Broken reset circuit / 555?

[mikro]

Interestingly, Videl now doesn't show up as a primary suspect despite the fact that DE is not active at all. So yeah, maybe I'll start with Videl, despite all the measurements.

And I'm glad I did because of course it was Videl's doing:

PXL_20250813_103159891.jpg

PXL_20250813_103221715.jpg

(yeah, I'm such a noob that I can't lift just one pin, I had to lift three of them and even while doing that I actually lifted the fourth one :D)

[Badwolf]

Interestingly, Videl now doesn't show up as a primary suspect despite the fact that DE is not active at all. So yeah, maybe I'll start with Videl, despite all the measurements.

And I'm glad I did because of course it was Videl's doing:

Good work.

Unlikely with two failed Videls, though. :(

BW

[mikro]

I suspect, following @dml's speculation, that in the first Falcon it was Mighty Sonic 32's job (inserted wrongly) because not only I can't get it operate 100% reliably (e.g. I can't switch it to 32 MHz but there's still a slight chance that the bad CPU on the motherboard somehow interferes...) but it took with it the CPU, Videl and the GALs.

What however have these two Falcons in common is the fact that they are both not recap'ed and using some ATX PSU in a tower housing. So yeah, one voltage peak and Videl's dead.

Other people told me today that Videl being dead is far from uncommon, there was a guy who also fixed two broken Falcons and both had dead Videls. IIRC even @exxos had to replace one.

[exxos]

Dead videls seems common to me. I've replaced several when I fixed falcons for people. Those falcons generally started with users messing around with the PSU or its in bad shape. Nobody replaces the bulk cap either. Ive screamed from the rooftops for years about it all.

But also don't forget I've had ram issues because of "noise" on the videl. It's possible that could also be a factor in them failing. It's why I designed the spike immune STram board a couple years ago as lots of RAM kept failing and I kept seeing random RAM errors. All those problems went away with the prototype board I've been running for the past year or so in my falcon. Others have had RAM fail in their falcons as well. It's certainly not helping if the PSU is spiking like crazy. Think I've seen up to 9volts IIRC.

EDIT
viewtopic.php?p=118667#p118667

[stephen_usher]

Would a Zener diode between +5V and ground help clamp the spikes or are they too slow to respond?

[mikro]

So, Videl's out! Of course, I can't do any real testing yet but I could at least try to prove one thing: does Falcon boot with absent Videl?

The answer is: maybe. ;)

Remember, this is my CT2'ed Falcon so I can't easily "revert to stock", I have to work with what I have got. So... since I know that all Ax/Dx lines are fine now, I tried to boot in the normal (non-turbo) mode, booting from ROM and not seeing anything from the CT2 except CPU/FPU.

Nothing on serial line.

Strangely, I got an idea to try the same thing in the turbo mode, booting into TOS 7.04 with TT RAM support etc. Surprisingly, after several presses of the reset button:

PXL_20250815_152158179.jpg

Hmm. Taking https://www.exxoshost.co.uk/forum/viewt ... 143#p47751 for help, it seems that CPU tries to write zeroes to address 0x0008 and it reads garbage back.

You see the problem - since this is happening only in the turbo mode, I can't really say whether this is due to missing Videl, a fault in the non-turbo mode (like bad memory card / pins?) or a series of faults which CT2 somehow mitigates into somewhat booting state.

We shall know for sure once the new Videl has been mounted.

[Badwolf]

Strangely, I got an idea to try the same thing in the turbo mode, booting into TOS 7.04 with TT RAM support etc. Surprisingly, after several presses of the reset button:
PXL_20250815_152158179.jpg

Whoa, so that's the diag cart outputting over serial without a Videl?

Good to know if so.

BW

[mikro]

Whoa, so that's the diag cart outputting over serial without a Videl?

Yes, but with TOS being read from CT2's flash. As mentioned, I can't be really sure what we are seeing here, maybe absent Videl = TOS on the motherboard is unable to boot, who knows.

Actually I could have measured it but I'm lazy to take the PCB back to the office with the scope. Will mount Videl later and we will get our answer. I certainly root for the result that everything will be working again and therefore missing Videl is a showstopper. ;)

[dml]

We shall know for sure once the new Videl has been mounted.

Will be interesting to see - but with the Videl missing I think the worst it can do is just not respond to the Combel. It is not configured to be able to write to RAM so I think even floating pins couldn't do it. So probably at worst, it would halt the machine - but we learned from ijor that can't really happen either.

But who knows for sure :) try and see!

[mikro]

So, the final update. I hope.

After replacing the Videl ... both the normal and turbo mode work. So that puts the ability to boot without Videl really into question.

However: there was one, erm, situation. After Jookie finished with the Videl, he noticed that there were two legs (!) lying around. Most likely from the Videl removal operation. We were both quite surprised, Jookie was cutting one leg at time, it wasn't like he had a dremel or similar. So it's quite suspicious where the legs came from.

Anyway, it's possible that the weirdness observed in the post above was caused by those legs. Too bad we noticed this *after* soldering the new Videl. :-( So the mystery didn't get a 100% evidence to support or reject that claim.

Anyway, Falcon boots, so that makes two saved Falcons in two months, nice!