• Please review our updated Terms and Rules here

Testing and/or troubleshooting Model III floppy drives

N

natevw

Member
Joined
Oct 22, 2019
Messages
34
Location
Richland, WA (USA)
Hello, I am trying to get the disks working in a Model 3 which I bought from a friend as a child, and recently brought home from my parent's house.

As a kid, I never had disks for it until a kind soul mailed me one all the way from Australia! Much to my disappointment at the time, I couldn't get it to work and soon gave up. Now that I have a little more patience, a little more dexterity, a little more electronics knowlege and a little more budget I would like to get to the bottom of this! [Though some emphasis is deserved on the "a little" on all preceding accounts ;-]

I had ordered some X1 capacitors and so tonight I opened up the case to replace them and try refurbish the floppies, but noted the following:

- there is a single power supply, labeled "Compaq Computer Corp, Copyright 1987" with some very aftermarket looking cabling
- I pulled a floppy drive out, thinking I would try moving things along the rails like I had seen in a YouTube video. But nothing seemed readily accessible and so I figured I should do some more homework (at least re-watch https://www.youtube.com/watch?v=-p2GkL9fjV8) before fussing with them.
- Powering the computer on with the drive re-connected to power, I did notice that neither of the 50Hz/60Hz markings seemed to stand still under a newer incandescent bulb.


I have a Model 100/200 floppy that I threw onto an (otherwise-unrelated) order of Model 100 parts from arcadeshopper. When I put that in, the machine fails to boot at all — the disks spin for a while, then turn off, still blank screen. If I leave the door open I do get a "Diskette?" message. When hitting Break+Reset to boot into ROM BASIC, `OUT 244,1` turns the LED on for the bottom drive. Both spin (?). `OUT 244,2` turns the LED on for the top drive. Again, both spin. No other noises, i.e. besides the spinning there's no clicks or clatters like e.g. our Apple /// used to do.

I need to dig around and see if I can find the TRS-DOS disk in some boxes, but it didn't work ten years ago and I'm not hopeful it will again. How might I proceed in troubleshooting? Do I risk damaging disks by trying them? Are the drives relatively robust if I open them up and try clean/move the mechanisms manually?

As far as making disks, is the process in https://www.youtube.com/watch?v=JM_eW7h07ew pretty foolproof if I pick up any random [update: 360K as stated in video] 5.25" drive off eBay and chuck it in an old Windows XP machine? I don't want to be troubleshooting the Model 3 drives if the disk itself isn't "known good".
 
I'd start off by cleaning and lubing the rails and using a head cleaning disk. I've had the setscrew for the stepper motor loosen up on many Tandon and Texas Peripherals drives, so check those. It's behind an open square in the frame and may have a clear plastic window over it.
As far as adjusting drive speed...the strobe effect isn't going to work with an incandescent bulb. You need an OLD fluorescent bulb. Anything modern like CFL won't work for some reason. When I was a kid I used to use my fluorescent desk lamp but that disappeared over the years. I found that a small neon sign from my game room works great.
 
The first step should be to use the head cleaning disk mentioned above. It is the most like procedure to achieve results without the danger of damaging any hardware components.

Make copies of any disks you will use so that you won't run the risk of damaging the original or last one.
 
What is the model number of the floppy drive? If you have an after market PSU, the drives may be third-party, too. There's a lot of info on the web, but it's very much vendor and model specific.

You can make real floppies from images using an old DOS-based machine (no higher than Windows 98 ). I use a Pentium 233MMX machine and a TRS-80 emulator written almost 20 years ago by David Keil. Depending on the images, there are different software tools that can be used.
 
Last edited:
Stone said:
The first step should be to use the head cleaning disk mentioned above. It is the most like procedure to achieve results without the danger of damaging any hardware components.

Make copies of any disks you will use so that you won't run the risk of damaging the original or last one.
Click to expand...
And don't forget to make a copy of the cleaning disk for future use ... ;) ok I´m making jokes :cool:

There are many videos on Yt how to clean a disk drive. The 50Hz/60Hz markings are important if a drive isn´t really working. Usually the disk drives are reliable. Try to use a boot disk after cleaning the drives. If everything is ok then there is no need for checking the rpm. The disk's head can be moved manually during the cleaning process with care but without power on.
 
Thanks for all the tips so far! I've ordered both a cleaning disk (which I suppose I don't necessarily need if I have the drives apart anyway) and a fresh bootable disk off of eBay, and I hope to carefully open the machine again this week or next and give it another go.
 
Got back to this project tonight, took the machine all apart again and opened up the floppy disks this time! Here's what I know:

The drives each have a "Texas Peripherals" sticker on the frame bottom, and 99-5050-001 Rev C Tandy Corp c 1980 on the mounted PCB. The head mechanisms were a bit scratchy on their rails but not seized or particularly griming. I swabbed them down a bit and added some fresh Wahl clipper oil that I had handy. They are now less scratchy, though not silky smooth.

The top drive had one of it's little white door latch pins missing this time I checked it. The little grabber it would have clipped into was also cracked. I superglued that and stuck a toothpick in place of the for now but haven't dared test the joint yet. (I'm usually 50/50 at _best_ for superglued plastic holding up more than 0 seconds of real usage…)

The bottom drive had a more interesting problem! I noticed it had a gaping hole above the drive head where the other one had an insert. AFAICT this is a plastic piece with some sort of felt attached, and I imagine this must be to push the medium up against the head. So maybe we can't expect that drive to work. (Oddly this drive has two metal posts for its door latch, rather than one metal post plus plastic pins?!)

So I took the [formerly] top drive, replaced its latch with the sturdier one, and installed it in the bottom slot. The [formerly] bottom drive back in up top, with its latch still flapping in the breeze for now. I also re-seated all the jumpers around the PCB on both drives just for kicks. But alas:

Still nothing :-(


Questions:

- what are the chances I could find the missing part? does it have a name? I suspect parts are rarer than the drives themselves though?
- am I correct to assume that the cable position will select the primary drive? is there any other termination or configuration that I'm overlooking in order to swap the two drives?
- should I be hearing the head grinding and clicking a bit before and/or after I close the door on a disk?

From the video in my OP, I learned that I should leave the door open at first, then close it when I get the "Diskette?" prompt. With no disk, I get no prompt just a black screen. With a disk in but latch open, I get the "Diskette?" prompt. After I close the latch, nothing changes…it just spins and spins for a while with no apparent attempt to sync up with the tracks. It doesn't click or grind or any sort of "seeking" noises, just spinning. What should I test or try next?
 
As far as the Drive Latch is concerned, there are folks that have designed the Door Latch Parts
for Texas Peripherals and the Tandon TM-100

You just need to locate someone with a 3D Printer or Resin Printer to print some SPARE Door Parts.
You can get some Brass Brazing rod in 1/8" and 3/32" sizes and cut some Brass Rod to make a full length
Hinge part. Toss those small White Nylon pins in the TRASH, and insert the full length Brass Rods
for a Hinge that will outlast you. Most older Floppy drives came out with the Brass Pins, at least the
Tandon TM-100 series did.

You can wipe down the Guide Rods that the Head carriage assembly slides on with A Cotton Swab
dipped in Alcohol. I use a Motorcycle Lubricant called Dri-Slide (which doesn't attract dirt) to lubricate
those Head carriage assemblies. (It's also good to use on flatbed scanners). Any Motorcycle shop
should carry the Dri-Slide, and a bottle will last you a lifetime.

I have a Basic RPM Test Program that works on the Model 3, but you need to load it from Basic.

Probably the best way to test the Floppy drive is to remove it, and connect it to an old DOS Computer.
Then download Dave Dunfield's Imagedisk and use the Utilities to check the RPM, and step the
head back and forth as a test.

Cable Position - Some of the TRS-80's had Pins removed from the Cables to make sure you plugged
the right floppy in the right position. But, the Drives that are Internal need to be strapped for Drive Select 0
and Drive Select 1. That is typically a punch out header strip that is located on the Floppy. I've removed
those and installed Dip Switches, so it makes it easy to change positions. The Model 3 and 4 does not
use the Terminators on the (Internal/External) Floppy's. But, typically all other computers use a Terminator
(DIP Package) on the last Floppy on the End of the Cable. I know that is true for the Model 1.


Thanks.

Larry
 
Last edited:
Appreciate the door latch tips! To be clearer, the "missing part" I am concerned about is this:

> a gaping hole above the drive head where the other one had an insert. AFAICT this is a plastic piece with some sort of felt attached, and I imagine this must be to push the medium up against the head.

I didn't quite dare take the insert out of the more complete drive for fear of snapping it, but it looks like more than just plastic. Has anyone had success replicating that part?
 
Small update: last night I replaced the cable between the main board and the disk controller. The old one was definitely decaying but after I took it out did have continuity on all lines. No improvement, no change with the new cable. Get the "diskette?" on boot followed by black screen soon after I close the drive doors. No head seeking movement/noises.

Unless someone knows a common cause for that symptom, I guess the next step is to start studying the service manual for all the details of how these drives and controllers are supposed to work. I do not have any old DOS computers handy although possibly a Tandy 1000 which might be a fun combo if I can find it in working condition itself. I wasn't sure how widely compatible these drives were but am getting the impression they were quite similar at the interface level at least.
 
I have been able to confirm the following general background information:

- the closest thing these drives have to selection jumpers are some exposed traces which are currently all intact
- the drives respond to DS on all four lines (10/12/14/32), the cable itself has "pulled pins" to select the drives
- the Shugart cable interface is a bit different than the later "PC standard" and only has one motor enable line

So to answer some of my earlier questions, it's expected that the motors for both drives would run whether I do `OUT 244,1` or `OUT 244,2`. And swapping the two drives is also supported without needing any further configuration due to the cable select design.

Finally, I was able to exercise the drives using a SuperCard Pro [a USB floppy drive controller used for flux-level] last night. It does not really support the Shugart-style drive control signals, but by taping over the DS0/DS2/DS3 pins on both of the drives I was able to use them — one at a time — as the "B:" drive where the Motor On signal pin matched. The results:

- the SCP drive utility sees both drives spinning at very close to 300rpm when requested, so their spindle motors and index sensors are working
- the head stepper motor and zero track sensor seems to be working fine as well, the SCP app's test routine is able to detect a maximum of 43 tracks on both
- one of the drives (originally the top drive as I found it) is able to image a disk with the expected results — in short, THE DRIVE SEEMS FINE!
- as I suspected, the drive with the missing part does yields ± blank images since the media does not get pressed against the sensor properly

But now what? I have already tried the WORKING drive in the bottom slot of the Model III itself and it does not read disks. Is it possible for the FDC controller to fail only partially — spindle control working, but stepper signals and/or sensors not?
 
Trying a shorter update and with no links to see if it will be approved by moderators this time:

- The PSU cables are fine 12V/5V to the drives
- One of the drives isn't reading data due to the missing "head load pad"
- But the other drive is fine as tested via a SuperCard Pro floppy controller

Since the machine won't boot off the good drive, I'm starting to suspect it must be the floppy controller but it's certainly not completely dead. Is that likely or is there something else to test next?
 
Did another round of troubleshooting tonight, this time with a Gotek floppy emulator running the FlashFloppy firmware.

On the FlashFloppy USB stick were 8 images, four different Model III disks captured via SuperCard Pro and converted into two different formats (HFE and IMG) with the HxC software. I had previously tested the drive emulator itself by reading back the fake images into SCP which worked fine. So essentially I have a "drive" with perfect alignment, reading "disks" that show no errors in the HxC visualizer thing.

And my Model III still won't boot from disk!! :-/

Pretty much the same story with the floppy emulator as with the real unit: it finds the drive, spins it and says "Diskette?" and then does nothing more when I insert one. So tonight I let the beast take over my workbench again and took out what felt like 50 screws to get all the way to the disk controller board. Made sure the cable from the motherboard was seated properly, etc.

The only thing visibly amiss is that there is a power jumper on the FDC board with four pins, but the lead connected to it only has three wires. Is that to be expected?

What else might I test here? There is some theory of operation in the service manual, but I don't see any troubleshooting guide or notes about what the TP test points on the board should read. I'll leave it apart for now but I'm kind of running out of ideas here… searching around the web it doesn't seem like the FDC card is particularly prone to any sort of failure but afaict mine is not doing its job :-(
 
A few more discoveries tonight: using the ROM version of BASIC, I am able to use INP/OUT to talk to the FDC and it responds generally as expected. I am able to step the head of either drive, return to track zero, etc.

I am not sure if BASIC can keep up with a sector/track read but experimenting with that will be my next step. So far just getting all zeros from `?INP(243)` so I may try attaching a logic probe to the cable to see what signals the FDC *should* be reading. Still haven't found a technician-style reference for using the test points on the controller board.
 
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
 
Appreciate the tips, Larry! There are three trimpots on my interface board, I won't touch them.

There's a helpful troubleshooting guide in the Sams Computerfacts for the TRS-80 Model III (CSCS5-B p. 7, on page 95 of the PDF) for "Will not read". While I don't have an aligment disk, it was still helpful in figuring out where to even start looking. There should be waveforms on pins 26 and 27 of the big main IC (U7).

Well on mine, I have a signal on pin 27 but pin 26 is dead. Looks like this is the RCLK signal and between the Sams guide and the official service manual it looks like this should be a clock signal generated in a PLL fashion via some other chips. I'm printing off the schematic to study a bit more since the Sams guide gets rather vague from this point: "If the waveform is missing at pin 26 of IC U7, check the logic readings on pins 1 thru 20 of Support Logic IC (U11) and pins 1 thru 16 of VCO IC (U14)."

That's a lot of pins to "check"… for what exactly?? :p This is where the theory of operation in the service manual will actually come in handy after all!
 
Heh, it looks like those three trimpots are related to this exact part of the circuit! Must. resist. urge. to tinker with them after all….
 
I did end up adjusting R7 until TP12 was reading 1.40V (it started around 1.54V) and adjusting R6 so that TP13 read as close to 2.0MHz as I could (when it hit 1.99962MHz I called it good).

While I run a read track loop, the WD1691 chip (U11) is getting:

WG and VFOE* around 240mV
VCO around 2.0MHz (Vmax 4.80V)
DDEN* depending on `OUT 244,1` [high] vs. `OUT 244,129` [low]
RDD* shows a sensible signal reading an arbitrary disk, i.e. periodic pulses down to 0V

But the RCLK pin is always low, and the PU/PD just stays at 1.40V. How likely is it that the WD1691 itself is bad?
 
You must log in or register to reply here.
Back
Top