SE/30 Floppy Troubles

[Ike]

I've recently acquired a very nice SE/30 with a great looking Logic board

I decided to wait with my conquest topic about it until I recapped the logic board.

Did successfully wash the board, did replace all the smaller caps while still leaving the two bigger caps in place, without one ripped or lifted pad

Now everything works... but one thing, the floppy drive.

It does read sometimes, it does write sometimes, but...

If I try to initialize a 800K disk it just says "initialization failed".

If I try to initialize a HD 1.44MB disk it gives me a bomb with sometimes a visible error about 'A math coprocessor not' being 'installed' or 'Address error'...

Now i know the Bourns-networks are vulnerable on these machines, is there a way I can test them without de-soldering?

I see no corroded vias, corroded chip-legs, solder blobs, pieces of metal or any other dirt on the board.

Am I missing something? Am I too stubborn by leaving the 220uf and 470uf caps in place instead of also just replacing them?

Would be fantastic if you could point me in the right direction!

 

[uniserver]

how bad was the cap goo leakage?

I'm thinking yours is pretty much doing the same thing as insaneboys.

there are some data lines that go to the SWIM and they get messed from rotted via due to cap goo.

http://maccaps.com/MacCAPS.COM/Schematics_Bonus_Info_files/SE30%20Schematics.zip

could you do me a favor, check out the schematics and use your meter and figure out what connections are toast?

then post your results? i think your fix will help many.

 

[Ike]

The capacitor leakage was present but not very aggressive in any way.

Some corrosion is visible on the chip legs and vias but nothing out of the ordinary as far as I can see.

Just checked all the lines going from the SWIM to the ribbon connector, those are all fine

Now going to check the rest of the bus to the SWIM... lets hope I find something there...

 

[uniserver]

also here was a good post on how to check the bourns filter

viewtopic.php?f=3&t=21096

I've just been through the Bourns filter problem on my 512Ke.
Fortunately, it is easy to test the filters without removing. With power off, just put a multimeter across each of the inner 8 pairs of pins, in the 200k resistance range. If it fluctuates up and down (filling the cap) then settles at around 50R on each cell, each RC network is OK.

New filters seem to be about $3-8 each, depending on supplier.

The FDD pinouts are also very picky and worth checking for each model that you plan to attach. Watch out for -12v on pin5/9 and whether your device uses that or should be isolated from it. I blew a filter finding that out the hard way.

Rick

also i found these for 98 cents 87 cents if you get >50.

http://www.onlinecomponents.com/dale-mdp1603470g.html?p=11218616

per my photo

viewtopic.php?f=7&t=18026&start=50

I doubt the network filter is the issue here. - by the errors you were getting.

i found that when these network filters go bad ( bourns ) you get random annoying messages saying floppy in drive is not initialized would you like to format it? ---- and there is no floppy in the drive or like the External will work fine but the internal will not. or the other way around.

 

[Ike]

Just finished measuring the SWIM. The Address bus lines and Data bus lines all connect from the legs of the SWIM chip to the corresponding pins of the PDS.

Thanks for the information about the filter networks! Just tested the it to be on the safe side, its completely ok. Looks like it protects the external port from frying the complete logic board when something is attached wrong like RickNel's case.

Just tried writing to a 1.44mb disk again, that works, reading off the disk mostly works fine...

just the formatting (especially on the 1.44MB HD disks) just causes instant bombing with a different 'bus'-related error before even writing one thing to the disk.

Would changing out the +5V decoupling capacitor C11 make any difference for the stability of attached drives?

...Whats next to test, any ideas?

 

[uniserver]

yeah you should replace the 220uf + 470uf caps as well… and all the 47uf 16v caps, + that 1uf 50v cap too.

 

[uniserver]

Also make sure your floppy drive is clean and oiled.

 

[Ike]

Okay I'll replace the big caps too, other ones are already replaced

Just saw your thread Uni, is that insaneboy's board you were talking about earlier, the symptoms are completely the same :O

If not, did you fix it?

EDIT: oh and yeah i lubricated the drive, it operates very smoothly.

 

[uniserver]

no, he did not want more then just a re-cap.

also i'v had one other board that did this too, they also did not want more then just a re-cap.

so it is a little bit of un resolved issue.

here is the first board with this floppy issue

viewtopic.php?f=7&t=22252

this one i even changed the SWIM chip…

how about even more steps back… ? the lines from the GLU chip to the UH7?

pretty sure some pass near the pesky leaking caps.

 

[Ike]

At least I can find some comfort in the fact that the SWIM itself is probably good.

When I get home I'll do some more testing, surely this can't be impossible to fix!

 

[uniserver]

i agree!

 

[Ike]

Starting to lose the hope here :-/ :?:

Just checked all pins of the UH7 chip to their locations, all is good there.

Checked the ram banks to be sure, they all connected as they should as far as I know.

What does the mac DO or NEED when formatting, that is DOESN'T need with normal reading/writing?

 

[uniserver]

here is something interesting.

this guy going through issues with floppy for an amiga 1200,

but maybe it will help us figure out our issue?

'Amiga Floppy Strangeness'

 

[genie_mac]

Hey Ike,

just a few random thoughts I had when looking at Uniserver's problem:

I guess you checked everything around the SWIM that has to do with writes: write (WR), write request (WRREQ), FPH3 (write strobe). I'd also just make sure that there are no shorts between any pins on the swim.

But I think you've done most of that already, so I reckon the write signal doesn't make it to the floppy drive. Did you check everything around UH7 (especially pin 4 and 16 to swim /floppy connector). Also check the pins on UH7 that go to the GLUE chip (page 3 of the SE/30 schematics that can be found on net) because that's responsible for inverting that WR line...if any of those are faulty could be a prob.

Finally did you check around L22 inductor/C67 cap and R30 resistor (around SWIM and floppy connector) as they are all part of the WR line...long shot though!

Hope you'll get it working

 

[bear]

If you didn't recondition the analog board, stop until that's done. Save yourself a lot of ghost-chasing.

 

[Ike]

-first of all, I'm an idiot with a scope since i have never used one... but I'm trying my best here " -

If you didn't recondition the analog board, stop until that's done. Save yourself a lot of ghost-chasing.

Thanks for the input but to me that does not make sense to me. To be sure I tried the Se/30 LoBo in my known working SE which gave me the same result.

Since the PSU outputs normal voltages needed for the floppy drive, I have no video problems whatsoever, and it does not make a difference with the hdd in the machine or in an extermal enclosure with it's own PSU I would not expect the analog board for this to be the problem.

But I found something very interesting today, let's call it a minor success!

I borrowed a digital multimeter-device from a friend of mine, that also has an oscilloscope with the output shown on a pc via software. It's slow - as expected - and annoying to use, but it does kinda give me the impression of what was going on. Results are in:

The floppy transmits and receives data from/to the SWIM, you see every pin actually do that on the scope.

Here is the amazing part: as soon as i kept the probe between ground and pin 4 of UH7, the WR-line, the disk started formatting as if NOTHING was wrong.

Hey Ike,~Did you check everything around UH7 (especially pin 4 and 16 to swim /floppy connector). Also check the pins on UH7 that go to the GLUE chip (page 3 of the SE/30 schematics that can be found on net) because that's responsible for inverting that WR line...if any of those are faulty could be a prob.

Finally did you check around L22 inductor/C67 cap and R30 resistor (around SWIM and floppy connector) as they are all part of the WR line...long shot though!

~

So this was kind of the right answer... the issue I have with my situation now is: The continuity checks just fine!

Why does prodding around with my scope solve the problem temporarily?

 

[uniserver]

solder a wire wrap from a to b.

i think you have found it… a bad trace my friend.

 

[genie_mac]

Hmm interesting! Are you sure the continuity between pin 2 SWIM and pin4 UH7 is good (i.e. as close to 0 Ohms as possible)?

I suppose the scope adds a bit of capacitance to the circuit, so might act as a filter or maybe create a bit of a delay. Is the signal on the scope clean or is there lots of noise? Might also be a ground issue on the IC, did you check for good / noise ground and good power?

I guess you could always put a cap of same capacitance as the scope leads in the circuit to see if that fixes it, but it shouldn't need it so there must be a problem somewhere...

Keep up the good work

 

[Ike]

It's just weird. It's actually the WR● line coming from pin 4 of UH7 to pin 2 of the SWIM... the issue is that my multimeter gives close to 0 ohms... Yet it just doesn't want to work without a direct lead between the two pins, or the scope between any of the above pins and ground...

Grounding on UH7 is okay, pretty sure I checked that as well.

Such a shame. i really hope to just get around the final issue, without having to resort to the spaghetti-wire-method.