Testing and/or troubleshooting Model III floppy drives

[natevw]

oh quite possible

Thanks, I ordered a replacement from Eastern Europe so hopefully will have an update in a few weeks. It's labeled as a WD1691-PE versus the WD1691V in mine but hopefully that doesn't make much difference.

 

[natevw]

Success! On the controller front at least.

The replacement chip arrived yesterday and I put it in tonight. It booted right into the images off the Gotek/FlashFloppy in the Drive 0 cable position and for the first time ever I had something besides Cassette BASIC running on my stalwart old TRS-80

No luck booting from an actual disk in the one ± working drive yet, but a simple BASIC test program is able to at least see some sort of data stream coming off of it. Maybe just a matter of alignment now. I'm debating whether to just mount the Gotek permanently in this, or plow through and get one/both drives working again; probably gonna go for the latter.

 

[shank]

look for the model III diagnostics and run the drive alignment program. you won't have the alignment diskette so you can't check head radial (which isn't your problem, anyway, if your drive isn't even stepping), but you should be able to check for index pulses and track zero indications. i admit that i never ran these diags without an alignment diskette so YMMV. don't do any alignments, just check signals... that may show you a path forward. with the additional info listed below, you should make sure that you have the proper drive handshake/status signals making it out of the drive and to the FDC. you should have a scope for this, but at a minimum you can construct a logic probe inexpensively. it won't give you timing, but should show you activity. lots of plans on the internet and they are even available from that great internet retailer...

if you haven't downloaded the MIII Technical Bulletins, now would be a good time to do so. they'll give you some info about the machine. as well, Notes and Jumpers will provide additional information. more information is better! Also, service information for those drives is available... either RS or Tandon. Should be a theory of op in there. if you're not familiar with how the drives work and what they expect be sure to find a good theory of op before poking around.

On the FDC alignment: so it's been a zillion years since i did one, but it's not difficult if you have all the test equipment. i don't recall if the alignment diskette is required for that. i don't suggest you do this, just FYI.

i've attached a Model III file from TS training that gives a tech-centric overview. ignore the weird formatting--this was before we had nice GUI word processors... anyway, there's a little bit on the FDC and drives which might help.

 

[natevw]

look for the model III diagnostics and run the drive alignment program. you won't have the alignment diskette so you can't check head radial (which isn't your problem, anyway, if your drive isn't even stepping), but you should be able to check for index pulses and track zero indications. i admit that i never ran these diags without an alignment diskette so YMMV. don't do any alignments, just check signals... that may show you a path forward.

Thanks, good stuff to review. Note that my drives ARE stepping though. For example, if I start the drive on a different track it does seek to track 0 before it tries to boot up. This test program tweaked from the SAMS Computerfacts for the Model III works fine to move the head back and forth:

10 OUT 244,1: OUT 240,3
20 FOR T=1 TO 700: NEXT T
30 OUT 244,1: OUT 243,40: OUT 240,19
40 FOR T=1 TO 700: NEXT T: GOTO 10

For now, I've taken them out though for separate diagnostics via a SuperCard Pro controller when I get to it. Also grabbed a third Tandon-style drive, it's from an IBM and has dual heads so not exactly the same but the eBay seller claimed it was tested and functional so that might give me something to compare as I try diagnose any alignment or other issues with the drives themselves apart from the TRS-80 controller. Meanwhile I've been using a Gotek+FlashFloppy in place of the original drives in the Model III.

It boots fine from images on that. Or I should say, it "booted" fine from images on that initially but after playing around a little even those seem to have either gotten corrupted, or something in the controller circuit is degrading again. E.g. after trying to save a BASIC program to my TRSDOS image that one no longer boots past the time/date prompt. And an LDOS floppy image that initially worked fine afaict now goes crazy and starts seeking up to like track 220 when I try to get a listing via DIR. (Again this is on the Gotek so harmless to step that far but WHY???) Hopefully just something I typed wrong and not the new chip

 

[Patrick.B (TTR)]

Unzip slap this on your Gotek (just theses two files) and boot computer does it boot ? then it was file corruption of bad image. View attachment FF LDOS v531 Boot Image and CFG file.zip
and I would treat the gotek as drive 0 to be like a floppy on aTRS meaning do not plug it in until the system is turned on.
so turn on, insert USB, press reset, it is possible to "zap" a USB as well as it zaps a floppy when turning it on with media inside.

 

[Patrick.B (TTR)]

you know I was thinking about the "track" alignment method, would it not be easier to use super utility do a bulk erase (or use a demagnetize tool) then special format just track 20 and the use fdd software to lock on that track and then adjust with scope ? would make "other possible track signal" to not be seen ?

 

[shank]

you know I was thinking about the "track" alignment method, would it not be easier to use super utility do a bulk erase (or use a demagnetize tool) then special format just track 20 and the use fdd software to lock on that track and then adjust with scope ? would make "other possible track signal" to not be seen ?

sure, if you can do that. while you're at it you could write a track zero and use that for track zero alignment--if anyone really wants/needs to do it. you can grab the RS drive diags and take a look at what we did. the alignment was done with a scope. in some drives the index sector sensor can be moved and aligned.. something else...

it was just a thought. i don't know how far down the rabbit hole collectors want to go with this. we did it for 2 main reasons--to make sure the drives were aligned well enough to do their job (good way to find non-alignment drive type problems, too) and to make sure that diskettes would interchange between systems. whether that's necessary now is the question.

thinking a little more about this... if you do a disk for head radial, you should go ahead and do a track zero, too. you really need both working in a drive. if track zero is out of alignment the FDC can get lost. worse is when the drive does a recal to zero and gets behind track zero. that can cause problems. FDCs are easily confused.

 

[Patrick.B (TTR)]

I agree, and as far as "diskettes would interchange between systems. whether that's necessary now is the question" as one that restores and sells , I would hate to sell a computer which drives are isolated to the point if someone sent them a diskette they could not even read it because ti wasn't properly aligned, not to mention the troubles I would have to have a "machine specific" set of floppies and not just use my pool of testing floppies. standardization (to the closest we can do it of course) does make the work of restoring/helping others (no point if I cannot send a set of 10 floppies with os/games to someone because my drives are not aligned to spec).

 

[phil.bischof]

I will ask a simple question, did you inspect the power cable that run from PSu to card ?

Patrick, would a faulty power cable allow the drive to come on and seek, but fail to read/write? I have a situation where my drive will totally seek from 0/40, but will not read or write. Keeps failing a DIR with Error 04.

 

[phil.bischof]

Whatever you do, DO NOT ADJUST any POTS on the FDC card. I don't remember if there are any on the
FDC, but if there are DO NOT ADJUST.

One quick test you can do is with Power OFF, GENTLY Slide the Head Carriage Assembly in towards the Spindle
until it stops. Then move it back about 1/2". Then Power UP and see if it steps away from the Spindle. You should
hear it softly tick, tick, tick as it steps away from the Spindle to Track 0.

With a Disk inserted are you getting the Index Pulse?


Larry

Larry, I'm working on a very similar issue here with my drive and when I do the above test, it does not softly tick, tick, tick. Upon powering on, it waits a few seconds and then very abruptly just returns to track 0. It's quite rough. Is that normal?

 

[ldkraemer]

Phil,
It could be a normal sound. The best thing to do is Unplug the Read Head Cable(s) and
remove the two screws holding the Logic PCB. You can then swing it out of the way to
access the Head Guide Rods. Wipe them down with a cotton swab dipped in Alcohol.
After that I drip about 3-4 drops of Dri-Slide a Motorcycle Cable Lubricant on the Guide
Rods. That will quieten then down lots, and make stepping easier for the stepper motor.
You can easily slide the Head carriage assembly full travel, then back to the center of travel
for the next test.

The Dri-Slide will not attract dirt/dust and will last for years. I use it on scanner rails too.
It's available at most Motorcycle shops.

Larry

 

[Patrick.B (TTR)]

Patrick, would a faulty power cable allow the drive to come on and seek, but fail to read/write? I have a situation where my drive will totally seek from 0/40, but will not read or write. Keeps failing a DIR with Error 04.

not likely no, Larry's suggestion above is very good, and if the issue continues I would look at the component on the board "the read circuit" from the schematics of the service manual of the floppy drive. maybe you have a blown part or dead cap or something.
if you don't have the service manual let me know what make/model of drives you're looking at. I should have it on hand

 

[phil.bischof]

@Patrick.B (TTR) @ldkraemer
Awesome advice guys, thank you! I did already clean and lubricate the rails, and running the BASIC program from another post where Larry provided it, my drive does seek out to 40 and back to zero, so it appears it must be something in the read circuit. The drive has TANDON stamped underneath, but also has a TEXAS PERIPHERALS sticker on there too. No other identifying info on the drive itself. The board on the drive is 99-5050-001 REV C. I have this attached service manual. IF there is another with more detail on the read circuit, that'd be great! Thanks guys!

 

[phil.bischof]

@Patrick.B (TTR) @ldkraemer
Well, I started down the path of checking the read/write circuitry and following just the first thing in my mind, I followed the wire coming of J1 back to the head itself. Flipped the drive over and removed the plastic cover to make sure it all looked good, and I could see what looked to be a wire just hanging free in there. Upon closer inspection, it was confirmed. One of wires that is connected to the underside of the head itself is broke off. No wonder it couldn't read/write.

There's no way in heck I'm fixing this thing, as the wire is as thin as human hair and then all wound up in a small coil that attached directly to the head.

Anyone got an extra Tandon Floppy head assembly? no worries if no as I'm sure those are pretty much impossible to find. Guess I'll just have to order another drive.

 

[Patrick.B (TTR)]

@Patrick.B (TTR) @ldkraemer
Well, I started down the path of checking the read/write circuitry and following just the first thing in my mind, I followed the wire coming of J1 back to the head itself. Flipped the drive over and removed the plastic cover to make sure it all looked good, and I could see what looked to be a wire just hanging free in there. Upon closer inspection, it was confirmed. One of wires that is connected to the underside of the head itself is broke off. No wonder it couldn't read/write.

There's no way in heck I'm fixing this thing, as the wire is as thin as human hair and then all wound up in a small coil that attached directly to the head.

Anyone got an extra Tandon Floppy head assembly? no worries if no as I'm sure those are pretty much impossible to find. Guess I'll just have to order another drive.

well I have quite a few drives in boxes. let me see if I have a Tandom TM-1 head for you. hit me in PM please as a reminder.

 

[Patrick.B (TTR)]

check your PM