Mac Classic - will it ever run again!!

[rubber_jonnie]

Quite a few Macs are very fussy about the type of SIMM, even manufacturer and sub-model.

It's not working even with the SIMMs removed, so whilst I agree with what your saying about the SIMMs, it's the card itself that's failing.

Unless of course it's the logic board that has a problem, as this is the 2nd card that isn't working with just the soldered on RAM and no SIMMs.

There are some support ICs and buffers, and I have spare ICs from the 2 other mainboards, so I have some options, but the card is just a PCB. I suppose I could double check the passives on the card to be sure none of them have failed.

It's just a bit odd that it would work once then fail. I do hope no RAM has died, or being killed by the mobo.

[alexh]

Quite a few Macs are very fussy about the type of SIMM, even manufacturer and sub-model.

They won't be fussy per-se it will all be down to the DRAM chips used and their configuration. e.g. 1mbit or 4mbit, x8 or x16, EDO or FastPage, 50ns, 60ns or 70ns etc.

I'm sure if you wrote down the specs of each SIMM I'm sure you'd find a correlation between ones which work and ones which don't.

[rubber_jonnie]

I've tested out the passives on the expansion board, all ok. The schematics from Bomarc are a little unclear, but there is a sense pin on CASPAL, UL1, that determines if the external RAM is connected, it looks like it may be pulled down to GND when connected, and if it isn't then that is a problem, same for the SIMMs fitted signal.

I have plenty of spares, so it'd be easy enough to replace the logic ICs one by one , see if that fixes it.

[Tomswork]

Maybe a ram chip with a internal short is another possibility some of those parts die on the mb


Tom

@rubber_jonnie

[rubber_jonnie]

Maybe a ram chip with a internal short is another possibility some of those parts die on the mb


Tom

@rubber_jonnie

The machine boots fine to the ROM without the RAM card fitted, so it's not the onboard RAM.

Since two cards don't work, I think it might be something that connects the mainboard to the RAM card.

I guess we'll see.

[Tomswork]

The ram card has a meg of ram soldered to it nd 2 simm slots ?

Tom
@rubber_jonnie

[rubber_jonnie]

The ram card has a meg of ram soldered to it nd 2 simm slots ?

Tom
@rubber_jonnie

There is 1MB on the logic board, 1MB on the card and 2 SIMM slots on the expander as well, so the expansion card can add a total maximum of 3MB, so 4MB total in the machine.

[rubber_jonnie]

Ok, so today was the day I decided that I should properly troubleshoot the external RAM issue in the Classic, since the new RAM board didn't work and gave the same issue as the old one.

First I dug out the BOMARC schematics for the classic. They are the best you'll find IMHO for these machines.

Then I pulled the mainboard and RAM card out and with my multimeter in continuity mode, quickly found that all the address lines were correctly connected, which was annoying since the error I was getting suggested an addressing issue.

Anyway, I continued on, quickly finding that pin 10 on the CASPAL (UL1) that senses the RAM board is installed was ok, as was pin 11 that indicates if SIMMS are installed on the ram board as well.

I carried on working through the connections to the socket, and noted that pins 21 and 23 of the CASPAL should connect to pins 5 & 6 of the RAM adapter socket respectively. There is also a 22R resistor connected in series in both lines, and this is where things got interesting.

Pin 21 to pin 5 was fine, measuring correctly, but pin 23 initially read over 300R. I dug out one of my donor boards and tested the same connections and both came out at 22R. So I traced the pin 23 to pin 6 connection, which turns out is on the bottom of the board and passes though a resistor marked R10.

Sure enough it was still reading high, until suddenly after some probing, it was reading a more normal 22R again. Time to break out the microscope!!

This is what I found:

PXL_20230219_171545928~2.jpg (289.72 KiB) Viewed 1510 times

As you can see, there is a crack across the solder and edge of the resistor.

During desoldering, prior to a reflow you can clearly see here a chunk missing from the end of the resistor:

PXL_20230219_172206264.jpg (334.84 KiB) Viewed 1510 times

And in the flux residue, just above the R from R10, there are the missing small chunks of resistor!!!

Happily, the resistor still reads in spec, so I re-soldered both ends, and the result is:

PXL_20230219_173617303.jpg (172.27 KiB) Viewed 1510 times

Yep, we are fully running again with 4MB of RAM and the System 7 OS boots nicely from the internal HDD

I also tested configurations with and without SIMMS and it also works just fine. In addition, 2 different sets of SIMMs worked just fine.

I've left it running for a bit, make sure it survives a decent period of operation, then I'll pull the mainboard out and make sure any remaining flux residue is cleaned off before I reassemble.

I have been wondering if it's been damaged for some time, and if inserting the memory card caused flex enough to cause the damage to R10. It could even been that way from the factory, since if nobody ever used a RAM adapter, you'd never see this issue at all, and if all it ever had before was OS 6, then the RAM adapter isn't necessary.

I guess that we'll never know, but it is incredible how small components such as this can cause such mayhem, and it's good to know that following good old fashioned troubleshooting techniques can lead you to the root cause.

A happy Sunday indeed

[Tomswork]

Very nice does this mean the battle is one

Tom

[stephen_usher]

Yay! But a bit annoying.