Oops, my obsession with computers from my childhood strikes again!

[PhilC]

@rubber_jonnie you could put something in the freezer for an hour and then sit it on top of the MMU and see if it runs for longer?

[rubber_jonnie]

@rubber_jonnie you could put something in the freezer for an hour and then sit it on top of the MMU and see if it runs for longer?

The thought had crossed my mind but I was concerned about condensation.

That being said, I broke out the compressor and airbrush and repeatedly made it fail, and after each attempt, used the airbrush to blow cold air over different ICs on each fail. However, either the air wasn't cold enough, or the flow was insufficient to get any meaningful results.

I had ordered freeze spray but just got a message it'll be delivered tomorrow now and not today

I did time failures today and from cold it took about 15 minutes to fail and from warm, with a 10 minute cooldown post failure, it was about 8 minutes to fail, so I really do believe it's something thermal.

I am considering a hot air blower to see if the IC's respond to that and the problem arises quicker, but it's still just a thought for now.

[stephen_usher]

Unfortunately, if you've changed the transceiver the only thing which could cause the problem is the MMU as there is literally nothing else in the path between the slots, memory and CPU.

What happens if you lift the /DIR pin of the transceiver and tie it to either gnd or +5 (via pull-up), whichever makes the slots be in memory read mode?

If it's the MD IN/OUT line then the 80 column graphics card will be able to read from memory but the system won't have any data from the slots poluting the data bus. The system should then remain stable if it's not the transceiver chip but instead the MMU controlling it that's the problem.

[rubber_jonnie]

Unfortunately, if you've changed the transceiver the only thing which could cause the problem is the MMU as there is literally nothing else in the path between the slots, memory and CPU.

What happens if you lift the /DIR pin of the transceiver and tie it to either gnd or +5 (via pull-up), whichever makes the slots be in memory read mode?

If it's the MD IN/OUT line then the 80 column graphics card will be able to read from memory but the system won't have any data from the slots poluting the data bus. The system should then remain stable if it's not the transceiver chip but instead the MMU controlling it that's the problem.

This is indeed my worry, I think the MMUs are like hens teeth, so if it has indeed gone, I'm stuck with a base machine that I can't expand

[stephen_usher]

I'll be surprised if no-one's reverse engineered the MMU chip by now.

[stephen_usher]

If there's no full recreation it should be easy to replicate the functionality of the BD IN/OUT line as this will be a combination of the address decode and bus read/write. You should be able to tap into those and use a GAL to replicate the functionality. I know the MMU does a lot more but for the expansion ports it would be a direct integration of the original Apple II circuitry.

It may be useful to 'scope the BD IN/OUT line to see what it looks like. Sometimes with these faults it's the line driver or internal pull-up gets weak. If it's the pull-up you could put a helper resistor, something like a 10K. If it's the driver then that's harder to fix.

[Tomswork]

Good source for rom and hal

https://downloads.reactivemicro.com/App ... JEDEC/IIe/

Tom

@rubber_jonnie

[rubber_jonnie]

I'll be surprised if no-one's reverse engineered the MMU chip by now.

Me too, and I'll be having a look.

[rubber_jonnie]

Good source for rom and hal

https://downloads.reactivemicro.com/App ... JEDEC/IIe/

Tom

@rubber_jonnie

I had wondered if the HAL could be easily replaced, and it seems the answer is yes

Thanks.

[rubber_jonnie]

If there's no full recreation it should be easy to replicate the functionality of the BD IN/OUT line as this will be a combination of the address decode and bus read/write. You should be able to tap into those and use a GAL to replicate the functionality. I know the MMU does a lot more but for the expansion ports it would be a direct integration of the original Apple II circuitry.

It may be useful to 'scope the BD IN/OUT line to see what it looks like. Sometimes with these faults it's the line driver or internal pull-up gets weak. If it's the pull-up you could put a helper resistor, something like a 10K. If it's the driver then that's harder to fix.

I'm slightly confused here, did you mean MD IN/OUT as I don't see BD IN/OUT anywhere?

I can scope it, yes, the main problem is knowing what's good and bad. I can video again though, rather than sit and watch for a change.