More SE/30 Woes

[SlateBlue]

First, I want to thank techknight for all of his help with this board. He was able to bring it back from terminal simasimac. I have an odd issue, and I'm not sure where to start. I've check continuity between several chips, but I quickly get lost in trying to understand the Bomarc schematics.

The issue: With 4 MB of RAM installed (1 MB SIMM times 4), I get boot chimes and successful start up. Mac runs just fine, no problems. When I replace the 1 MB SIMMS with 4 MB SIMMS for a total of 16 MB of RAM, I get terminal simasimac symptoms again. This happens also when all 8 slots are populated with 4 MB of RAM (32 MB total). Why?!?!

The board has been recapped, and some traces were previously repaired by techknight. I'm not sure where to start troubleshooting this issue, nor why 4 MB works fine, but not 16 or 32. Thanks in advance for any helpful advice.

New RAM purchased from WeLoveMacs: http://www.welovemacs.com/apple-memory-mac-ii--mac-se--classic-mac-iicx-4mb-apple-30-pin-simm--p-n-apl30p4-.html

 

[Unknown_K]

Check the new RAM on another machine and dig up some other 4 x 1MB SIMMs and see if your machine boots with 8MB (so you know if there is a problem with the other 4 RAM slots).

4MB working means that 1 bank is OK and the RAM is ok. When using both banks and new RAM you are introducing too many variables into the mix.

 

[SlateBlue]

Perhaps I wasn't clear in the first post:

Bank 1: 4 MB works (no simasimac)

Bank 1: 16 MB does not work (simasimac)

Bank 1 + Bank 2: 32 MB does not work (simasimac)

Should I focus on Bank 1? Why would it work with 4 MB, but not 16 MB?

 

[just.in.time]

It sounds like your RAM may be defective. At least that's what I think he means.

 

[SlateBlue]

Defective or incompatible? It's brand new.

 

[apm]

It might be a problem with one of the RAM address lines (the upper ones that matter for 4MB SIMMs, but not 1MB SIMMs). If so, the most likely cause would be a rotted trace from the cap leakage. Perhaps check around the SIMM socket near C1 to see if anything is opened up there. In the schematics, you'll find that many of the address pins are tied together on each SIMM in the bank, so you could check continuity between them.

 

[techknight]

Like I mentioned before, its worth investigating the 74F253 Mux ICs. I have had to change them out before in the past. 

 

[SlateBlue]

techknight: How does one determine if a mux is bad?

 

[SlateBlue]

As I understand, UA8 through UD8 are the muxes you speak of. UE8 is a shift register, and UF8 and UG8 are counters. Is this correct? If so, then all I need to replace are UA8 through UD8.

 

[nvdeynde]

I would be looking for a faulty Ram module first. I've had exactly the same problem with my SE/30 a few weeks ago. Do you have another Mac where you can test the 4MB modules  ?
It's likely that 1 or more of them are faulty. Are they 8/9-chip modules or 2/3-chip ?

The SE/30 apparently doesn't give memory errors, if you have a bad ram module it's Simasimac with chime of dead.

 

[techknight]

The muxes I speak of are located right next to the RAM sockets and are 74F253 labled ICs. 

as far as testing, thats a little more difficult to do without a multi-channel scope to verify they are able to switch and follow the truth table. I just change them. There are 6. 

Now, again, I would follow everyone elses advice and make damn sure it isnt bad or incompatible RAM. 

 

[SlateBlue]

I would be looking for a faulty Ram module first. I've had exactly the same problem with my SE/30 a few weeks ago. Do you have another Mac where you can test the 4MB modules ?

It's likely that 1 or more of them are faulty. Are they 8/9-chip modules or 2/3-chip ?

The SE/30 apparently doesn't give memory errors, if you have a bad ram module it's Simasimac with chime of dead.

These are newly manufactured 8-chip SIMMs. Will an SE accept 4 MB chips? I'm thinking that model maxed at 4 MB. Otherwise, I don't think I have any other computer to test these with.
@techknight - UA8 through UD8 are labeled: TIF035 BS F253.

UE8 is labeled LS166 XXJK040.

UG8 & UG8 are labeled: TIF048 BS LS393.

It doesn't make sense to me why there are three different part numbers. Based on the part number you provided, my mind says replace UA8 through UD8, and UE8 through UG8 are different chips. I'm leaning on you here for understanding.

 

[apm]

UA8-UD8 are the 74F253 chips. Some of the numbers will be manufacturing codes. "F253" is the important part.

UE8-UG8 are all involved with video. They shouldn't be the cause of your problem. UA8-UD8 are also video-related, connecting from the bus to VRAM, so I'm not sure I'd pull those off the board before checking out other options, especially incompatible RAM. You may have to find a different machine to test with, since the SE won't take 4MB chips.

Personally I'd also start by checking the connections between the SIMM slots and RP4, RP5 and RP6, using the schematics for reference. Cap goo could have rotted one of those traces. And anyway, testing continuity doesn't carry any risks other than some wasted time.

 

[SlateBlue]

UA8-UD8 are the 74F253 chips. Some of the numbers will be manufacturing codes. "F253" is the important part.

UE8-UG8 are all involved with video. They shouldn't be the cause of your problem. UA8-UD8 are also video-related, connecting from the bus to VRAM, so I'm not sure I'd pull those off the board before checking out other options, especially incompatible RAM. You may have to find a different machine to test with, since the SE won't take 4MB chips.

Personally I'd also start by checking the connections between the SIMM slots and RP4, RP5 and RP6, using the schematics for reference. Cap goo could have rotted one of those traces. And anyway, testing continuity doesn't carry any risks other than some wasted time.

Thank you!

 

[SlateBlue]

Using the Bomarc schems with my limited electronics knowledge, I did some probing around with the multimeter and found no broken traces. But here is the interesting thing:

I had previously tried the 1 MB SIMMs (8 chip) from an SE to upgrade the SE/30 to 8 MB. I kept the existing 1 MB SIMMs (2 chip) in Bank A and added the additional memory to bank B. This configuration gave simasimac symptoms just like to 16 MB and 32 MB configurations.

For grins, I resoldered the existing muxes onto the SE/30 board as some of the connections looked a bit corroded (not badly). Also, the battery from the logic board was removed prior to completing this task.

I then reinstalled the battery and began adding all the 1 MB sticks to the board, but this time I put the 8-chip SIMMs in Bank A and the 2-chip SIMMs in Bank B. Upon reassembly, I flipped the power switch, heard chimes, and watched the SE/30 successfully boot to the desktop.

Tonight, I'm going to remove the board again and reinstall the 4 MB SIMMs and see if the simasimac symptoms return. If so, I will suspect bad RAM. Otherwise, I'm not sure what else to check out on the board.

Thanks for all your responses.

 

[SlateBlue]

UA8-UD8 are the 74F253 chips. Some of the numbers will be manufacturing codes. "F253" is the important part.

UE8-UG8 are all involved with video. They shouldn't be the cause of your problem. UA8-UD8 are also video-related, connecting from the bus to VRAM, so I'm not sure I'd pull those off the board before checking out other options, especially incompatible RAM. You may have to find a different machine to test with, since the SE won't take 4MB chips.

Personally I'd also start by checking the connections between the SIMM slots and RP4, RP5 and RP6, using the schematics for reference. Cap goo could have rotted one of those traces. And anyway, testing continuity doesn't carry any risks other than some wasted time.

I forgot to specifically state that the connections between RP4, 5, 6 and the SIMM slots checked out fine.

 

[techknight]

Heating the pins on the 74F253 ICs will also heat the die. if there is an intermittent connection internal to the IC as well as a solder joint, its really hard to tell. 

Like I said, I have found more than once that one of those ICs were bad. 

Hopefully in your case its just the RAM, otherwise youll have to rework those ICs with new ones. 

 

[SlateBlue]

Heating the pins on the 74F253 ICs will also heat the die. if there is an intermittent connection internal to the IC as well as a solder joint, its really hard to tell.

Like I said, I have found more than once that one of those ICs were bad.

Hopefully in your case its just the RAM, otherwise youll have to rework those ICs with new ones.

I do have some 74F253's on order, they just haven't arrived yet. I'll report back with the results of swapping in the 4 MB SIMMs.

 

[SlateBlue]

Good news: The damn thing works with 32 MB of RAM. I dunno what fixed it... The only thing I can think of is that I cleaned the ROM SIMM contacts and reseated it. Beyond that, I changed nothing else. We'll see how long this lasts...