Model 4 with DSDD issue: RatBastard strikes again!

[clh333]

Let's skip over the sad story about a Model 4 that was inhabited by mice for 20 years and cut to the chase:

After many labors I now have a working Model 4 GA that has several new or refurbished parts. One of the modifications I made was to swap the OEM SSDD drives and cable for DSDD. Drive :0 is well behaved and reliable; the machine boots and executes programs under OS control. Other functions have also been tested both with diagnostics and practical tests - i.e. connection to peripherals. Drive :1, a repurposed IBM 5.25 DSDD, appears to be operating correctly until it is asked to format a disk.

D:1 will respond to a directory command by listing the contents of a disk. It will invoke a program that resides on its disk. When I ask the OS to format a disk in D:1 the drive responds by initiating the format, laying down tracks from 0 to 39. When it then tracks back to track zero and tries to verify the tracks it begins to falter, marking track after track bad. Disks that are created thereby are unreadable and generate seek errors when accessed afterward.

I thought the disk head was dirty, but got the same results just after cleaning it. I thought maybe the disk or batch of disks was faulty, but I tried several brands of NOS DSDD disks with the same result. I tried specifying different parameters for the format, single-sided, double-sided, different step rates, all with the same result.

Disk diagnostic tests suggest that the signals to and from the controller are getting through, and the fact that the drive can find and execute files from the directory suggest that the track zero switch and stepper motor are operating correctly. I'm beginning to wonder if the write electronics might be weak and leaving an unreadable signal behind. If that were the case, how would I go about testing this hypothesis?

Thanks for your suggestions.

-CH-

DSCN1239.jpg

[lafos]

What is the model number on Drive 1? Do you have a photo of the jumper block?

If you have a readable disk that you can sacrifice, try copying an executable file from :0 to :1 and then see if it will load and execute from :1

[ldkraemer]

The first thing I'd suggest is making sure the Drive(s) RPM's are 300 RPM. There are RPM
Programs that will test that. If the RPM's are correct, then I'd suggest checking the Manufacture's
Specification for the Stepping rate. I know that in Montezuma Micro CP/M OS you can select
various stepping rates. (30, 15, 6, xx) If the Drive is a Tandon TM-100, open the drive door and
check to see if there are two short White Nylon Pins used for the hinge points. If they are present
replace those two pins with a piece of 3/32 Brass Rod, so the hinge point is full length. (That helps
protect the Hinge part from getting broken. If it's broken let me know and I'll send you one hinge part
and a 3/32 Brass Rod cut to Length.)

I wouldn't fiddle with any of the FDC Pots thinking you will be making it better. Those are preset and
don't need tweaked.

Remember that ONLY the LAST Floppy Drive on the end of the cable needs the Terminator (150 OHM).
For selecting the DSx numbers I insert a DIP switch so I can quickly select DS{0..3} as needed. No other
Switch (toggle) is closed.

I think some where in my stash I have a RPM program that I wrote for the Kaypro II, and modified for
the TRS-80 Models 3 & 4. Let me search for it.

OH!, I forgot to tell you to check the RPM Boards attached to the back of the Tandon Floppy Drives, for
Acid from the Caps that has leaked and ate up the PCB traces on the RPM board. I'd found several
RPM boards that have this problem. I ordered some copper rivets from the Internet (Sweden) to
rebuild the plated through holes, and used small wire wrap wire to connect the traces. It's not
pretty, but makes the drive functional again.

Larry

[clh333]

Thanks to you both for your suggestions. I am now left with no other option but to remove the cover as d:1 is no longer responding to any commands.

I'll report back when I have more concrete information.

Thanks,

-CH-

[clh333]

Case came apart and I retrieved the offending drive. It is an IBM-branded Tandon T100-2A. In accordance with Larry's practice, before installing I had removed the T.RES IC and substituted an 8-position DIP switch in the jumper socket that controls drive selection.

I tested this drive and two other similar DSDD Tandons with the SuperCard Pro this afternoon. Each was recognized - provided I had the DIP switch set properly - and spun briefly when SCP software utility scanned the bus for drives. I was able to run the maximum track test, so the controller board is not dead, but I was unsuccessful running the speed test. The drive activity light would flash briefly and then go out. The drive would not spin.

I removed the motor control board on one of the three and inspected the traces for signs of leakage around the caps. The three electrolytics showed signs of leakage: I removed them, cleaned up, tested continuity and replaced them, but after all that the speed test failed once again. The motor and spindle seem to move freely. I'll have to disassemble further to see if there is anything else worrisome.

But the fact that the drive spins briefly suggest another possibility: There has to be a feedback circuit based on spindle speed that governs the motor. It may be that this circuit is somehow inhibiting the motor's revolutions. More research will be necessary.

-CH-

[DeltaDon]

How many switch positions are closed on the DIP switch? I believe it should be two of the DIP switches in the on position. If I recall correctly for a Tandon you need a drive select closed (0-3 or 1-4) plus DIP number 8 (maybe 7?? - I need to dig out documentation) to spin the motor when selected. Else you get the no spin results you state.

[clh333]

How many switch positions are closed on the DIP switch? I believe it should be two of the DIP switches in the on position. If I recall correctly for a Tandon you need a drive select closed (0-3 or 1-4) plus DIP number 8 (maybe 7?? - I need to dig out documentation) to spin the motor when selected. Else you get the no spin results you state.

I had consulted the Tandon manual, which described the options for jumpers. Beside drive select (0...) the options are multiplex (for single-drive systems, not applicable here), power saving and head-loading of a solenoid, again not applicable. The drive formerly had only one option enabled, drive 1, selected with a wire jumper. And the M4 d:0 only has one option, d:0, selected, yet it is working properly.

I did some more research and found this topic has been discussed here as early as 2010. My thanks to Chuck(G) who referenced a Sam's Computer Facts for the T100 and a procedure for testing the motor servo control board. A VCF member named "per" found a couple of ICs at fault and was able to restore his T100 to function. I think that is the path I will follow next.

I would have thought that ICs either fail immediately or work forever, but I guess that is a mistaken assumption. Maybe these were Indian ICs, huh, Chuck?

-CH-

[DeltaDon]

Here's the list of jumpers for a Tandon 100:

Programmable Shunt Socket (1E):
HS - unused; must be cut
DS0..DS3 - Drive select
MX - multiplex; used only in single-drive systems when drive logics should remain enabled at all times. Must be cut in a multi-drive system.
(Spare)
HM - unused; must be cut
R50/R51 - Selects power save mode via 0-ohm resistor:
R50 - 3.8W saving at idle, by removing 200mA current to stepper motor when drive is not selected
R51 - no power save mode
W1 - Side Select:
Open - single-sided (TM100-1)
Closed - double-sided (TM100-2)
W2/W3 - Write Flip-Flop Control:
W2 closed, W3 open - Disables set and preset lines on write flip-flop only during internal N Write
W2 open, W3 closed - Dsiables set and preset lines on the write flip-flop continuously
W4/W5 - Write Protect Control:
W4 closed, W5 open - Write Protect Control responds to a write-protected disk
W4 open, W5 closed - Write Protect Control is inhibited
W6/W7/W9 - LED Control:
W6 closed, W7/W9 open - LED controlled by Drive Select
W6 open, W7/W9 closed - LED controlled with N In Use (J1-4)
W8 - Drive Select 3 Enable (DS3 disabled if open)
W10 - Door Lock (lock is disabled if open)

I was confusing the Tandon 100 with the Tandon 50 drives that has a Motor Sel. jumper on the same header as the Drive number sel.

[clh333]

Having tested the TM100 2A drives and found three with motor control issues I searched for more information on the subject. I found a reference to the Sam's Computerfacts diagnostic for the T100, in which the recommended procedure was to check input voltages (5VDC and 12VDC, passed), the motor winding's continuity (passed) and the resistance between pins 1 and 2 at #21 connect. This was specified as ~170 Ohms; the resistance I found was in the megaOhms to infinite range.

Tracing the circuit for continuity I came to some resistors which appeared not to be conductive. I checked other resistors nearby and found some were, others were not. I removed a few from the circuit to be sure of my findings, and found R3 and R10 non-conductive. This was a surprise to me as I expected to find the problem with one of the ICs or capacitors, but it seemed the smallest of the resistors had failed.

After more checking on line I find that it is possible for a resistor to fail, that when they fail they usually fail open rather than short and if this does occur it usually is the result of overvoltage somewhere in the circuit. So my suspicion of the ICs may not be misplaced, but this raises another set of possibilities: How many other components are likely to have failed if this overvoltage scenario actually occurred?

Any suggestions are welcome; I'm in tall cotton now.

-CH-

[rittwage]

Based on your original description of the problem, it seems like what I experience when I don't have the termination correct.