CRG wrote on 2023-06-18, 13:08:

It's almost there, does now boot and I've discovered the a cache error but I'm waiting delivery of parts to try and resolve that.

Latest video on it is live, give it a look if you're interested.

https://youtu.be/BVxUnFwFdWk

I was waiting for this video 🙂. Thank you, it's great - as always!

Regarding failing gas spring piston - it reminds me of furniture type, something used for shelfs. Quick Google search shows there are hundreds of varieties and I'm out of my depth 🙂.

I really don't know how to resolve problem with the cache. Maybe I would test chips in another board first. That ram tag maybe could be tested with retro chip tester pro if you have it. Really not sure. Just throwing ideas around.

CRG wrote on 2023-06-18, 13:08:

It's almost there, does now boot and I've discovered a cache error but I'm waiting delivery of parts to try and resolve that.

Latest video on it is live, give it a look if you're interested.

https://youtu.be/BVxUnFwFdWk

Nice video as usual.

Regarding the hard drives that won't spin up, keep in mind that Seagate sold some SCSI HDDs with a modified firmware that is set to always wait for the motor start command from the controller, completely ignoring the jumper settings. I know this for a fact because I have a Seagate Decathlon that behaves like this. Upon detection by the controller, it reports itself as a Seagate ST5660N but with an added "SUN" string unlike the regular consumer models. To me this indicates that these have a different firmware intended for Sun servers/workstations.

So I would try them again but connected to the SCSI controller before junking them, you never know.

Regarding the cache issues, maybe I missed it, but did you test the chips with the TL866? It may not be an exhaustive check but it can provide some clues. Also, my guess is that all the 9 identical chips are both of the cache banks + tag, with the tag being the one that is closer to the Opti chipset. Not sure what the smaller chip is as I cannot read the markings on your video, but maybe it's used for the dirty bit in a Write-Back configuration? Just speculation on my side.

I also reiterate my suggestion on modding the DS1387 chip with a coin cell if you haven't done that, the battery is definitely not rechargeable. If you notice it didn't work at first but suddenly started to work it may be on its last legs.

And definitely set the proper amount of memory (16MB) on the EISA configuration! It doesn't auto-detect everything. I have an ECS EISA board that crashes in strange ways if I don't set the correct memory values on the EISA config.

CRG wrote on 2023-06-18, 13:08:

It's almost there, does now boot and I've discovered a cache error but I'm waiting delivery of parts to try and resolve that.

Latest video on it is live, give it a look if you're interested.

https://youtu.be/BVxUnFwFdWk

Only on episode 2 and I'm already hooked! 😀 - nice video again (thx for the shoutout btw)

Not sure why none of the other online resources mention it but I think @TheMobRules is correct re the mystery smaller 10th chip being high speed RAM for dirty bits (see http://www.uktsupport.co.uk/acer/mb/j3.htm )

Pretty sure you don't need to load a config file in the ECU for this cpu card (although some longer types will if they also connect to the paired EISA slot)

Probably best, if you do solve the cache problem, to return the bios to ICL as the configuration files for the J3 board in the Altos 7000 and the TeamServer Ei do seem quite different

Thanks again guys, you've been most helpful. I've got a full set of replacement cache coming and on testing the original cache with the TL866 there is indeed one faulty (I need a full set for another project anyway). I don't think I can test that 64k x 1 chip though. Its a p4c187-15pc. I can't find this exact part anywhere but I've ordered a Cy7C187-15PC. I'm fairly sure its compatible, its certainly got the same pinout, in fact had to order 5 so if anyone needs one just drop me a DM.

I've tried a single bank with the faulty chip out and it's still not working so there must be a fault elsewhere and hopefully within said dirty ram chip. Must say I've never heard of dirty ram before so I'll have to do some reading on it.

I own the same mobo I haven't run it yet but I can see that you need that extra power supply for the board (I've highlighted in the picture). Can someone take exact pictures of what is how connected so I can make such a connector myself.