Western Digital MFM controller repair (IBM 5150)

[Skip94]

Hi all
Now I have my new 5150 functioning as it should and getting towards where I want it, having built an XT CF and Adlib clone in the last couple weeks, I want to turn my attention towards the MFM HDD and controller it came with.

The hard drive is a "Microscience" HH612 10mb drive and the controller is a WD1002-WX2-F002.
The drive spins up and when the dip switch on the back is set to "exercise" mode, the stepper motor runs back and forth through its range of travel smoothly with no funny noises.
However with the card plugged in, all I get on boot is the drive LED flashing a bit, then 1701 error.
Looking on minuszerodegrees, the 1701 error could be just about anything!
I've tried the simple stuff. All contacts are spotless, the drive is selected correctly on the dip switches, correct cable, reseated the socketed chips on the board etc.
Beyond that, I have no real idea where to start. This is my first time tinkering with MFM stuff, everything else I have done is IDE.
I know that the drive and controller have to be low level formatted together, but as this drive and card seem to gave been used together before, I would assume tgey should be formatted. Trying to open Spinrite gives me something along the lines of "Controller bad, do not use".
I would prefer not to LLF the drive, as it would be nice to see what was on it in terms of software, before wiping it. Also my gut feeling is that the controller is faulty, rather than the drive.
Does anyone have any thoughts on next steps I can take to diagnose the problem? One thought I had was to socket all the 74LS series chips, so they could be tested, but I'd prefer to track down the actual problem, rather than just start yanking chips out.
Sorry for the essay
Cheers
Andrew

 

[modem7]

.. then 1701 error. Looking on minuszerodegrees, the 1701 error could be just about anything!

The 1701 does inform us that the 'BIOS expansion ROM' functionality on the controller (only part of the controller) is probably fully functional, i.e.
1. The computer's power-on self test found the card's BIOS expansion ROM, was happy with the contents, then executed the initialisation code within.
2. The initialisation code within generated the 1701 error.

I know that the drive and controller have to be low level formatted together, but as this drive and card seem to gave been used together before, I would assume tgey should be formatted.

High probabilty, but not ruled out.

Trying to open Spinrite gives me something along the lines of "Controller bad, do not use".
I would prefer not to LLF the drive, as it would be nice to see what was on it in terms of software, before wiping it. Also my gut feeling is that the controller is faulty, rather than the drive.
Does anyone have any thoughts on next steps I can take to diagnose the problem?

Perhaps Spinrite asked the controller to do a self-test and the controller responded a negative answer.

You could see if that is the case by seeing if SpeedStor software indicates same. SpeedStor asks the controller to do a self-test via the menu navigation of: {Diagnostics}{Controller}

Or perhaps use DEBUG per [here], seeing what the responses are for functions 00h, 12h, and 14h.

 

[modem7]

I have created a DOS program named RAYXTMFM.
See what that reports.
Details and program at [here].

 

[Skip94]

Hi, thanks for the responses,
I got a little time this morning to have a play.
Speedstor will not let me even select the diagnostics menu, as far as I can tell it doesn't even think that there is a controller installed.
Running RAYXTMFM, it executed parts 1 and 2 correctly, then in part 3 it says something along the lines of cannot identify the BIOS and exits.
Running the debug commands, I get the results pictured. not sure I understand the result though I'm afraid.
Cheers
Andrew

 

[modem7]

Speedstor will not let me even select the diagnostics menu, as far as I can tell it doesn't even think that there is a controller installed.

On my IBM XT, I removed the hard disk controller. Executing SpeedStor 6.03, it reported at the bottom of the screen, "** ERROR ** Cannot activate disk BIOS".

I then disconnected the control and data cables from the controller. At IBM XT power-on, I saw a 1701 error, which was expected, because my controller's BIOS ROM did not see 'hard drive 0 is ready'. At start-up, SpeedStor 6.03 did not report an error. On SpeedStor's main menu, 'ManualSetup' and 'Exit' were the only options enabled, and if I chose 'ManualSetup', the only enabled menu option of those shown was 'Quit'.

Before I could do any further experiments (e.g. simulate I/O read/write failure on the controller), my IBM XT decided that it no longer wanted to play with me. So I'll have to fix it later.

Running the debug commands, I get the results pictured. not sure I understand the result though I'm afraid.

The first screen shows you executing function 0 (controller to reset itself). On exit of function 0, I see the carry flag set (error) with the AH register set to 05h. 05h equates to "Reset failed".

No hard drive should need to be attached for that function to pass. That for sure is why Spinrite displayed something along the lines of "Controller bad, do not use".

Also my gut feeling is that the controller is faulty, rather than the drive.

At this stage, that is where the finger points to.

I am sure that if you had another suitable computer, you would have tried the controller and drive in that, mentioning that experiment in this thread.

Partial ISA slot failure - have you tried the controller in a different slot ?

 

[Skip94]

Thanks, that makes a lot more sense.
Oddly, I do now have a different computer to try it in, my newly purchased 5160 turned up yesterday. However it presents exactly the same symptoms. I have tried several different slots, and have given all connections a thorough clean.
I shall do some more head scratching
Cheers
Andrew