Need help diagnosing my PAL Amiga 500+

[SamBushman]

So I DO have a Logic Probe. While I won't be able to identify specific bytes or anything coming across the bus, I can at least speak to whether we are seeing activity.

I tested Pin 1 of U10, U11, U12, and U13. I also tested Pins 1, 2, 18, and 19 of each RAM chip. All showed activity and all their signals changed pattern as DiagRom went through it's different screen colors (which I believe represents different kinds of hardware checks).

I have also just broken down and ordered a real oscilloscope. Once that comes in I can get some more specific information on what's coming down the various lines of the data bus.

[solarmon]

Even if you see the signals getting to the ram chips, it still could be a failed ram chip.

For my A500+ after fixing the first time from battery leak it failed 6 months later and I had to socket the ram chips before I identified the bad ram chip.

[SamBushman]

Would bad ram chips cause the serial output I'm seeing? My understanding is DiagRom will output the results of RAM checks and report failed memory. If bad memory would cause the serial data to corrupt, then I don't understand how these checks could ever work (and therefore how the RAM could be to blame here).

Don't get me wrong, I've bought more RAM chips in case I found such problems. I just haven't followed through with socketing and replacing them yet on account of the above assumption (I don't want to just be swapping stuff out blindly). If it may actually be a root cause I'm open to giving it a shot.

[robinsonb5]

Originally Posted by SamBushman

Would bad ram chips cause the serial output I'm seeing?

I don't think so, no - the only way I can imagine bad RAM chips causing scrambled DiagROM serial is if they're failing in such a way that their output drivers remain on when they shouldn't. If a short or other fault were causing the _CAS signal to get stuck that could do it, too.


However, I still think you have a datapath problem. The four chips U10 - U13 are responsible for connecting the CPU's bus to the chipset bus. You've already replaced U12? I think my next step would be to replace U13.

[SamBushman]

@robinsonb5 Thanks for the confirmation. Much appreciated

I've ordered replacements for U11 and U13. Those should be hitting my mailbox Monday (I was hoping to update the thread with the results of my work, but the mail is slow ). I plan on doing the swaps then.

[SamBushman]

https://drive.google.com/uc?export=v...eTMIp5ZayAjoeX

Just replaced U11 and U13

https://drive.google.com/uc?export=v...kjuBxzKNQUJMCE

HOLY

CRAP!!!!!

We got serial characters that make sense! We have a picture!!!!! SWEET!!!!!

Ok, so I intend to run tests and verify the chips and whatnot work. HOWEVER, a new problem seems to have presented itself. After letting the machine run (just sitting at the DiagROM menu for a bit) the screen started to flicker (almost as if it was losing sync). I felt the chips and only Agnus seemed even a little warm.

I power cycled the machine, and same thing.

I let the machine sit for a while (and Agnus cool), then turned it on again. The picture was stable.

I bought some heat sinks for the custom chips already, however I wouldn't expect Agnus to misbehave on it's own like that.

Any thoughts on where to start looking?

[wazza]

Have you reseated Agnus and given the socket a good clean? That corrosion gets everywhere!

[DigitalKeeper]

Hey - Sounds Good

Have a look at the Agnus Socket and Agnus itself - coz of Contact Problem's

Otherwise - as i sad before - some Agnus have a Failure Problem with Higher Temperature

Cool down or change with another one...

[SamBushman]

I've taken a look at the socket. It looks pretty clean. I sprayed some DeOxit in there and used a toothbrush to clean it a bit. Reseated Agnus and tried again. Still eventually gets unstable.

Also put a few heatsinks on Agnus, no difference.

I think it's time to replace Agnus.

[SamBushman]

Well, I'm an idiot. I was using a GBS-8220 as a scan doubler. I'm pretty sure my display issues were due to that. Plugging the Amiga into my Framemeister with a SCART cable shows a rock solid image all the time. So on the plus side, Agnus is likely ok.

All my memory tests good and I've even been able to play The Humans! YAY!

But I can't just have a win......the floppy drive stopped detecting disks. It no longer clicks when waiting for a disk, and the Amiga is behaving as if no disk is inserted at all times. This behavior started while I was attempting to prep a hard drive. I failed to read a file from disk, and attempted to eject and re-insert the disk. From that point on, the drive stopped detecting the disk was in at all.

Drive heads have been cleaned, all rails re-greased, drive cleaned of dust, Write-protect and disk detect switches have had a little Deoxit sprayed on them. The Multimeter tells me the switches are working properly and we are getting proper voltages to the drive. I've reflowed the data and power connectors. No change :/

Last idea I have is to recap the drive. Capacitors are on the way, but it will be a couple days. I really hope this works.......

[mtb]

Congratulations for the fix.

Why don't you use a Gotek? We did it recently and it's really really nice! -> https://compvter.blogspot.com/2020/0...-su-amiga.html

[solarmon]

Check the floppy power and data cables are connected correctly and the right way. What is the status of the floppy LED - does it come on at any point, or is on all the time?

Try swapping the CIA's to see if that changes anything.

[DigitalKeeper]

That is a Failure of the "Micro Switch" for the "Insert Disk" Signal

Ya can desolder the Switch, open it, clean the Silver Plates inside, coz they are "Black" with Oxyd, then re-assemble it and all would be fine.

WARNING -> the Drive Heads are extremly Sensitive!!! Don't push, pull etc at the Heads

The "other thing" -> leaking capacitors in the drive

[SamBushman]

Why don't you use a Gotek? We did it recently and it's really really nice!

I mean, I could. I do have a bunch of Floppy Disks I'd like to use though.

Check the floppy power and data cables are connected correctly and the right way. What is the status of the floppy LED - does it come on at any point, or is on all the time?

Try swapping the CIA's to see if that changes anything.

I've verified proper voltage is getting to the power port on the Floppy Drive using the DC voltage setting on my multimeter. I trust the floppy ribbon is connected correctly, as the drive was working for a period of time before it stopped.

The LED does not come on sadly. It remains off the entire time.

Swapping CIAs is a good idea! Just tried it and sadly no change in behavior

That is a Failure of the "Micro Switch" for the "Insert Disk" Signal

Ya can desolder the Switch, open it, clean the Silver Plates inside, coz they are "Black" with Oxyd, then re-assemble it and all would be fine.

That's a good idea too! I have already sprayed Deoxit on the switch and verified the switch is shorting the expected contacts and those signals reach the connected board using the multimeter. That gives me confidence that the switch is operating correctly.

The "other thing" -> leaking capacitors in the drive

Yep, this is my current guess. I'll be pretty stumped if replacing the caps makes no difference.

[robinsonb5]

Since you have diagrom, try the drivetests menu there - see if you can get any signs of life out of the drive.

Using your logic probe, see if you get anything on CIA-B pins 10 and/or 11 when you use the step in / out options.

[SamBushman]

Got my new caps in. Recapped the floppy drive. No change in behavior. Tomorrow morning I plan on checking the pins robinsonb5 suggested.

[SamBushman]

Ok, so running the drive test I see a couple things:
1. The on-screen status of Disk-in and write protect don't seem to be valid all the time.
2. Whenever I step in or out of tracks my logic probe detects the changes as I'd expect (high to low for the _DIR signal and a chirp as each step is taken). The drive take no action though.
3. Using the multimeter, it appears that the appropriate pins on the CIA are making a connection to the pins on the data connector of the drive.

I'll go ahead and double check that the other connections between the EVEN CIA and Floppy data are connected. Otherwise I'm questioning if something I'm not thinking of has failed on the drive.

[robinsonb5]

Is the Sel signal reaching the drive?

[SamBushman]

All appropriate pins on the 2 CIA chips, Gary, and U36 appear to be connected to the data lines on the Floppy connector (as well as the row of grounds). Any suggestions on what sort of signals I should expect on these lines as I run through the drive test?

[robinsonb5]

Originally Posted by SamBushman

All appropriate pins on the 2 CIA chips, Gary, and U36 appear to be connected to the data lines on the Floppy connector (as well as the row of grounds). Any suggestions on what sort of signals I should expect on these lines as I run through the drive test?

This might help: http://amigadev.elowar.com/read/ADCD.../node018E.html


But basically make sure you're seeing some activity on the SEL pin on the drive when you do the drive tests, because it will ignore all the other signals if that one's not asserted.


Also, if you have any random PC floppy drive to hand, you could try that to eliminate the drive as your problem. A PC drive won't work without modification but it should at least click when empty, and spin up if you reset the Amiga while a disk is inserted.