I was at the Norwich Games Festival last week, along with six of Amigas and a BBC Master 128 with Valiant Turtle. Whilst I was there, OMTG Retro handed me his Amiga 1000 and asked if I could fix it for him, it wouldn’t boot. So, I had a look into it.
The Amiga 1000 does not have the entire “Kickstart” boot system in ROM, which is what you would expect with other models. Instead, it has a special extra RAM area called a WCS (Write Controlled Store) and on boot, the “Kickstart” ROM is loaded into this from a floppy disk. It is the only operation you can do on power-up. Instead of getting an image asking for Workbench, you get this image:
Once you have provided this, the boot happens, and it requests the Workbench disk as normal, or of course, whichever disk you want to run.
Now, the problem with this Amiga was that the Kickstart disk was not working. After reading the disk, the machine would lock up with a black screen.
The first thing I tested was the floppy drive, the Kickstart disks don’t appear to have great error checking. So, if this was bad, then it would explain things.
The easiest way to test this is to plug the drive into my Greaseweazle and have it read the disk.
The disk read fine. I also created my own floppy disk with the Kickstart on it, and this also was not booting.
This pointed me in the direction of something possibly being faulty in the WCS area.
A couple of years back I modified the “buptest” in PiStorm, a utility which tests communication between the PiStorm and system RAM, to talk directly to the WCS RAM. This is extremely useful as a diagnostic tool, as you don’t need most of the Amiga working to be able to run this test, the PiStorm is talking directly on the address/data bus to the required chips.
I tried this out and errors were definitely found.
The highest nibble on odd bytes was getting corrupted. This would make sense because the RAM is in the configuration of 4bit chips. Very likely that one of these is faulty.
This is what the motherboard looks like, it is a later model with the WCS soldered directly to the board, instead of requiring a mezzanine board.
I actually haven’t worked on this type before, so it was interesting to diagnose. The WCS RAM is the eight chips above the screw at the front middle. These are 41464 ICs, pretty common DRAM for the era.
I worked out based on the schematics that it was likely the first chip (U1J). This is due to the endianness of a 68000 CPU, it is actually the nibble on the low byte on the data bus. U1K is used for the second half of the RAM.
I desoldered this chip and popped it into my Backbit Chip Tester. I ran it on the 4464 DRAM test, which is logically identical to the 41464.
Yep, this chip is bad. Luckily, I have lots of new/old stock of them, so I soldered in a replacement. Different manufacturer, but identical specification.
Now time to test it out…
Success! It loads the Kickstart fine now and asks for the Workbench disk as expected. I then ran a battery of tests with Amiga Test Kit and they all passed. Finally, I tried loading Lemmings, which is quite a disk intensive game.
Everything worked fine, this Amiga 1000 will need a little bit of a physical clean, but the motherboard is now in much better shape.
In my previous post in this series, I managed to diagnose and repair three very…
I finally got Jops to generate a good DiagROM serial output, but the video output…
All the motherboard issues were resolved in my previous post in this series, now it…
With this Amiga 2000, I previously got it into a state where it would boot…
Last time I worked on Jops, I left myself a lot of work to do.…
I recently acquired an Amiga 2000 for £350 which was in an unknown state, but…
View Comments