You may remember that the A4000 known as “Jools” went to its new owner for an event in a semi-working state. Unfortunately, due to an injury, he wasn’t able to bring it back to me to finish until now. There was a bit more work to do than I thought, but here we go!
Jools was left in a state where we knew the following work was needed:
I first started with the joystick ports. These old ones are not going to be used again, I therefore find it easier to snip the mounting clips on the top side to make them easier to desolder on the bottom. The new sockets are looking a bit shinier.
Next comes the SIMM sockets. I took a photo of the bottom so that I could remember what patches I have made in the past.
I then desoldered the sockets. There are 360 solder joints to do this. It took a while. I also quickly removed the 7407 from in front of them.
I cleaned up the board here and soldered the new sockets, reapplying the patches as I went. The bottom patch I made on the top side under the SIMMs.
Once the sockets were in, I tested them using a Hafnium board. This was difficult to photograph, but D0 was intermittent on the Fast RAM sockets.
What I found was the trace from the buffer chip had cracked near the chip, a touch of solder fixed it. It looks messy with flux right now, but it will clean up nicely. It is pin 2 on this SMD IC.
The machine booted into Amiga Test Kit, but the mouse wouldn’t move at all. I then used my custom version of Amiga Test Kit which uses a mouse in port 2, that also didn’t work.
The mouse and joystick inputs use a 74HCT166 to push the data serially out to the chipset. After checking the connections were still OK, I removed this chip and tested it.
Pin 13 is the serial output, that explains a lot. These chips seem to be more vulnerable to battery damage than most in that area for some reason. There are two next to each other. One controls the ports, the other is for a chipset upgrade that never existed. I replaced both of these, and noticed a damaged trace whilst I was at it, which I patched. This wasn’t so important, but it is nice to have the chip pull-up connected.
This time, I could control Amiga Test Kit, and both ports worked correctly. I popped in 16MB of Fast RAM along with the maximum 2MB Chipset RAM, and it all worked!
Success! I ran some other tests, such as RTC, to make sure everything else was still working. I did accidentally disconnect a patch wire for the RTC when swapping the 74HCT166 ICs, which caused that to fail. After re-connecting that patch, all was good. The LED pins were now working correctly too.
The motherboard was cleaned and dried, 100% working and ready to go back into the case. I won’t show the reassembly at this point because the final look of this machine will be amazing when revealed, but it is not ready yet.
Anyone who has worked on Amiga 4000s before will tell you that removing the faceplate can be a pain. The plate is held on using lots of two-prong clips. Some of these are in an awkward position, and it is difficult to pop them out using fingers.
Using pliers is possible, but there is a risk of damaging the clips. Instead, I designed a tool which you push onto the clip, and it pushes the two prongs together whilst simultaneously pushing it through the hole.
This makes it really simple to pop all the clips, and you can have the entire face safely removed in seconds.
I’ve made this tool available on the Retro Supplies store. You can also find Hafnium there, which is the diagnostics tool for the SIMM sockets.
I haven’t forgotten about the hard drives that came with these machines. They are too old to support LBA, which means that they can’t be used with USB IDE adapters. Instead, I need to use a period correct machine that supports CHS hard drives. I tried to archive them with the 486 PC that I just built. Unfortunately, the first one I tried had a minor explosion.
Luckily, this isn’t fatal, but it does mean I’ll need to do some work on the drives first. This was a tantalum capacitor on the 12v rail of the drive. They were often under-specified at the time, and when they fail, they fail short circuit. Some lovely magic smoke was released at this time.
I’ll be working on these at some point in the near future.
Hot on the heels of repairing the Cirrus Logic VLB card, I acquired a dead…
It seems that every time I document how to do Xilinx JTAG programming from a…
Fellow vintage computer collector, Gammitin, challenged me to fix four VLB graphics cards, as well…
You may remember my recent 486 PC build, I enjoyed doing it, but I had…
One of the projects I wanted to do with my new 486 PC was to…