Mac IIcx no chime after board re-cap and startup circuit repair

[rodders]

The only ROM images I can find are complete, presumably for use in emulators. Do you know of a source of split ROM files i.e. for each of the chips, or some way to de-interleave them without having to resort to coding. Someone must have done this before.
BTW are the ROM binaries simply interleaved (first byte for ROM1, second for ROM2 etc) or is some byte swapping necessary since these MACs are big endian? Thinking about it I suppose that just affects which way round you insert the ROM chips.

 

[phipli]

The only ROM images I can find are complete, presumably for use in emulators. Do you know of a source of split ROM files i.e. for each of the chips, or some way to de-interleave them without having to resort to coding. Someone must have done this before.
BTW are the ROM binaries simply interleaved (first byte for ROM1, second for ROM2 etc) or is some byte swapping necessary since these MACs are big endian? Thinking about it I suppose that just affects which way round you insert the ROM chips.

If you have the TL866+ style (or its replacement that I forget the name of) programmer, the software it comes with supports writing ROMs like this.

I found out after writing a python script to do it for a RiscPC, and then when I went to write the ROM files I'd made, I saw the option built in

So I think you want the bottom four options in the menu shown in the screen shot.

You could also use this to load each of the four bytes and then use the verify function to check the original ROMs, instead of writing new ones.

BTW are the ROM binaries simply interleaved (first byte for ROM1, second for ROM2 etc) or is some byte swapping necessary since these MACs are big endian? Thinking about it I suppose that just affects which way round you insert the ROM chips.

I don't think there is byte swapping and the endian doesn't matter. It doesn't change which way you insert the ROMs, it doesn't really make a difference. Just defines what you do with the bits when you use them.

 

[rodders]

If you have the TL866+ style (or its replacement that I forget the name of) programmer, the software it comes with supports writing ROMs like this.

I found out after writing a python script to do it for a RiscPC, and then when I went to write the ROM files I'd made, I saw the option built in

Doh! I hadn't thought to look at the TL866+. It sounds like you have trodden a similar path.

the endian doesn't matter.

I was just wondering which chip went where but I see they are labelled LL, ML MH and HH which is presumably 4th, 3rd, 2nd and 1st (MSB) bytes respectively.

I'm also wondering if the ROMs have been tinkered with before as three are AMI chips and one is a VLSI which seems odd as you would think they would be installed in sets.

 

[phipli]

I was just wondering which chip went where but I see they are labelled LL, ML MH and HH which is presumably 4th, 3rd, 2nd and 1st (MSB) bytes respectively.

I don't think that would make sense, surely the low byte would be the low byte, regardless of if you were LSB or MSB? I think you're perhaps over thinking it. Easy to check by reading the removed chips.

I'm also wondering if the ROMs have been tinkered with before as three are AMI chips and one is a VLSI which seems odd as you would think they would be installed in sets.

I wouldn't jump to any conclusions - Apple mixed chips fairly often. This is a Performa 630 :

 

[rodders]

Having splashed out on a de-soldering station I decided to remove the ROM chips one by one to see if the select line started to behave properly.
I started with UA1 (HH) no change, UA2 (MH) no change, UA4 (ML) no change and finally UA5 (LL) and success, the SEL line is now pulled low.
it looks like UA5 was clamping the line, unfortunately it doesn't mean that the others are OK, I'll see if I can read them with my TL866.
I plan to put sockets in rather than soldering the ROMs directly in case its not the end of the story. Just waiting for parts to arrive.

This thread was started by other guys with these symptoms, I wonder if their ROMs could be bad too?

 

[phipli]

Having splashed out on a de-soldering station I decided to remove the ROM chips one by one to see if the select line started to behave properly.
I started with UA1 (HH) no change, UA2 (MH) no change, UA4 (ML) no change and finally UA5 (LL) and success, the SEL line is now pulled low.
it looks like UA5 was clamping the line, unfortunately it doesn't mean that the others are OK, I'll see if I can read them with my TL866.
I plan to put sockets in rather than soldering the ROMs directly in case its not the end of the story. Just waiting for parts to arrive.

This thread was started by other guys with these symptoms, I wonder if their ROMs could be bad too?

That sounds like good news!

One thing I'd check, try measuring the resistance between SEL and 5v on each of the ROM chip footprints - to check if there is something holding it high on the board. Edit : Was being stupid.

Another thing is given it only worked after you removed the final ROM, that might indicate that the SEL signal is attached but very weak - perhaps the trace is almost broken and just connected by a sliver of copper.

A couple of things that can be checked while you're waiting for parts if you get bored

Socketting the chips is sensible. If nothing else it makes it easier to write custom ROMs if you wanted to for any reason.

 

[rodders]

One thing I'd check, try measuring the resistance between SEL and 5v on each of the ROM chip footprints - to check if there is something holding it high on the board. Edit : Was being stupid.

Doesn't sound that stupid.

Another thing is given it only worked after you removed the final ROM, that might indicate that the SEL signal is attached but very weak - perhaps the trace is almost broken and just connected by a sliver of copper.

Good thought, I'll have a look.

I've tried reading the ROMs as 27512s and its a bit weird. I managed to read the three Apple chips, LL and ML didn't have anything recognisable, I need to compare them with the downloaded ROM image. MH however was all FF which is odd. The VLSI (HH) chip couldn't be read as the reader complained of a pin detect error for pins 15-19 which implies that its a different pinout. Its labelled as a VT23512 but I couldn't find any support for that in the TL866.

 

[phipli]

LL and ML didn't have anything recognisable

They wont, remember that you're looking at quarter of 32bit words cut up, not continuous data.

MH however was all FF which is odd

Remove it from the socket, clean the pins and try again, it probably just isn't reading properly, or is possibly faulty.

The VLSI (HH) chip couldn't be read as the reader complained of a pin detect error for pins 15-19 which implies that its a different pinout.

That is the one that you suspect is bad, so that isn't a surprise. If it works in the IIcx, it will have the same pinout as the others. Or should have.

I need to compare them with the downloaded ROM image.

Put one (lets say LL) in the reader, load the low byte from the saved rom from the internet into the read/write software, then just click verify. If you change the ROM type you'll need to re-load the saved ROM because I think it flushes out what you have loaded.

 

[rodders]

That is the one that you suspect is bad, so that isn't a surprise. If it works in the IIcx, it will have the same pinout as the others. Or should have.

No, the one I suspect is LL, the last one I removed. I haven't been able to find a datasheet for the VT23512 to check it out.
Do you know if the TL866 uses the select line? I was a bit surprised that it managed to read the one I think is duff.
I've ordered some W27C512s so I think I'll end up burning a set.

 

[phipli]

No, the one I suspect is LL, the last one I removed.

Ah sorry, I thought you'd removed them in order for some reason.

Could you upload a photo of your ROMs?

I was a bit surprised that it managed to read the one I think is duff.

Might be driving harder, or the fault is still on the board - the chip might be completing a circuit that is caused by a short or something weird like that. No fault of the chip. Possibly.

What it is is a clue, which is nice

 

[rodders]

Could you upload a photo of your ROMs?

Here's a picture, UA1(HH) on the left UA5(LL) on the right.

I managed to read the VLSI chip by turning off the pin detect so there's something odd there but I'm guessing the chip type (AM27512) so there may be an incompatibility there.
When verifying against '1988-09 - 97221136 - Mac II FDHD & IIx & IIcx.ROM' UA1 (HH) successfully verifies against byte 1, UA4 (ML) against byte 3 and UA5 (LL) against byte 4. As for UA2 (MH) it reads successfully but only produces all FFs so I'm pretty sure its dead. Verifying it against byte 2 obviously fails. What is puzzling is why removing it didn't fix the /SEL problem which implies there is more to find.
Further inspection of the /SEL traces hasn't revealed anything, there is no resistance and the pullup resistor measures at 4K7 which is correct.
The fact that /SEL was clamped high until I removed UA5 (the last to be removed) suggests that it also has a problem. Its surprising that it reads and verifies OK but maybe the TL866 doesn't use the /SEL pin.

 

[phipli]

I managed to read the VLSI chip by turning off the pin detect so there's something odd there but I'm guessing the chip type (AM27512) so there may be an incompatibility there.

Was it turning off the pin detect, or did you turn off "check ID"? Or is that already off I guess?

The pin detect might be failing due to it being a mask ROM - the erase pin might not be being detected as normal or something. No idea.

When verifying against '1988-09 - 97221136 - Mac II FDHD & IIx & IIcx.ROM' UA1 (HH) successfully verifies against byte 1, UA4 (ML) against byte 3 and UA5 (LL) against byte 4. As for UA2 (MH) it reads successfully but only produces all FFs so I'm pretty sure its dead.

OK, at this point, its probably best to assume your primary issue is that the chip UA2 is dead. This would stop the computer booting, so work on fixing that before anything else.

Further inspection of the /SEL traces hasn't revealed anything, there is no resistance and the pullup resistor measures at 4K7 which is correct.
The fact that /SEL was clamped high until I removed UA5 (the last to be removed) suggests that it also has a problem. Its surprising that it reads and verifies OK but maybe the TL866 doesn't use the /SEL pin.

I'm going to suggest leaving /SEL for now. See what happens once you have a full set of working chips installed. Perhaps any unusual behavior of /SEL is just a result of not getting the expected response from the ROMs.

If a good set of ROMs doesn't fix everything, then come back to /SEL as the next investigation.

 

[rodders]

Was it turning off the pin detect, or did you turn off "check ID"? Or is that already off I guess?

The pin detect might be failing due to it being a mask ROM - the erase pin might not be being detected as normal or something. No idea.check ID was already off, it was pin detect errors I was getting but turning off pin detect allowed the ROM to be read

i had already turned off check ID, it was turning off pin detect that enabled the ROM to be read. I agree that it may be some peculiarity of this particular device. The TL866 manual says that some chips may fail the pin detect which is why there is an option to turn it off.

OK, at this point, its probably best to assume your primary issue is that the chip UA2 is dead. This would stop the computer booting, so work on fixing that before anything else.

I totally agree. When the parts arrive I'll burn a full set of ROMs and see if the machine boots. I may swap in the old ROMs to see if they are any good later.

 

[rodders]

The chips have arrived! Sockets soldered in. EEPROMS burned and fitted. Then this!

Not being familiar with Macs I'm hoping this is the happy mac sound!

 

[phipli]

The chips have arrived! Sockets soldered in. EEPROMS burned and fitted. Then this!
View attachment 19955
Not being familiar with Macs I'm hoping this is the happy mac sound!

That is a sad chime, and means the basic hardware test failed (sort of similar to a flashing floppy disk light on an Acorn).

But...

Given you don't have any RAM installed, I wouldn't expect anything else. So given the context, this is excellent news! Well done!

You need to put four 30pin memory simms into Bank A to get any further.

This is what a Happy Mac IIcx sounds like if I remember correctly :

 

[phipli]

Oh, and if it isn't too late, take super care, moreso when removing SIMMs, but the catches on those SIMM slots are insanely brittle.

It can help to heat them a little with a hair dryer first. But regardless, avoid moving them any more than needed, and try to apply force to more than just the very tip.

 

[rodders]

I've been very careful with those nasty plastic SIMM clips
Anyhow, I've put it all back together and, as Adrian Black would say, "It freakin' works!", even the monitor!
I've managed to boot from a 6.0.8 floppy and everything seems fine except for a couple of dead keys on the keyboard.
I'm powering it from a lashed up PSU at the moment as the original that I'd 'upgraded' with the guts of a cheap Chinese ATX PSU has extremely noisy outputs. Its as though the output filter isn't working at all. I need to look at that next.

 

[phipli]

Anyhow, I've put it all back together and, as Andrew Black would say, "It freakin' works!", even the monitor!
I've managed to boot from a 6.0.8 floppy and everything seems fine

Excellent news! The drive should be a 1.44MB drive, which means with a bit of creativity and software you should be able to write disks from a PC. I'm actually impressed the floppy drive works, might be worth giving it a careful clean and lubrication if you haven't already. A stiff mechanism can lead to a broken eject gear sometimes.

except for a couple of dead keys on the keyboard.

I see you have one of the really nice keyboards. That one has mechanical alps switches. I actually prefer them to the huge "surfboards". Given it is mechanical, I'd probably just press the non-working keys a couple of hundred times and they'll probably start working again. Otherwise, caps off and some contact cleaner. I recommend not trying to dismantle spacebar or the return key (or possibly enter). They have little plastic clips that are broken on some of mine from other people trying to take them apart before I owned them.

Let me know how it goes with the PSU. I need to rebuild some of mine at some point. I have enough to swap between machines so haven't put much effort in. There is only so many machines you can use at once sort of thing.

Also feel free to ask about what software you should find based on your interests etc. Games / drawing / music / animation etc etc.

 

[rodders]

Otherwise, caps off and some contact cleaner.

Hmm, I've already given that a go but maybe need a few hundred more clicks!

Let me know how it goes with the PSU. I need to rebuild some of mine at some point.

If you plan to rebuild some, have a look at this https://github.com/roddersuk/Apple-IIcx-PSU-ATX-Adapter. I have some spare PCBs if you're interested.

Also feel free to ask about what software you should find based on your interests etc. Games / drawing / music / animation etc etc

Thanks for this, I'll let you know when its in a fit state to use i.e. not lashed up on the bench!

 

[phipli]

If you plan to rebuild some, have a look at this https://github.com/roddersuk/Apple-IIcx-PSU-ATX-Adapter. I have some spare PCBs if you're interested.

I have some of these on the side : https://github.com/GeorgeRudolf/OldMacATX I tend to just use them for bench testing though. When I don't actually just use a period PSU. Thank you for the offer though. I will have a look at the design out of interest too.

Thanks for this, I'll let you know when its in a fit state to use i.e. not lashed up on the bench!

One IIcx thing - it is one of the few machines that is fully compatible with the colour version of Space Quest. I think it requires an older version of the QuickDraw software and a newer version was baked into the ROM on later machines? Something like that. Also, the sound doesn't work in emulators, so it is rare to see it working if you don't have an early Mac II. From what I understand anyway... I haven't actually tried it on my IIci or anything newer, only my IIcx. It should also work on a II, IIx or SE/30 with a colour card - they're all basically the same computer.

(By the way, the sound is really unimpressive for a Mac - most games that run on a Mac II have sample based midi style music that runs in software).

Lemmings would be a better example of what the IIcx can do music wise :