exxos's Atari, Amiga & retro forum

Sorry, yes, BD IN/OUT.

Good/bad would be voltage levels. It’s a single wire obviously supposed to be pulled up by the MMU. If you’re getting noisy or not very good voltage levels on that line then the transceiver will be all over the place and spuriously sending data in the wrong directions. It’s quite possible that the internal pull-up in the MMU is flaky or it can’t pull the line down fully. Either way looking at the signal should tell you if it’s not getting the correct voltages.

stephen_usher wrote: 08 Apr 2023 11:56 Sorry, yes, BD IN/OUT.

Good/bad would be voltage levels. It’s a single wire obviously supposed to be pulled up by the MMU. If you’re getting noisy or not very good voltage levels on that line then the transceiver will be all over the place and spuriously sending data in the wrong directions. It’s quite possible that the internal pull-up in the MMU is flaky or it can’t pull the line down fully. Either way looking at the signal should tell you if it’s not getting the correct voltages.

I had a look at MD IN/OUT and got what I would expect, mostly low and occasional highs, such as when I was typing in my test program. However, my freeze spray turned up, and it seemed to help when freezing the MMU.

Another however here, I decided to try alternative CPUs as I have 4 spares and suddenly I can't boot, not even with the original CPU. So I gently pulled the plastic frame off the CPU socket, and off fall two contacts that were clearly not happy.

So, any further investigation is on hold until I fit a new CPU socket. I can't help but think there are more problems like this lurking in some of the other original sockets, I may do some more replacements on the larger ICs just to rule them out as faulty. Bah humbug, I do so hate these older single wipe sockets.

@rubber_jonnie

I have had a couple of hals go on me and some of the custom ic are great candidates for ri pie heatsinks

Tom

Tomswork wrote: 08 Apr 2023 20:11 @rubber_jonnie

I have had a couple of hals go on me and some of the custom ic are great candidates for ri pie heatsinks

Tom

Well, it's all kind of gone sideways :pullhair: :pullhair: :pullhair:

Now the machine won't boot at all with the new CPU socket fitted, and I honestly don't know if it's the CPU or the MMU or even something else :(

Well, I say won't boot, it powers on but all I get is on screen garbage, like the CPU is removed and it never clears the screen RAM. Too late for any more now and tomorrow we're out, so it'll have to sit and wait.

I would say, it was booting with the MMU/IOU sockets replaced, so whilst it could be the MMU has suddenly gone, it could also still be the new CPU socket, so I need to buzz it out, but god damn it GRRRRR.

It could still be the HAL I guess, more digging to be done, but a quick check did suggest there was no clock at the 6502, so it could be that as there seems to be 5v and /RESET is high as expected.

Merdé

PS I agree re the RPi heatsinks.

Time for the scope start at the clocks then halt line then data and address see if its getting out at all. And the pcb are bad for getting shorts under sockets

Tom

@rubber_jonnie

Tomswork wrote: 09 Apr 2023 05:50 Time for the scope start at the clocks then halt line then data and address see if its getting out at all. And the pcb are bad for getting shorts under sockets

Tom

@rubber_jonnie

Yep, it is indeed scope time. Plus I'll double check my CPU socket for shorts.

Such is the life of a collector of old computers.

Socket has no shorts, but I need to check all the pins connect to where they should go as its like there is no CPU fitted ATM. Oh, and let's not forget that the clock signal that goes direct from the LS125 is a little off...

Ok, so from the dire straits of yesterday where it didn't want to boot at all and behaved as though it had no CPU, it is actually now booting to the basic prompt again, however it is very unstable as the clock is somewhat unstable.

I'll take that though as it at least give me something to aim at. Interestingly, with the 'scope leads connected to the CPU clock in and the clock out on the 74S02, it does seem to stabilise a little, at least enough to type in some basic commands.

It could still easily be an issue with the MMU as well, so I need to figure out where I go next with it, but at least it is able to boot up now.

I was wondering if residue from the freeze spray may have been conductive and was causing issues since it's difficult to get it neatly in one spot, so I figured if I spray into and under sockets where I'd used it, with IPA, it may help matters.

So I took an extreme step yesterday and took the board down to my garage with my little compressor and airbrush and 'jetwashed' the sockets and board with IPA. It wasn't fierce, but it was strong enough to clean everything nicely.

It's a technique I expect I'll use again!

Well this is this morning:


It's running Elite just fine and the clock signal is nice and tidy and stable again. It's now been up and stable for well over an hour.

So, I'm back to square one, and can start looking at the 80 Column/RAM card issue again.

Phew.

Tomswork wrote: 08 Apr 2023 03:38 Good source for rom and hal

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

Tom

@rubber_jonnie

Out of interest, I tried programming a GAL16V8B with the appropriate JED file and it didn't work...

Have you tried a replacement yourself?