Should I install 64MB or 65MB in my SE/30?

[phipli]

If I use my stock ROM, the system shows 2MB of RAM instead of 65MB - it will boot the MacTest ISO - but it won't run because not enough free memory.

Do I remember you saying you put the 64MB in Bank B?

I'd start by putting the 16MB SIMMs in Bank A, which is what the Apple instructions say. This will especially help when you're using the stock ROM.

While you have them out, clean the contacts on both sets of SIMMs using IPA (rubbing alcohol).

I can't boot from the ISO - it hangs at "Welcome to Macintosh". The README on the ISO says it includes a System 7.5.3 bootable installation on the CD. I think because of my replacement ROM, I'd need to update the "gusd" in the boot image, which I had to do for my hard drive image. Except the ISO is read only since its a CD device Instructions I used for my HD image are these:

You don't really need to boot from the ISO to run the tests, I never do. I wouldn't worry about patching that OS just to run the RAM test.

 

[djc6]

Do I remember you saying you put the 64MB in Bank B?

I'd start by putting the 16MB SIMMs in Bank A, which is what the Apple instructions say. This will especially help when you're using the stock ROM.

Are there instructions somewhere? The "Apple Memory Guide" I found doesn't mention 16MB SIMMs at all - page 4 is about the SE/30:

http://www.applerepairmanuals.com/the_manuals_are_in_here/Apple_Memory_Guide.pdf

I mostly don't want to move things around too much out of fear of breaking the clips Mine are all plastic.

 

[phipli]

Are there instructions somewhere? The "Apple Memory Guide" I found doesn't mention 16MB SIMMs at all - page 4 is about the SE/30:

The rule is generally "put the big SIMMs in Bank A" on the SE/30, rather than it specifically talking about 16MB SIMMs. 16MB SIMMs weren't readily available when the SE/30 was released and due to a court case from the Mac II/IIx era, Apple only listed configurations they'd tested.

I mostly don't want to move things around too much out of fear of breaking the clips Mine are all plastic.

Yeah, those clips are a nightmare, but I really suggest putting the SIMMs in the conventional configuration - it isn't that it never works the other way in various machines, it is just that it is one more thing to take out of the "why might it not be working properly" pot. Truthfully, cleaning the SIMMs is more likely to fix the issue, but as it stands, you have the SIMMs in backwards vs the normal way of doing it for an SE/30.

Also, I don't know why that document has the SIMMs the other way around for 17MB - if you look at the IIcx page (the IIcx is basically the same computer in a different form factor) the arrangement is always big SIMMs in Bank A. I suspect the 17MB row on the SE/30 might be a typo.

 

[phipli]

I mostly don't want to move things around too much out of fear of breaking the clips Mine are all plastic.

BTW, if you use a hairdrier to warm them up, they are a little less brittle, but you still need to be super careful. You shouldn't need to ever touch it again after cleaning them and swapping them over.

If you have some contact cleaner, spray it in the slots while you're at it. Deoxit is expensive but works well.

 

[djc6]

The rule is generally "put the big SIMMs in Bank A" on the SE/30, rather than it specifically talking about 16MB SIMMs.

OK, I populated only Bank A with the 16MB SIMMs - nothing in Bank B - and the test is running now!! Its still going but its no longer hung. I am running the system outside the case - I need to see if these SIMMs will physically fit while in Bank A with the logic board back in the chassis they are quite a bit taller. Bank B has a lot more headroom.

Re: Deoxit - I sprayed some deoxit and used a new soft bristled toothbrush to clean the slots, without having read your recommendation yet

I am curious if I should bother with the Bank B 4x256KB. I'm going to do some experiments later what I can boot with this configuration - stock ROM, system 6.0.8, etc.. before I decide to install that 1MB of RAM.

Getting the RAM out was pretty easy to do delicately, it was re-installing the RAM that required a little more force than I wanted to exert. I used a plastic spudger to bend the tabs versus my fingers thinking I could apply finer amounts of force. I may try the hair dryer!

Thanks!

 

[phipli]

I am curious if I should bother with the Bank B 4x256KB. I'm going to do some experiments later what I can boot with this configuration - stock ROM, system 6.0.8, etc.. before I decide.

Using 16MB SIMMs in Bank A and nothing in Bank B is commonly reported to cause no-boot issues with the stock ROM. There is no reason not to put small SIMMs in Bank B, normally, and it improves compatibility. But you want to clean the contacts on those SIMMs. Given you're having more luck without them installed, they may be the issue. Do you have any 1MB SIMMs? They tend to be cheap or free and you could put them in instead.

But, if it currently works, and you're never planning to use the stock ROM, I'd consider just leaving it as it is with the 64MB you have in there.

 

[djc6]

Now the gremlin I am trying to track down is that MacTest Pro short and extended tests PASS in 64MB or 65MB configurations (with 1MB in bank B) but the "Long Test" hangs immediately in either configuration.

"Long Test" also hangs immediately with 7.5.3 rev 2 and Stock ROM and MODE32 installed.

Curiously, my Garrett’s Workshop GW4402B ROM SIMM has a dip switch setting for "stock IIsi" ROM and it won't boot with the 1MB installed in Bank B. It will boot with only 64MB in Bank A, nothing in Bank B and IIsi stock ROM selected. It will boot with DIP switches set so it boots from IIsi ROM with RAM test disabled which is default position it ships with.

 

[djc6]

Now the gremlin I am trying to track down is that MacTest Pro short and extended tests PASS in 64MB or 65MB configurations (with 1MB in bank B) but the "Long Test" hangs immediately in either configuration.

I no longer think its hung doing the "Long" memory test. The mouse cursor changes to a beachball and I cannot move it. However, the Status message "Now testing RAM at address $##### actually does update at intervals just under every 2 minutes. Only the first two digits increment so it seems to be checking blocks of memory. So now I think I am just impatient. I am thinking the "Long" test could take a day or more to complete? The progress bar actually shows a sliver of progress after an hour or so, but I still cannot move the mouse.

The VRAM test is pretty cool - psychedelic almost in the patterns it draws on the screen