I’ve been slowly but surely branching out into Amigas from Commodore’s 8-bit range of computers, and as such I recently got my hands on my first Commodore Amiga 1200.
The A1200 is a 32-bit computer from the 1990s, and is effectively a more capable version of the A600, featuring a 14MHz 68020 CPU (as opposed to an 8MHz 68000), 2MB Chip RAM (as opposed to 1MB) and AGA graphics (as opposed to ECS graphics), and a full-sized keyboard – like the A600, it also features an onboard IDE and PCMCIA interface.
This A1200 was a bit of a basket case, not only in need of a good clean and service, but also some repairs – the computer booted correctly but had distorted audio, and the internal 3.5″ FDD would not load disks (even in another machine).
As it turned out, both of these faults were down to the Achilles’ heel of the later Amigas.
Unlike the A500 which was manufactured using through-hole electrolytic capacitors, the A1200 and its little brother the A600 primarily use surface-mount (SMD) electrolytic capacitors – being a relatively new technology for the time, these capacitors are especially unreliable and frequently leak, causing a lot of problems in this era of computers.
Physically (as opposed to electrically) leaky capacitors not only have compromised performance, but the leaking electrolyte can have corrosive effects on surrounding components and traces, damaging or weakening the PCB.
Electrolytic capacitors can still leak in storage, so the only practical option is to replace them – this is a fiddly job which is not easy do properly, even with the practice and correct tools required for SMD soldering/desoldering, but is definitely worth it.
The first step, therefore, was to replace all of the electrolytic capacitors on the A1200’s mainboard – the SMD ICs are generally very reliable, and one electrolytic capacitor in the reset circuit is known to cause black screen faults.
When doing this, I remove the PCB from the case then remove the RF shield and plastic insulating layer; then, I desolder (using a Duratool D00672 set at 350C) and remove the four through-hole electrolytic capacitors as well as the keyboard connector, which gives better access to the SMD capacitors along the back edge of the board; then, one by one I remove all of the SMD electrolytic capacitors using a hot-air rework station (set at 250C and high flow, with a small nozzle installed) and a pair of tweezers, carefully covering any surrounding components with heat-reflective tape and, for large areas, aluminium foil.
I find that if you aim the heat gun at the base of the capacitor for long enough, it will fall of its pads when pushed sideways with barely any force, minimising potential PCB damage.
I then apply rosin flux and leaded solder to all of the newly exposed SMD pads, then desolder it using the desoldering gun, leaving a clean pad – I then clean up all the remaining flux or heat marks using 99.9% IPA.
I would recommend using a fume extractor when doing this kind of work, primarily because of the solder and flux fumes, but also because leaky SMD capacitors have a habit of going pop and letting out all the magic smoke once they’ve heated up.
Once the board has been cleaned up, it’s time to install the new capacitors – for this project I used new capacitors with a solid electrolyte, which won’t suffer from leakage in future.
For this, I install all of the SMD capacitors one by one, making sure to prioritise the least accessible ones, and making sure to install them in the correct orientation – if not, they’ll go pop when the machine is powered on. I usually apply rosin flux to both pads on the board and capacitor, then hold the capacitor in place on the board using BluTack, then tack the capacitor in place using leaded solder and a small-tipped soldering iron (set to 350C).
Then, the through-hold capacitors and keyboard connector get installed – some connections require a higher temperature setpoint on the soldering iron, as the ground plane draws a lot of heat away from the vias.
After testing, the A1200 appeared to be working fine again – happy days! Time for some upgrades – both for cosmetics and for performance.
First off, I dismantled and thoroughly cleaned the case and keyboard.
I also dismantled, cleaned, and lubricated the original FDD (a TEAC FD-235F), but this didn’t solve the disk loading problem. The FDD appeared to be mechanically fine, so the problem was likely electrical: I cleaned all the connectors to no effect; I then noticed that an SMD electrolytic capacitor on the drive logic board was quite crusty, so decided to replace all the electrolytic capacitors in the drive (a 4.7uF 25V SMD and a 100uF 10V radial). After doing so, the drive started working perfectly.
Then, I fitted a KickStart 3.1 ROM, internal IDE-CF adaptor, and an internal WorkBench 3.1 CF card, allowing internal storage without the need for a HDD; I also fitted an 8MB Fast RAM expansion with onboard RTC and 40MHz FPU.
Soon I’ll be treating the keyboard for yellowing, and fixing the PCMCIA reset issue.
I’ve also had an Amiga 600 on the bench – pictured below alongside the A1200 – also requiring a recap. This has already featured on the blog, here.