Amiga

Restoration of a barn find Amiga 2000: part 2

With this Amiga 2000, I previously got it into a state where it would boot from the main CPU and the accelerator card. Now it is time to continue working on it.

PSU Testing

I tested the PSU with a hard drive for a load, but it kept restarting. Judging by what I could figure out about the circuit in the PSU, it requires enough of a 5V load to start up properly. I knew the voltages were safe, so I decided to plug it into the motherboard and try it.

If you ignore the fact my test desk is a bit of a mess right now due to working on several projects, the results are quite good. The Fluke desktop multimeter is connected to the 5v line and is showing 5.1v, this is within tolerance. The 12v line is showing 12.19v, which is pretty good too. It is now stable.

The fan is a little noisy compared to a modern solution, but it is riveted in, so I may hold-off replacing it for now.

Keyboard connector

The metal plate on the keyboard connector had been hit hard with corrosion, I tried cleaning up the corrosion, and then I noticed corrosion all over the legs, so I decided to replace it, as it is likely the corrosion is inside the connector as well.

The replacement doesn’t have an outer metal shield, but should work well.

Fuses

The Amiga 2000 has several soldered on fuses. These can fail as all fuses can. I tested all the fuses on this motherboard, and sure enough, one of them had blown, F3.

This provides 12v to the external floppy drive port. These fuses are Littlefuse model 0251004, which are a fast-blow 4A fuse. I sometimes replace these with PTC fuses, but this time I decided to use the real thing.

This is the removed fuse, as you can see, there is no visible way of telling that it has blown. But, I checked again after removing it, it was dead. I replaced it with the same fuse type.

Battery overkill

There are many options to replace the RTC battery, in this case I chose to do use a Li-SOCL₂ battery. This is a lithium AA battery, which is designed for very long life at low power and no leakage. Whilst it is in an AA form, the output voltage is 3.6v, it is also not rechargeable. It should last around 10 years keeping the RTC running.

So, I did this modification in two parts, the first was to solder a connector to where the battery used to be.

Then I took an AA battery holder, soldered a diode in series with it to prevent charging, and added pins to connect it.

Testing

Time to test that everything on the motherboard works correctly. I’ve set up a test rig, it’s not pretty, but it should work.

I ran all the AmigaTestKit tests, and it all looks good!

PSU fan

I decided to go ahead and install a quieter PSU fan after all. I’m a fan (pun not intended) of the Arctic P8 Silent right now, so I’ll be installing one of those.

First I had to remove the rivets holding in the old fan. I managed it, but there were some casualties along the way.

Then I fitted the Arctic and spliced a space Noctua extension cable onto the PSU’s fan power wires so that the fan could be easily unplugged removed if needed in the future. This is the end result.

I then closed it up, hooked it onto the motherboard and tested. The fan ran so silent that I had to double-check it was actually spinning!

FDD testing

I wanted to make sure the floppy drives were good before reassembling. So, I hooked each one up to my Greaseweazle, ran a cleaning disk and then an Amiga disk.

Both drives worked perfectly!

Accelerator RAM

In my previous post, I mentioned that the 040 accelerator card was showing as 2MB RAM. I thought this was a little odd because there were 8 SIMMs in there, unless they were 256K that doesn’t check out.

I did a little bit of research into this and found a couple of things.

First of all, the card initialises 2MB of RAM in the 24bit Zorro II memory space (which can have a maximum of 8MB autoconfigured RAM). You then need to run a tool to configure the rest of the RAM in the 32bit space. This card would likely pre-date the Zorro III memory area autoconfiguration. According to the documentation, the RAM in the 32bit area runs faster than the 24bit RAM.

Now, for the SIMMs themselves. The chips on them have custom branding by the board manufacturer (Progressive Peripherals & Software) and are difficult to identify. I managed to find one of these SIMMs for sale and found they are 1MB 80ns SIMMs. Which means this card has 8MB of RAM. It is capable of taking 8x 4MB SIMMs to provide 32MB of RAM. This would have been huge for an Amiga 2000 at the time.

Once I have everything running together, I’ll investigate this further.

Next time

It is time to reassemble the machine. I have a few things coming to help finish it off, and I really want to see what is on the hard drive. Due to delivery delays that are typical over December, it might be a couple of weeks before I have another update, but watch this space!

LinuxJedi

Recent Posts

Restoration of a barn find Amiga 2000: part 3

All the motherboard issues were resolved in my previous post in this series, now it…

2 hours ago

Two special Amiga 4000s: Repairing Jops

Last time I worked on Jops, I left myself a lot of work to do.…

2 weeks ago

Restoration of a barn find Amiga 2000: part 1

I recently acquired an Amiga 2000 for £350 which was in an unknown state, but…

2 weeks ago

Diagnosing an Amiga 1200 Data Path Fault

I recently acquired an Amiga 1200 motherboard in a spares/repairs condition for about £100 recently.…

4 weeks ago

Bare Metal “Hello World” on an STM32MP135F-DK

Whilst I do like working with STM32 development boards, some basic information I need can…

4 weeks ago

Two special Amiga 4000s: Diagnosing Jops

When I last left this blog series, the first of Stoo Cambridge's A4000s had gone…

1 month ago