SE/30 + dougg3 rom simm testing help requested

[bigmessowires]

And if you program a stock SE/30 ROM image to the SIMM, and put it in one of those machines, what happens?

 

[techknight]

Dont have a programmer. I got my simm preloaded a couple years ago.

Otherwise i would find out right quick.

 

[bigmessowires]

Has anyone ever used the Dougg3 ROM (or a stock IIsi ROM) to successfully add HD20 support to an SE/30? And really exercised it to confirm that it works? With my test setup and custom IIsi-derived ROM, SE/30 HD20 support (from Floppy Emu) kind-of-mostly works, and can boot from HD20 or mount HD20 after booting from another disk, but then it subtly corrupts the HD20 disk after opening and closing a couple of files. I quickly start getting all kinds of strange disk errors when accessing the HD20 in the Finder, and I have to replace the disk image on the SD card with my master copy. The same ROM SIMM, Floppy Emu board, and SD card work fine in HD20 mode on a IIsi or IIci, so I don't think it's a bug or problem with the Emu.

 

[olePigeon]

I don't have a FloppyEmu anymore.  I need to order one.

 

[bigmessowires]

I can bring you one when we meet for the SE/30 - send me a PM with what type you want.

 

[techknight]

my personal machine boots the doug ROM, and yes the HD20 mode does work with my HD20 drive. No, I have not tried the floppy emu though.

 

[bigmessowires]

Thanks, that's helpful to know! So this is a real HD20 drive, and you exercised it enough with running some programs and copying some files to be comfortable that it's A-OK? 

I just got a second SE/30 from olePigeon. Haven't had a chance to try it yet, but this will help me determine if the Floppy Emu HD20 disk corruption I saw with the first SE/30 was due to a hardware problem on that machine, or if it's a general issue. I'm rooting for a hardware problem.

 

[bigmessowires]

Good news, the mystery is mostly solved. The weird HD20 disk corruption problem was specific to that one SE/30. With a replacement ROM in the new SE/30, Floppy Emu HD20 worked great.

I also confirmed my earlier theory about certain RAM causing problems for any IIsi or IIsi-derived ROM in the SE/30. This behavior was consistent in both SE/30 machines. There is a particular set of 4 mismatched 4 MB RAM SIMMs that work fine in both SE/30's with the stock ROM, but won't boot in either SE/30 with the IIsi ROM or the BMOW ROM. I spent quite a while swapping RAM and ROM between the SE/30's and other systems to convince myself this was a real effect. I know Techknight said he's seen cases where the same RAM worked with the dougg3 ROM in one SE/30 but not other, but my hunch is there was some additional problem at work there.

The troublesome set of 4 x 4MB RAM was this:

USA SY 9304, 4P1004JDJ-M8
81C1000A-70, 9418 F64Z, Malaysia PJ
AAA1M300J-08, NMBS 2091
TI-80, TMS4C1024DJ, HEI 114A R

 

[techknight]

Again, RAM could be an issue, but I can bet my life that its not the entire issue. 

I have serviced alot, and I mean ALOT of SE/30 boards over the last few years, and the one thing that they had in common was I used the same EXACT RAM between them all. some boot the doug ROM, some do NOT. I never had an in-between one though. 

The other thing in common was the "darker" PCBs had issues with the doug ROM, and the lighter green models did not. That was another thing I found as well. 

I only saw a few socketed models, and if I recall, none of those booted the modified ROM SIMM. the non-socketed ones did. 

My guess is CMOS timing variance, or some sort of board change through PCB revisions. 

My honest opinion is ROM access timing or maybe the flash memory outputs cant really drive the bus on certain board, or maybe memory mux timings vary. But who knows for sure. 

 

[bigmessowires]

Yeah, I wouldn't be surprised if there are also other factors besides the RAM. I haven't noticed lighter and darker PCBs. Do you know if those correspond to the socketed vs non-socketed boards?

The first SE/30 I tested was a socketed model. It didn't work with the ROM SIMM at first, but it was OK after I swapped out the mismatched RAM. The second SE/30 was the soldered-in model, and it behaved the same as the first one with respect to the RAM.

It's an interesting mystery.

 

[techknight]

Yea I know. I would absolutely love to find out the root cause. its stumped me for awhile. 

The darker boards were almost always socketed, but I can recall one not socketed. But it varies. 

if all SE/30s are truly identical, what would stop one board, but not the other? I am assuming chip timings are really that critical, or maybe bus currents vary. I dunno. I wonder if the flash memory versus real ROM has different drive current specs? Maybe buffer it. 

 

[Macdrone]

I have a IIsi rom that works in 1 SE/30 and not another with the same ram, so age may be playing a factor.

 

[bigmessowires]

Macdrone, is there any obvious difference between the SE/30 that works and the one that doesn't? Like one is a socketed CPU and the other isn't? Or one is recapped and one isn't? Does pushing on the SIMM PCB in its socket help at all? Dougg3 thought unreliable contacts on the ROM SIMM socket might be an issue sometimes.

 

[Macdrone]

They are both soldered SE/30's, put 8mb matching 1 mb simms in, both recapped. No obvious differences.

 

[olePigeon]

Any corrosion on any of the chips?  Are any of the ROMs socketed versus soldered?  I've noticed that even between similar motherboards, sometimes the ROMs are socketed and sometimes they're soldered.

 

[Kazu]

I am a newcomer.

Yes! I have used a stock IIsi ROM to successfully add HD20 support (Floppy Emu) to my SE/30.

It can boot from HD20 or mount HD20 from another disk without problems.

The combination of IIsi ROM SIMM and Floppy Emu HD20 emulation mode is very useful for me.

I got my SE/30 from olePigeon and Floppy Emu from bigmessowires last year.

Here are other conditions.

    HDD : SCSI2SD

    RAM SIMMs : 68MB  16MBx4(BankA) + 1MBx4(BankB)

But I noticed that I could not use Daystar Turbo 040 together with IIsi ROM SIMM.

Sometimes it works but then (almost always) it corrupts the HD20.dsk and the FDD hangs-up for 1.4MB FDs (800K FDs are OK).

I don't know why and how to solve it.

I would be happy this will help to solve our root problem.

 

[bigmessowires]

Hey olePigeon, you are the only other person I'm aware of who has tried the combination of SE/30 + IIsi ROM SIMM + Daystar Turbo 040. Did you experience any problems with the floppy drive or HD20 with that combination?

 

[olePigeon]

That wasn't me, I don't have a Daystar SE/30 adapter.  I have a IIci with Daystar.  There's one person on here with a Twin Spark and an older Daystar w/o cache card, but theirs wouldn't boot.  I don't know if they had a IIsi ROM or not.

If you happen to get an SE/30 adapter, I'd be happy to lend you mine.

 

[Kazu]

Here are RAM SIMMs specs.
     OWC30PS 16MB x4
     THM91000S-10 1MB x2
     CENTURY 1MB x2
  Toal: 68MB

And I attached the picture of the motherboard.

I think it is a socketed model and with "darker" PCBs.
Correct?
 

 

[james_w]

I have a couple of SE/30s, plus doug's ROM SIMM and programmer. Happy to do some testing if it's still required - just let me know!