• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Should I get a Rominator II?

croissantking

Well-known member
Nice tidy recap job there. The board does indeed look clean but I only mentioned rinsing with IPA as water/dishwasher liquids usually leave residues on the board and IPA will displace those.

Check F3 fuse, if my memory is correct, that's for internal SWIM/SCSI.

After running is the SWIM (UJ11) red hot?

A bad SCSI chip is also a possibility if there's no SCSI activity at all.
I'll do some more troubleshooting with these pointers in mind and report back.
 

croissantking

Well-known member
Nice tidy recap job there. The board does indeed look clean but I only mentioned rinsing with IPA as water/dishwasher liquids usually leave residues on the board and IPA will displace those.

Check F3 fuse, if my memory is correct, that's for internal SWIM/SCSI.

After running is the SWIM (UJ11) red hot?

A bad SCSI chip is also a possibility if there's no SCSI activity at all.
So it boots fine off a disk tools floppy with the HD disconnected. With a drive connected to the SCSI bus either internally or externally it will just hang at the grey screen, with a moveable mouse pointer visible but nothing else.

I checked the F1-3 fuses in situ and they all show continuity.

Should I suspect a bad SCSI chip at this point? Luckily, I’m competent at SMD soldering.
 

joshc

Well-known member
Yes, seems like a bad SCSI chip is likely then. I had the same problem with an SE and replacing the SCSI chip fixed it.
 

croissantking

Well-known member
Yes, seems like a bad SCSI chip is likely then. I had the same problem with an SE and replacing the SCSI chip fixed it.
Where might I be able to get one of these NCR53C80 chips? Do any Macs use this chip other than the SE/30? Also I have seen a different chip used here on a later revision of SE/30 board.
 

croissantking

Well-known member
Ok, I realised there are different manufacturers for these chips (NCR, AMD, Zilog, etc.) and have ordered a used ‘AM53C80AJC’ off eBay

I also scored a used IIsi in good condition on eBay for just north of £100, I am very excited about that.
 

croissantking

Well-known member
I checked all the pins on the SCSI chip and found a broken trace going to the Glue chip. I’ve fitted a bodge wire and my beautiful machine is working once again.
I can see from photos I took when I first recapped that this trace had been damaged by the previous owner who really did a number on the board around the three caps next to the power connector. What’s strange is that it only completely stopped recognising SCSI devices after I put my board in the dishwasher. Maybe the trace was partially connected before (but unreliably) and washing it totally severed the connection. Or maybe my earlier instability issues are nothing to do with this and are still present. I’ll have to use the machine a bit and see how it goes.
 

croissantking

Well-known member
Nope still crashy. I’m going to replace the ROM socket next. I have a feeling the machine is having trouble accessing its ROM SIMM.
 

Chopsticks

Well-known member
before you replace the rom socket, have you actually cleaned up both there Rom and ram sockets as well as the rom/ram Simm contacts using ipa? you could have intermittent connections there. also replacing the rom socket might not actually be easy as getting those sockets (64pin SIMM I think??) are very difficult these days.

Id strongly recommend that to anyone who washes their boards in a dishwasher to make sure they also thoroughly clean them with ipa afterwards. there's several different chemicals that are often used in dishwashing liquid/powder/cubes etc that can actually conduct just enough electricity to either cause resistance or shorts based issues.

as for the hdd issue, I likely could be a scsi chip issue but it also could be an issue still with the power supply. its possible that when ou turn the se/30 on with the hdd connected that the initial power draw on the 5v or 12v lines is getting loaded down to much causing issues with logic states be3ing invalid. you should try using a separate psi just for the hdd to rule that out. those se/30 psi's are known to be a little on the under powered side and as those scsi hdd's are getting very old now its common that they are for a lack of a better word a little 'stiff' due to baring (please note im not referring to stuck heads on those quantum drives, that's a different issue)

probably a good idea to measure the +5 and +12 voltage rails on the db15 floppy port during power up to see if the voltage remains steady..

someone also I think mentioned issues with using ram with less then 8 chips, well that's bollocks, I have 3 se/30 machines here and they all use 3 chip Simms with only 2 chips populated. the 3rd chip or the 9th chip on 8chip sims is just a parity bit chip that Macs don't even use.
in regards to say a 2 or 8 chip (or 3 and 9 chip with parity bit) the reason they are like that is because ram chip have been made using different amounts of bits stored per chip. so a 8/9 chip Simm stored 1 bit per address on each chip, where as a 2/3 chip Simm will use a 4bit per accessed address. in fact you could use a single chip solution technically though Im not sure if they made ram chip in that config back in those days.

also a trick I use for my custom rom sims I made is to actually apply 2 or 3 layers of caption tape on the back on the sim near to contacts just to provide some extra pressure. in fact I had to do that to one of my se/30 ram banks once too as I had issue with random crashing at times.

Im not sure if any off that is of any help but if not perhaps its given you some other ideas to investigate.

lastly yes there is a se/30 logic board orb reproduction available. but unless you can arrange a group order on here or somewhere else just know that is not exactly cheap if you only need 1 board and have to order for a PCB manufacturer. I plan to order some boards at some stage but I have no idea when tbh


edit: forgot to mention Force32 won't work on se/30 with stock ROM. stock rom is 32bit dirty so forces32 won't work unless you use a iisi/iifx rom, a romulator or other similar product that has 32biut clean rom on it.
 

cheesestraws

Well-known member
edit: forgot to mention Force32 won't work on se/30 with stock ROM. stock rom is 32bit dirty so forces32 won't work unless you use a iisi/iifx rom, a romulator or other similar product that has 32biut clean rom on it.

You can also use MODE32. Force32 will work correctly with that. In fact, that's what I do.
 

croissantking

Well-known member
@Chopsticks Thank you for your thorough input.

I think I need to get creative to find the issue.

My ROM socket has a broken clip and I have to force the SIMM to sit as far forward as possible with rubber bands. I just got a IIsi which has a ROM SIMM socket with metal clips that I was thinking to donate to the SE/30, since this will never be used due to it having onboard ROM chips.

I may get the Rominator II as I’d like to try a different ROM SIMM, if changing the socket doesn’t help. I’ve heard these can occasionally go bad, like RAM.

I’m not sure if dishwashing the board was a good idea anymore but note that I didn’t use any detergent.

My SCSI issues are solved but I will check the voltages on boot up as you suggest. The good thing about my IIsi aquisition is that I can now test my HDs and RAM outside the problematic SE/30.

Is it normal in MacsBug btw for there to appear a line that says ‘the ROM symbol file could not be read’? I attach an image of one of my routine crashes.BCB54FFE-E21D-4346-AC3F-5237B4A4794B.jpeg
 
Last edited:

Chopsticks

Well-known member
You can also use MODE32. Force32 will work correctly with that. In fact, that's what I do.
oh ok, yeah thinking about it I guess that would work if force32 loads aftefter mode32. I haven't used mode32 in quite a few years and I don't recall any stability or software issues from it myself, im assuming you haven't either. So using mode32 could save him some money rather then buying a romulator II.

@Chopsticks Thank you for your thorough input.

I think I need to get creative to find the issue.

My ROM socket has a broken clip and I have to force the SIMM to sit as far forward as possible with rubber bands. I just got a IIsi which has a ROM SIMM socket with metal clips that I was thinking to donate to the SE/30, since this will never be used due to it having onboard ROM chips.

I may get the Rominator II as I’d like to try a different ROM SIMM, if changing the socket doesn’t help. I’ve heard these can occasionally go bad, like RAM.

I’m not sure if dishwashing the board was a good idea anymore but note that I didn’t use any detergent.

My SCSI issues are solved but I will check the voltages on boot up as you suggest. The good thing about my IIsi aquisition is that I can now test my HDs and RAM outside the problematic SE/30.

Is it normal in MacsBug btw for there to appear a line that says ‘the ROM symbol file could not be read’? I attach an image of one of my routine crashes.
well since you already have a socket then its certainly a good way to rule out if its the issue or not. not sure if you have access to a 3d printer or not but if you do then you can actually print out two clips to that are meant to hold the Simm in if the original plastic clips of on the socket have broken. from what Ive read on the forums they seem to work pretty well.

as for the romulator, I have my own 8mb rom Simms I made ages ago with the modified IIsi rom that supports the floppy hdd option and the built in rom disk. now realistically being that its a se/30 with scsi I never use the floppy hdd option (I have a floppyemu device that supports that format) as its just painfully slow, its nice to have the option but I kinda understand why apple never included it in the stock rom. As for the RomDisk support, I spent quite awhile tweaking what files/utilites I put on it, in fact I must have reflashed the rom a dozen times only to rarely use the feature either. it has been handy a couple times to speed up recovery after a bad crash though. as cheesestraws mentioned he uses mode32 with force32 so that could certainly save you money...

As for dishwashing the pcb, well I can't say I've ever put a board in one to wash it, but many people have and haven't had any issues so just because I don't do it doesn't mean im going to tell someone else not do it. I just wouldn't use tap/drinking water to clean a pcb due to the minerals in the water. regardless there are a few different views and thoughts regarding cleaning/washing and dealing with corrosion, best advice I got on that is use the methods you feel comfortable using.

ill be curious to hear how you go testing the ram an hd on the IIsi
 

croissantking

Well-known member
As ill be curious to hear how you go testing the ram an hd on the IIsi.
Me too. I have to recap it first as there’s juice all over the MLB and PSU like every other IIsi out there.

I guess I could actually test out the ROM SIMM from the SE/30 in it, I guess nobody else has done this because why would you; but if a IIsi ROM works in an SE/30 it would surely work the other way around.

Interesting to hear you don’t really find the RII all that useful. To my mind, booting off the ROM disk seems a cool feature that could also help in isolating SCSI issues.

The IIsi also, curiously, has come with 4x 8MB SIMMs. I believe these would not work in an SE/30?
 

Chopsticks

Well-known member
Me too. I have to recap it first as there’s juice all over the MLB and PSU like every other IIsi out there.

I guess I could actually test out the ROM SIMM from the SE/30 in it, I guess nobody else has done this because why would you; but if a IIsi ROM works in an SE/30 it would surely work the other way around.

Interesting to hear you don’t really find the RII all that useful. To my mind, booting off the ROM disk seems a cool feature that could also help in isolating SCSI issues.

The IIsi also, curiously, has come with 4x 8MB SIMMs. I believe these would not work in an SE/30?
I have no idea if the se/30 rom will work in the IIsi or not, I don't have a iisi myself but there's likely a pin headed jumper you need to swap to change from the built in rom to using the Simms.

the romulator is cool and quite hardly depending on the situation. being that I have a 3 se/30's and amount my collection I guess im lucky enough to have enough parts to narrow down issues and fault find quicker. but definitely the built in rom disk has halped me couple times to dig some issues quicker. I mean it boots pretty darn fast using a rom disk that's for sure

I wasn't aware that they made 8mb 30pin sims, I mean technically I think there's 12 address lines available on a 30 pin sim so in theory I think that allows for unto 16mb sims but im not sure they se/30 can address more then 128mb total using 4mb sims aways...
 

croissantking

Well-known member
I have no idea if the se/30 rom will work in the IIsi or not, I don't have a iisi myself but there's likely a pin headed jumper you need to swap to change from the built in rom to using the Simms.

the romulator is cool and quite hardly depending on the situation. being that I have a 3 se/30's and amount my collection I guess im lucky enough to have enough parts to narrow down issues and fault find quicker. but definitely the built in rom disk has halped me couple times to dig some issues quicker. I mean it boots pretty darn fast using a rom disk that's for sure

I wasn't aware that they made 8mb 30pin sims, I mean technically I think there's 12 address lines available on a 30 pin sim so in theory I think that allows for unto 16mb sims but im not sure they se/30 can address more then 128mb total using 4mb sims aways...
Here’s a pic of the SIMMs. They’re basically 16MB SIMMS with half the chips missing.98BF08D5-E2A2-4123-A137-6B6011771F79.png
 

cheesestraws

Well-known member
oh ok, yeah thinking about it I guess that would work if force32 loads aftefter mode32. I haven't used mode32 in quite a few years and I don't recall any stability or software issues from it myself, im assuming you haven't either. So using mode32 could save him some money rather then buying a romulator II.

MODE32 does some deep magic, parts of it load very early indeed.
 

Chopsticks

Well-known member
MODE32 does some deep magic, parts of it load very early indeed.
doesn't mode32 just patch some 24bit rom code that's still left over in the dirty roms, place changed code in memory and change the pointers to look there for the code rather then the rom space address?
that's just my assumption really? like the actually implementation is probably much more complicated due to needing to load very early in boot, though couldn't they just hijack one of the early PTCH resources to get this happening early on in boot?
 

cheesestraws

Well-known member
doesn't mode32 just patch some 24bit rom code that's still left over in the dirty roms, place changed code in memory and change the pointers to look there for the code rather then the rom space address?

pretty much, and adds some other code to do memory tests as well. But that is pretty deep magic by extension standards and requires fairly intimate knowledge of the way the system operates. I meant more 'it's not just a normal INIT which loads at INIT time', so Force32 will basically always load after the bits of it that matters, even if it runs before the INIT resources in Mode32.
 

croissantking

Well-known member
pretty much, and adds some other code to do memory tests as well. But that is pretty deep magic by extension standards and requires fairly intimate knowledge of the way the system operates. I meant more 'it's not just a normal INIT which loads at INIT time', so Force32 will basically always load after the bits of it that matters, even if it runs before the INIT resources in Mode32.
How is an INIT able to run before INIT time?
 

Chopsticks

Well-known member
pretty much, and adds some other code to do memory tests as well. But that is pretty deep magic by extension standards and requires fairly intimate knowledge of the way the system operates. I meant more 'it's not just a normal INIT which loads at INIT time', so Force32 will basically always load after the bits of it that matters, even if it runs before the INIT resources in Mode32.
absolutely, im sure even with all the resources available today that to be able to patch the system boot process to do all that stuff would require some serious skill.

How is an INIT able to run before INIT time?
from memory, and im not a coder, ive just read about it but when the system file boots there are a bunch of PTCH resources that get loaded up to either add/replace specific macOS functions etc. this happens before the rest of the OS starts up btw. when system 7 came out there were certain functions on the core OS that Apple wanted to change however because of how the macOS works the only way some things could be added was by patching a pre-existing resource that's part of the boot up process to basically added the function they wanted that gets loaded before the rest of the system starts loading, Apple did this a lot over the following years and while im sure many other companies did the same it certainly wasn't a common knowledge type of skill set for a Mac programmer to have at that time.
So Mode32 as far as I understand it is a INIT but mostly its just a file for holding the actual 'data' that gets used to patch the system file with code to replace 24bit ROM routines with 32bit versions.
they possibly could have actually just patched the system file using a patching app and not used an INIT at all, however im just speculating here and might be wrong. there might be a reason im not aware of as to why they used an INIT.
 
Top