JezC - the second (and laST) H5C1 build

[exxos]

@Steve I think your board had the reset troubles before the OSC fault was discovered though ? viewtopic.php?f=89&t=5600

If it was that board you sent me ages ago, it booted up fine for me, but then again I was using a different power supply.. But I suspect it was a problem with the OSC itself. The original batches were AXO brand, to my knowledge there wasn't any troubles with those.

I think the trouble only started with the later batch OSCs where I had to have them custom-made by YXC. The resistor modification made it into the H5C3 revision anyway as I think any current H5 builds will have the same problem.

[Steve]

Ah yeah, time flies.. so many things happened with my H5. I remember though, I did send it to you & you did these various fixes. It came back to me and it still didn't work! Even though it worked for you. Not 100% sure what I did next, but replacing the chipset with SMD variants sticks out in my mind.

[JezC]

Ok, I remembered to bring the resistors home for the 32MHz OSC fix, so will try to fit one after work today...

[JezC]

....and no difference

As I forgot to bring a 'scope home from work for the weekend, I'm a bit limited on checking if this has improved the signal levels on the clock or not.
Using a DMM I'm seeing a resistance between 0v and the 32MHz Osc output (after the inline resistor) where it was showing 0/c before, so I think the resistor is working (and I also checked it is indeed 470 ohms).

I'll try removing & re-inserting the PLCC Glue & MMU (and possibly 68k & 68901 & EPROM) later to see if I can find/fix any poor connections on them.

[exxos]

I don't know if the OSC would kill the shifter. May be worth swapping to see if it helps...

[JezC]

I don't know if the OSC would kill the shifter. May be worth swapping to see if it helps...

Ok, might try that before I reseat the PLCC devices - thanks!

[JezC]

I don't know if the OSC would kill the shifter. May be worth swapping to see if it helps...

Ok, might try that before I reseat the PLCC devices - thanks!

Nope, swapping the shifter hasn't changed anything on the new H5C1...and the shifter still works OK in the other H5C1 (shows 1 floppy connected on the boot screen - modified TOS 1.04 with IDE support & then boots to desktop showing A and B drive icons as long as there is no floppy disc in the drive, gives 2 bombs if there is - after the initial boot screen shows the 1x floppy connected).

So, the shifter ahs survived (and the 68K, MMU and GLUE are OK in the other H5C1 as well)...

I'll try to have another look at the floppy access issue on the original H5C1 while it's on the "operating table" - but any suggestions about how the accessing the floppy disc differs compared to checking a floppy is connected during boot might help localise the initial connection checks I guess...?

[exxos]

Difficult to say without looking at TOS sources. But I'd assume TOS "pings" the floppy drives and if it gets a reply then the drive is there. Then it must access more info like getting FAT and then fails. So sounds like corruption somewhere between the 1772 to DMA or DMA to the main databus. But it could be anything like a bad clock to the 1772 etc..

[JezC]

Ok, had a spare minute or two at home after work (with the 'scope this time!)...

The booting H5C1 (with floppy issues) has an AXO osc fitted and even that is showing 32MHz with 5.5v peak-peak...(without the 470 ohm resistor to GND on the output).

I'll get the second H5C1 set up later this evening & check what the clock is doing on that...

[JezC]

Hmm, not quite later the same night (ended up having to borrow the 68k PLCC from an STE and the MMU/GLUE from another STFM)...

So 32MHz clock is running OK and just over 5v p-p but still no screen activity on boot or reset.

Next job is to go through the non-boot diagnostic steps...and see whether there is any activity.