The restoration of two Amiga 1200s

      No Comments on The restoration of two Amiga 1200s

As a brief interlude, and before getting back to FPGAs, I wanted to document something I consider to be tremendously important: the restoration of my beloved Amiga 1200.

I own two of these machines: the original one I bought in the summer of 1993 just before starting university, and a “spare” I bought off eBay about 10 years ago. As an upfront warning, I did not expect this project to have a great result, so the before pictures are nonexistent.

Roughly speaking the eBay machine was in a reasonable condition as I received it. It’s biggest issue was, and remains, the quality of the video output: vertical interference lines are prominent, making the machine fairly unpleasant to use. In terms of the quality of the plastic, which is a sure sign of the age of the machine, the eBay machine wasn’t awfully yellowed. Not new looking obviously, but not bad either.

At the point I started the restoration of my original machine, it had two main defects: It was very seriously yellowed, and the sound output was nonexistent; over time the volume of the sound produced became quieter and quieter to the point that the output became imperceptible.

Because my original machine was and is so precious, I decided to trial my restoration techniques on the spare machine.

The first step then was to dismantle it with a view to recapping the main board. That is, replacing both the SMT and THT electrolytic capacitors. Recapping it would give the board many more years of good service, and hopefully also cure the problem with the video output.

After gaining access to the board a visual inspection revealed nothing out of the ordinary, just a fair amount of dust. I gave it a quick clean with IPA and a toothbrush. The shielding around the PCB was also slightly rusty, as happens with these old machines.

The process for recapping an Amiga 1200 is not worth describing in detail here, since there are so many videos describing the possible methods, such as this one. The first thing I did was purchase a kit of capacitors from eBay, like this one, but in the end I bought a box of capacitors from Amazon as well as I needed a second pack and I thought they would be useful for other projects. The method I went for was to use hot air, as shown in the linked video, to remove the old electrolytic capacitors and to solder the new ones. The principle reason for wanting to use hot air, with a lot of aluminium tape, was to prevent damage to nearby components. I don’t trust myself not to slip with the iron and damage a nearby plastic part:

For what it’s worth, I find that aluminium tape is a lot better at reflecting heat then Kapton tape, which some people prefer.

The above picture shows some bulging at the top of at least one of the capacitors, indicating that they should indeed all be replaced.

The use of hot air presented it’s first problem: I had two exploding capacitors, despite attempting to heat them in the quickest, yet not too quick, way. The likely reason is simply the age of them. I suppose next time I will have to test myself with the iron.

A second problem I had was a damaged pad. This was nothing to do with the usage of hot air, but rather was caused by the removal of a through-hole capacitor with an electric solder pump. In the end I managed to attach a new capacitor by soldering the lead to the top of the board:

I wasn’t best pleased with damaging a 30 year old board, but at least the recap job was done. After briefly testing the board on the bench I discovered that the video issue was not resolved by switching out the electrolytic capacitors. A little bit deflated, I moved onto retrobrighting the keyboard keycaps, and the case.

The first job was to pull off the keycaps from the base of the keyboard to give the whole thing a very good clean. This was fairly physical work, and I managed to break the plastic clip that holds the retaining bar to the space bar. This was extremely frustrating; I’d basically turned an antique keyboard that was a little yellowed into a paperweight. Luckily this wasn’t my original Amiga 1200!

Nonetheless I continued on with the retrobrighting in order to familiarise myself with the process. There are a few different methods, but I went for using 6% Hydrogen Peroxide in a cream form. The keys were placed on clingfilm, which in turn was placed on a tray, coated with the Hydrogen Peroxide, and then the clingfilm placed over the keys to create a semi-airtight seal:

I was amazed at how they looked after only about 4 hours in the sun, and it wasn’t even a particularly hot summers day.

The case was retrobrighted in a very similar way. I had to move the case around so each part was in the sun, but after only a couple of afternoons the case was looking like new.

The result was I had a near pristine looking Amiga 1200, with only one problem: it had no working space bar.

For my original Amiga 1200 I opted not to perform a full recap. The experience with a broken pad and exploding capacitors was enough to dissuade me from that course of action. Instead I was only concerned with fixing the problem with the sound output, and performing a full retrobright on the keyboard and case.

After dismantling the machine, I performed the usual visual inspection. I was fairly surprised by how decent it looked; the expected amount of dust and a very small amount of corrosion notwithstanding was nothing to be too concerned about. Again I cleaned the board with IPA.

A great thing about working on Amigas is the availability of schematics at the Amiga Wiki. Looking at the Amiga 1200 schematic, C324 and C334 are key capacitors in the left and right audio amplifier paths. I replaced both of those, which was fiddly based on the fact they were near the plastic keyboard flex connector, but no improvement in sound output was noticed.

I spent quite a few hours trying to figure out the problem. After eliminating capacitors as being a cause of the problem, and having a read of several webpages including this one and measuring some key points in the audio circuit, I came to the conclusion that it must be the LF347 (PDF) quad op-amp:

This picture was taken after giving the board a clean, and I’m still not convinced that the discolouration on some of the pins is proof that the part was faulty but that was indeed the case. Swapping the IC was done extremely carefully as I was determined not to cause any further damage to the board and the result was a total success: I had working sound on my precious Amiga 1200 after more then a decade of silence.

Retrobrighting the case and keyboard was done as described above. This time I was very, very careful when pulling keycaps from the keyboard! I also retrobrighted the mouse.

Here’s a picture of my Amiga 1200 reassembled and plugged into my Dell ST2210 monitor. This monitor is exceptionally useful for this setup as it is capable of showing all screen modes an Amiga can generate:

The Amiga is stock with no case modifications or any non period-specific hardware. The only addition is a Blizzard 1230 MKIV with 50MHz MC68030 and MC68882 I purchased in 1996. After getting the occasional Guru Mediation from a 32MB SIMM (which works fine when used in MAXI030) I swapped in an 8MB stick, which has been working reliably now for a few months.

Back in the day I used a 500MB 2.5″ harddisk in this machine, but these days a CompactFlash acts as storage.

In terms of software, I’m running the last iteration of Workbench I had setup, which I used up until about 1997. These days games can be loaded from the CF with the excellent WHDLoad, which makes starting any game ever released for the Amiga a simple matter of double clicking on an icon.

The only negative to this whole process was the damage done to the keyboard of the eBay machine. Many months later I am still frustrated with myself for causing this damage. I suppose I will keep it for spares, or possibly sell it for parts. I imagine even in its current state it would still fetch a reasonable amount.

My original 1993 machine, and its Blizzard accelerator, will never be sold however.

The next post will be back to the topic of FPGAs and microcontrollers…

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.