peters Milan 040 issues

[peters]

It all went wrong trying to get slot 1 to work. I got a SIMM recognised in it but then slots 2 and 3 stopped working.

I've managed to get 128 + 0 + 64 +128 = 320 to work again although the clip on slot 3 cracked and I've had to super glue it together so I'm not sure how reliable it's going to be.

The behaviour is a bit odd. A few times it dropped back to 192. Sometimes it would start making long beeps on power up suggesting no RAM detected or more serious issue.

At the moment 320 is the best I can get to boot. I will leave it like that for a few weeks to see how it goes.

IMG_20251230_162611857_HDR.jpg (339.28 KiB) Viewed 1162 times

[peters]

@exxos thanks for splitting the thread. My posts got a bit out of hand.

It's now booting reliably with 320mb and I'm switching my attention to the second HDD.

I have a sata drive with an adapter plugged into the back of it and an IDE cable to the motherboard.

It works sometimes. Last night it worked for about an hour and then the machine locked up.

You suggested a dry joint on the motherboard socket and I'm beginning to think you are right. Could it be that it fails as it warms up ?

[dml]

It does sound like some kind of resistive/intermittent connection somewhere but I think the board really needs a full inspection with a microscope and maybe resistance checks from the socket pins to the trace destinations on at least one of the main ICs.

It would also be helpful to use a decent memory test - just detecting the ram in a given socket doesn't mean it is working correctly - that could be your freezes. Proving some part of the ram is good on multiple passes would at least let you know the machine works properly with a single slot and a specific RAM card and then maybe narrow down which slot and/or ramsticks are causing problems. If you have more than one problem going on it will be a confusing picture and hard to understand without a reliable starting point to compare with.

[peters]

I don't have the tools or skill to inspect the board. I would probably miss something that is actually a problem.

I ran M Schwingen's memtest app a few times so I think the RAM is ok apart from the poor nature of the sockets. They are just flimsy. That design wasn't great with the 1mb Simms and isn't suited to the much larger EDO Simms.

I think slot 1 is broken in some way. It did work briefly yesterday.

What is obvious is that the board is held in by clips along the back and screws in the middle but there are no supports on the corner where the RAM sits. There are no holes in the case for mounting screws.

I also think the second IDE port has problems.

[peters]

My Milan ram has been gradually getting worse. Now even slot 0 is rarely booting cleanly.

I suspect dry joints so I'm going to carefully disassemble everything and then try to reflow the solder on the simm sockets.

Fingers crossed I don't break anything.

[rubber_jonnie]

My Milan ram has been gradually getting worse. Now even slot 0 is rarely booting cleanly.

I suspect dry joints so I'm going to carefully disassemble everything and then try to reflow the solder on the simm sockets.

Fingers crossed I don't break anything.

If you're dead set on doing this my advice would be first clean with IPA and add flux before reflowing, the gel kind as it will help everything flow. If you don't have flux I'd suggest you get some, but NOT plumbers flux as that is super corrosive.

The next bit of advice is to be patient and take your time. Rushing usually end up with damage, I've fixed enough of my own and other peoples rush jobs to know that patience is a valuable commodity.

Lastly, before you break out the soldering iron, could this simply be bad RAM/SIMMs? Also I can't quite tell what that component is between the pins, but I had a Mac Classic that was having a ton of RAM issues when the RAM expansion was plugged in and a resistor, looked similar to that component, was cracked. I replaced it and the problem was gone. Check stuff like that too before soldering.

[peters]

Thanks.

I have about 8 Simms and they all have problems so I'm fairly sure it is the sockets or the soldering.

I removed all but one simm and it has been ok for a few months but now it takes 5-6 attempts at wiggling the simm before it boots. Then it works fine for a couple of hours. The next day it's bad again.


Using a powerful hand magnifyer the soldering looks ok but a bit oxidised with age. I'm no expert.

I'm not confident I can reflow them successfully. I've just given them a good brush with IPA and letting it dry.

[rubber_jonnie]

I'm not confident I can reflow them successfully. I've just given them a good brush with IPA and letting it dry.

Flux is very important, don't miss it out.

[peters]

I decided not to solder.

Slot 0 has a weak clip on one side so about 15 years ago I had to apply a small blob of hot glue. It's been good all that time.

I managed to get the simm out without damaging anything and gave that slot a better clean.

It has now booted several times without problems.

I tried a second simm in slots 2 and 3 but they didn't detect the RAM.

Ps.
I managed to get 128+64 to boot up. This was how I ran it for about 20 years when I could only get 1 of the 128mb to work.