Amiga case it wouldn’t sit quite flush.  When I looked closely I could see why – the motherboard D-plugs had pronged mounting pins which went through the board but these were hitting the posts in the case.

GBA1000 D plug pins

I had to cut these off in order to get it to mount flush against the case standoffs and then I could use the self-tapping screws through the top of the D-plug to secure the plug – and the board – to the case.

GBA1000 D plug pins 2

So, trial fit of the motherboard into my A1000 case:

GBA1000 trial fit

Everything seems to fit ok, let’s carry on fitting the PSU

GBA1000 board fitted with PSU

and then the rear panel

GBA1000 board fitted with rear panel

The complete system up and running.

GBA1000 running

Right now it’s booting Workbench 3.1 from a CF card and I’m in the process of loading up WHDLoad and a chunk of old games.

As part of refurbishing some Amiga 1000s I’d picked up, I wanted to clean up the cases which had yellowed over the years.  I’d tried retr0bright on the cases and whilst this worked in a few instances, in others I either got a bleached effect or I just couldn’t get all of the staining off.  Plus some parts of the cases were scratched, which retr0brite doesn’t help with anyway.

I’d been looking for something to respray the plastic with.  Conventional plastic spray paint seemed to be too messy and too thick – you needed special primer and then a plastic top coat.  Cases I’d seen sprayed had lost their texture and looked obviously resprayed.  I found some Vinyl spray paint which purported to do exactly what I wanted – change the colour of plastics without needing a primer and because the coats were very thin it didn’t affect the texture.  However, when I tried some of these paints I just didn’t get a particularly good result – I found even with very thin coats that the paint never seemed to really dry properly and would just get wiped off with IPA or other cleaners.

I kept searching and eventually stumbled onto Kolorbond.  This was paint specifically designed for plastics – specifically uPVC window frames – but also including ABS which is what the Amiga cases were made of.  I contacted the manufacturer and asked if they would colour-match against a piece of the Amiga case (the memory slot cover).  They said they would, and supplied me with a 1L tin of Matt-finish Kolorbond, along with WindowPrep which is used to clean the plastic before spraying.

Kolorbond WindowPrep

I stripped the Amiga cases down then started to spray them, first with dust coat of Kolorbond and then two “proper” coats which was enough to cover all the discolouration and the light sanding marks from where I’d cleaned up the scratches with wet & dry.  These are the pictures I took along the way.

Keyboard before:

A1000 respray keyboard base before A1000 respray keyboard top before

RAM Cover and Front Panel before (notice the original colour behind the RAM cover)

A1000 respray RAM cover before A1000 respray front panel before

Top cover and rear panel before (again, notice the colour staining – very bad on the rear panel)

A1000 respray top cover rear panel before A1000 respray rear panel before

One coat applied, getting better

A1000 respray RAM cover during A1000 respray front panel during

and the completed parts

A1000 respray keyboard side after A1000 respray keyboard top after

A1000 respray front after A1000 respray rear after

A1000 respray one case and keyboard after

I have to admin to being a little apprehensive about spraying these A1000 cases and possibly ruining them, but actually I’m really pleased with the results.  The only problem I had is that there was some “Commodore” text on the front panel which I had to spray over, but I think I prefer the look without this anyway. The colour match is almost perfect and the original texture of the plastic has been kept.  The paint doesn’t chip or peel off and also doesn’t wipe away with IPA.  They just look like new again.

Now on to building them back up into working machines again.

I built my Hypex-based 5-channel power amp a few years back, but it never quite had the sound I was looking for.

Since I was moving to Vinyl with my Thorens TD-160 Super, a friend advised me to try a valve amp.

I looked around for DIY kits and most were either horrendously expensive or underpowered.  I eventually found “Eiclone” – a 60W monobloc kit based on EL34s with PCBs and a build manual that could be purchased from Parks Audio.  The kit looked fairly easy to build due to the PCBs and could be tweaked quite easily to run different tubes if required.  I ordered the kit and sourced the parts required so I could start the build.  This is the build thread I copied from the forums over at


After ordering the PCBs I ordered all of the parts required, mainly from Mouser but as I’m in the UK the Hammond transformers came from Philip is extremely knowledgeable and recommended I use JJ KT88s instead of EL34s, a Tube Doctor 7025 RT010 instead of a 12AX7 and a Tung Sol 6SN7GTB. This was all ordered from who were excellent and cheap.

The box of parts arrived so I could start the build:


Here’s the bare Eiclone Rev B PCB:


I soldered all the small components on first and then the larger ones. I’m going to mount the PCB upside down in the chassis so the tube sockets are on top and the other components on the bottom.
I used Orange Drop coupling caps but only because I mis-ordered the original ones and figured these looked cool 😉
Component side (will be the bottom) completed:


The tube sockets were a bit of a nightmare to solder straight and level. There must be a trick to this but I didn’t find it.


Both boards finished, now onto the chassis.
This will be my least favourite part – I’m terrible at any sort of metalwork. I’m using a Hammond 16x8x2 1441-26BK3 chassis which is steel rather than aluminium so not the nicest thing to cut.
I planned the design using CAD software – output transformer at the back of the chassis, power transformer at the front and the choke directly underneath. I printed this out to scale and literally stuck it onto the case with tape:


I centre-punched the drill marks and then proceeded to drill all of the holes required. I used a step drill bit to do everything other than the tube sockets. I did those with a punch which made a nice clean job of it.


The cutout for the IEC power socket I did with a Dremel and a cutting disc. I had to grind a little more material away afterwards in a few places but a trial fit with the IEC socket showed that most of my terrible metalwork would be hidden by the bezel.


There was quite a bit of deburring to do on the inside of the chassis – I did this gently with the step bit.


Time to trial-fit everything together now. I bought some nice black button head screws to try and keep everything looking neat. The holes for the tube sockets were slightly off-centre as I found it almost impossible to keep the drill and the case still when drilling the pilot holes for the punch, but luckily it doesn’t look too bad. I’m covering up the access hole for the bias pots with a rubber blanking grommet. These cases were originally a stop-gap until I had a nice wooden chassis made, but since they turned out so well I’ll probably keep them like this.


Onto the final assembly now – the internal wiring. I tried to keep the power as far away from the input signal as possible. Lots of sticky pads and cable ties used. I’ve kept some of the unused cabling from the transformers in case I want to change things around later or reuse them for something else.


After a couple of false starts in the testing phase (the bias test points TP1 and TP2 don’t work until the tubes are plugged in – I set -50V using Pin 5 on the KT88 sockets) I was ready to power up.
The both worked first time! I’ve set the bias on the KT88s to 65ma (650mV at TP1 and TP2) at the moment but will go back and re-check at regular intervals.

This is the end result:




and here they are installed:


I’m very happy with the sound but they’ve been running for less than 2 hours so far so I’ll give them time to bed in and then re-check everything.
I’m off to listen to them now, thanks for reading.

This computer was purchased by my Dad back in 1982.  It was pre-ordered by him in 1981 but delivery delays meant it wasn’t shipped until something like June 1982.  I remember it shipped with the pre-release OS1.0 ROMs (quickly upgraded to OS1.2) and a “Provisional User Guide”, which I still have.

It was originally a Model A and has a Revision 2 motherboard.  The hole on the left side of the keyboard was supposedly for expansion ROMs and was delivered uncovered as in the pictures.  Later models had this hole covered.   Note the early style text (“bbc microcomputer”) on the clear plastic function key strip.

This machine was upgraded to [almost] Model B spec a little later – an additional  16K RAM was added (to total 32K) and the user port, 1MHz bus and Tube connectors fitted to the motherboard.  I think it’s still missing a couple of logic chips for the Tube interface but I can’t quite remember.  A little later still, a Solidisk DDFS disk upgrade was fitted along with a 5.25″ 40/80 track floppy drive.  Then a Watford Electronics RAM/ROM board was fitted, with 3 banks of “sideways RAM” and a couple of extra ROM sockets.

I found this buried in the loft.  It still has it’s original packaging which was just a plain brown box and poly inserts.

When I powered it up the PSU went pop due to old capacitors which had gone bad.  Once they were changed it had some strange display problems which were eventually cured by replacing a couple of RAM chips.  The Watford RAM/ROM board was rattling around inside the case so I had to remove it – I’ll leave fixing that for a later date.

I stripped the machine down completely and retr0bright’d the case and cleaned up the motherboard and keyboard.

It now works perfectly and looks nice and clean too.  I’ve since sourced some joysticks and will set it up at some point to play Elite on it again.

BBC Top 1

BBC Left Side

BBC Right Side

BBC Top Case Open

The Boozhound Labs pre-amp I built needed a power supply.  The basic option was just to use a basic “wall wart” DC power supply, but knew these were generally low quality, unregulated and quite noisy.  Since pre-amp quality is pretty dependent upon power supply quality, I decided I wanted to use something better.

I found a project by AMB Labs for a Sigma 11 power supply.   It’s a high-quality regulated single-rail power supply.  I’d configure it for 24v.

Again, I used a case from Modushop and a rear panel from Schaeffer.

Here are the images of the completed build.

PSU Side


PSU Top 3

PSU Rear

I wanted to replace my basic Pro-ject Phono Box preamp with something a little better.  After reading through some threads on I found a project for a JFET RIAA pre-amp from Boozhound Labs.

The kit came with a PCB and the parts required to populate it (including Russian PIO capacitors) but needed a power supply.  I made that separately.

As with my Hypex Power Amp, I bought a case from Modushop and had a rear panel machined and engraved by  Schaeffer.

Here are the pictures of the completed build.

Preamp Top 3

Preamp Top 2

Preamp Rear

I’m very happy with the result and the sound quality and have been using it with my Thorens TD-160 Super.

I wanted to new multi-channel amp to replace my previous Marantz MM9000 power amp.  After reading various reviews on I decided to use Hypex modules.  Specifically, UcD180HG 180w amp modules along with Hypex’s SMPS power supply modules.

I used a chassis from Modushop and had a rear panel machined and printed by Schaeffer.

These are the images of the competed build.

Amp Front

Amp Top 2

Amp Rear Lid off

Amp Rear

Because the amps expect a balanced input I also needed to make some phono to XLR input cables.

I was very happy with the result though later replaced the front two channels with a tube amp as I found the sound a little too clinical.