Amiga drive signal issue (SOLVED via bodge wires)

[Turrican3]

EDIT: perhaps the thread was more suited to the repair section? I don't know....

=============

Hey there,

I own an Amiga A1200 that has recently shown (likely due to a reverse connector insertion I did, but that's beside the scope I guess) a wrong behaviour.

Basically, the floppy disk motor is always spinning, except briefly at reset/cold boot.

I've had a look at how it's supposed to work and I see this signal is generated by a custom IC called AA-Gayle, depending on the state of two input signals.

Upon further inspection it seems one of the aforementioned input signals that dictate the motor status, _MTR, is always at 0 Volts, and is shorted to ground. I have removed the other two chips that are related to _MTR (one is the chip that generates it, the other takes it as input) to check whether this would make any difference, but I can still see the short to ground.

So my questions are:

1. is it reasonable to assume AA-Gayle is actually faulty? My next planned step is to remove this chip (waiting for proper hot air nozzle) and see if the short to ground of that pin disappears

2. I understand _MTR and the other input signal just go through a flip flop, I wonder if I could then put an external flip flop of the proper type and reconstruct the signal, lifting the relevant AA-Gayle pins

Any help would be appreciated.

[m k]

Reasonable is a bit subjective expression.

Have you removed U26?

_MTR0 is just a spinner, surely you can fix it with external connection.

[Turrican3]

Yes I have removed both U26 and the CIA 8520 that generates _MTR, but the signal still appears to be shorted to ground even with those two chips removed, that's why I thought Gayle might be faulty.

[m k]

If _SEL0 is up and drive is still spinning then it starts to be reasonable to think that Gayle is bad.

Drive itself is also a possibility and that E599 filter.

[Turrican3]

Forgot to say thanks before, I'll do it now. :-)

Having said that, floppy is definitely working, as connecting it to an A600 (same cables) shows no issue.

I'll see what happens with _SEL0 and report back (I *think* I've already checked that too, but I want to be sure before confirming)

[Turrican3]

I can confirm (see attached logic analyzer capture) that regardless of its level, _SEL0 doesn't seem to affect the drive, which keeps spinning.

All the signals have been taken directly from AA-Gayle, and I can also confirm CIA U8 doesn't seem to be able to drive the _MTR signal, which is stuck to 0V and shorted to ground.

[m k]

It's pretty clear then.

I seem to have a recollection where something similar has happened.
But I also have a recollection that wrongly positioned 8520 VIA didn't do a thing, if that was the original situation.

Since only _SEL0 goes through GAYLE one can wonder how theoretical second internal drive is used.

Your external connection may need a delay, so simple gate chip may not be enough.
But that's possibly just a cosmetic thing, where the drive possibly turns on for a moment during power up.
When writing this I had a next recollection where inserted disk and wrongly turning on drive had some issues but that can be from a completely different machine.

[Turrican3]

Thanks!
Right now I'm waiting for a replacement Gayle plus proper hot air nozzle.

Assuming the short to ground is indeed related to Gayle I expect it to disappear as soon as I remove the chip.

If that's the case I'll see if I can source the right flip flop (latch?) and try that first, since I don't really care that much about possible external drive issues. But if it doesn't go as planned, I'll just replace Gayle.

Will report back anyway in (hopefully) a few days.

[Turrican3]

Managed to remove AA-Gayle but... perhaps surprisingly, even with the chip out of the board, /MTR still appears to be shorted to ground.

Could someone suggest any further checks?

EDIT: ok here's a crazy (?) idea... perhaps since the /MTR signal only goes to U5 and U26 (and the former now it's likely ok, I'll put it on a socket anyway) I might just route the signal directly from the CIA U8 with a couple of bodge wires?

[darkspr1te]

Managed to remove AA-Gayle but... perhaps surprisingly, even with the chip out of the board, /MTR still appears to be shorted to ground.

Could someone suggest any further checks?

A quick look at https://www.amigawiki.org/dnl/schematics/A1200_R1.pdf shows that _MTR has 100pf cap on the line (page 9, right hand side ), check to see if this is shorted to ground on your board


darkspr1te

[Turrican3]

Thanks for getting in touch (please have a look at my EDIT if you can, too!)

Unfortunately you're talking about the /MTR0 signal, which is a different one from /MTR; it's on the very same page but on the left.

[darkspr1te]

Ah, you are talk about the signal PRE gayle, the signal from the 8520 _MTR and not after the gayle chip _MTR0 , my bad i missed that,


darkspr1te

[Turrican3]

Don't worry it's ok! :-)

And yes that's correct, the signal shorted to ground is pre-Gayle. On the other hand post-Gayle _MTR0 seems to be ok, as during the reset procedure Gayle consistently forces it to high for a few milliseconds IIRC (I'd have to check for the exact value, can't remember right now), as per specifications.

[m k]

Could someone suggest any further checks?

Have you had all 3 components off simultaneously?

Check also that there's nothing under 8520 socket.
It's also a multi layer module so possible trace hole can have a ground connection.

[Turrican3]

I haven't had all 3 ICs off yet, but please note _MTR already appeared to be shorted to ground before I installed the socket for the 8520 (i.e. with Gayle being the only IC installed of the three set).

I can pull the XOR IC off again anyway and see what happens with the 8520 removed from the socket (nothing unusual seems to be there, at least to my untrained eye, but I could be wrong by the way)

[m k]

You can also have more than one fault.

[Turrican3]

Oh absolutely, but I hope that's not the case!

CIA removed from the socket, Gayle and XOR Gate ICs removed, short to ground is still there.

I think I'll try the bodge wire solution, assuming I'm able to properly isolate the relevant PLCC ICs pins... yes I could just cut the tracks and call it a day, but I'd prefer a non-destructive, reversible fix.

[darkspr1te]

if you can isolate the the line from the rest of the board why not try injecting power into the line and using alcohol along the trace to see which heats up ?
or cut the trace at a point repairable and see which half still has the short

darkspr1te
 

[Turrican3]

Being a newbie I'm not 100% sure about the suggested power injection method. This signal basically goes from chip A to chip B and C, where there appears to be continuity. I suppose I'd need to add a resistor to avoid burning everything and apply a low voltage (less than 5V possibly?), low current (no idea here) between A and C and see what happens?

Regarding the latter, as I said I'd prefer a reversible fix since this is a relatively rare PCB. I'm not willing to sell it mind you, so I might probably avoid worrying too much about this, but I wouldn't mind keeping it as genuine as possible nonetheless...

[m k]

I'm not exactly a fan of fault finding cuts either.
Easiest cutting place is a bit away from free through hole, then repairing can use the hole as a wire support.
Wooden stick will also be a great help.

You can also try to find a milliohm meter and measure which point is closest to ground.

Seems that you may have done something mechanical somewhere earlier.
Go back and think what are those things that can be the cause, and maybe even more importantly, think also situations that can't be it.
Finally you may have a small and forgotten detail that doesn't actually fit and leads you to the source.

Check the under side again also and as carefully as you've done the upper side.
Extra particle of something somewhere wouldn't be the first time.

[Turrican3]

Seems that you may have done something mechanical somewhere earlier.

I definitely did something wrong before this fault emerged, but seemed more like an electric/electronic thing rather than mechanical: I put an RTC clock expansion the wrong way to the nearby internal connector.

Now, whether this was a mere coincidence or not, I have no idea, but I guess since I cannot go back in time and avoid this silly mistake on my part the only thing I can do now is try to fix the current issue. :-)

I will report back as soon as possible.

[m k]

So you shorted something with the RTC board and the current went through one of those _MTR chips and burned something external after that?

[Turrican3]

It's possible but I can't say for sure, as I've ran many extensive tests after discovering this issue, and nothing unusual emerged. The RTC board works perfectly fine too once connected the right way.

The CIA that generates _MTR is confirmed working because I tested it on an A600, for the record.

Right now I'm trying to see if I can route the signal via bodge wires. Unfortunately I've inspected both sides of the motherboard and the track shorted to ground doesn't show any visible damage/bridges or something like that.

[Turrican3]

It's working!

So as I suspected after removing it, Gayle was fine too. It's probably not the best professionally looking repair but I guess it could be worse, too?

Really happy about this fix, thanks for helping me everyone.

[jmelson]

It's possible but I can't say for sure, as I've ran many extensive tests after discovering this issue, and nothing unusual emerged. The RTC board works perfectly fine too once connected the right way.

The CIA that generates _MTR is confirmed working because I tested it on an A600, for the record.

Right now I'm trying to see if I can route the signal via bodge wires. Unfortunately I've inspected both sides of the motherboard and the track shorted to ground doesn't show any visible damage/bridges or something like that.

MANY years ago I had a fault on a minicomputer mag tape controller.  The reset/ line went to a bunch of chips, and was shorted to ground.  in desperation, I made a few foil cuts and found a 1" long section of trace that was shorted to ground.  There was no ground anywhere near that section.  But, I peeled back the bad section, there was no visible anomaly under it.  I patched it with insulated wire wrap wire and the unit worked fine afterwards.  That was ONE CRAZY fault!
Jon

[Turrican3]

MANY years ago I had a fault on a minicomputer mag tape controller.  The reset/ line went to a bunch of chips, and was shorted to ground.
[...]

I can see that must have drove you crazy!

[m k]

I have a recollection, yes, I know.
There the short was in a hole, recollection is also offering that the board was Amiga something but I don't believe it.
Anyway, the hole was originally misplaced so that between it and inner plate had only a fraction of space and then one day, for what ever reason, that fraction was none.