• Please review our updated Terms and Rules here

Computerland BC286 Hard Drive issues

falter

falter

Veteran Member
Joined
Jan 22, 2011
Messages
6,573
Location
Vancouver, BC
Decided to mess around with this BC286 I got a while back. I *think* this is technically an XT class computer running a 286 somehow, sort of like the Tandy 1000TL.

It has both an IDE/floppy controller, and an Adaptec ACB-2002 MFM hard drive controller. As it came to me, it had a (ugh) Miniscribe 3438 30-someodd meg HDD. The drive spins up, but flashes an error which, if I'm reading the code right, is some kind of weird seek error. This is a video of it here:
The drive spins up and the stepper motor does its thing, so it's not seized up.

In an ideal world, I'd find a way to make the machine run with it's original drive. But maybe that's not possible.

I thought of maybe using a Seagate 42MB MFM drive I have that I know is working, but I can't get past a 1701 error on startup.

It's been 30+ years since I had to deal with this sort of thing and have forgotten much. I'm trying to remember - can and IDE and MFM controller co-exist like this? I'm assuming they can since the machine came like this.

This machine has a battery backed BIOS with the usual Rev H options for hard drive settings. But I'm assuming those are for IDE, not MFM.

The Adaptec controller doesn't seem configured per the manual. The jumper is set for a Seagate 10mb - I would have thought it should be set to 'other'. It must be correct, at least for the Miniscribe.. otherwise I've no idea how it ever booted.

Anyone have any advice on this? I mean, I could just use the IDE controller and use one of the now numerous low cap IDE drives I have.. but I'd love to have the machine working with its original drive, or something resembling it if possible.
 
falter said:
As it came to me, it had a (ugh) Miniscribe 3438 30-someodd meg HDD. The drive spins up, but flashes an error which, if I'm reading the code right, is some kind of weird seek error. This is a video of it here:
Click to expand...
Based the 3438's manual, I see 1101, which is code D ("Seek error during burn-in or recal")

falter said:
It's been 30+ years since I had to deal with this sort of thing and have forgotten much. I'm trying to remember - can and IDE and MFM controller co-exist like this? I'm assuming they can since the machine came like this.
Click to expand...
I presume that the IDE controller is AT-class.
Your MFM controller (Adaptec ACB-2002) is an XT-class controller.
They can co-exist because they use different hardware resources - see [here].

falter said:
I thought of maybe using a Seagate 42MB MFM drive I have that I know is working, but I can't get past a 1701 error on startup.
Click to expand...
A 1701 error can be explained by a low-level format mismatch between the MFM controller and MFM drive, such a mismatch being expected after the drive has been substituted.

So, after substituting the drive with another, the normal sequence would be to:
1. Configure the ACB-2002 MFM controller to suit the replacement drive; then
2. Low-level format the replacement drive; then
3. Partition the drive (e.g. via FDISK.COM); then
4. High-level format the drive (e.g. via FORMAT.COM).

To low-level format the replacement drive, refer to the MFM controller's user manual (there are various ways, dependent on the controller make-model).

You should be able to get past the 1701 error, e.g. boot to floppy.

falter said:
This machine has a battery backed BIOS with the usual Rev H options for hard drive settings. But I'm assuming those are for IDE, not MFM.
Click to expand...
Correct, if we expand your "MFM" to XT-class MFM.

(AT-class MFM controllers and XT-class MFM controllers are different animals - see [here].)

falter said:
The Adaptec controller doesn't seem configured per the manual. The jumper is set for a Seagate 10mb - I would have thought it should be set to 'other'. It must be correct, at least for the Miniscribe.. otherwise I've no idea how it ever booted.
Click to expand...
See [here], which is for AT-class machines.

The same principle applies to XT-class controllers; you can configure them for a drive that has same-or-less cylinders, or same-or-less cylinders heads, or both. Some loss of capacity results.

But yes, for the Adaptec ACB-2002 controller and Miniscribe 3438 combination, one would expect the ACB-2002 to be jumpered for the "ALL OTHERS" setting, which results in the ACB-2002 becoming a 'dynamic' controller as described at [here].
 
Many thanks for the detailed response. I will use the debug commands spelled out in the Adaptec manual to do a LLF on the Seagate.

I'm assuming the seek error on the original drive is unrecoverable.
 
Heh. So I got into debug and was working away at entering parameters when the machine froze. I rebooted.. now it won't post. I've tried reseating cards, etc... nothing. The only thing I notice is the 80287 copro is really hot.. I'm wondering if that might be causing it to lock up.
 
Removing the 287 got me booting again. I also had to change the jumper on this NOS ST251-1 to DS1.. now the Adaptec is doing a LLF on it. Man they take a long time. So far so good.. only a few bad sectors.
 
Dumb question.. how long should an LLF take on an 40mb drive like this?

It's been going over 25 minutes.. and it seems like it's bringing up the same bad sectors over again. Like "bad cyl = 01A7, hd = 02, sec = 0B". it'll hit something like that, it'll carry on ticking away, and then it'll hit it again 5 minutes later. seems to be the same four or five that keep coming up. Is it possible it's caught in a loop?
 
falter said:
I'm assuming the seek error on the original drive is unrecoverable.
Click to expand...
Assuming that the error codes largely run in sequential order, then the errors of 'Unable to uncover Track Zero sensor' and 'Unable to cover Track Zero sensor' and 'Track Zero interrupter misadjusted' did not occur. For those, I expect 'short' seeks to be used (head seeking/stepping of only a few tracks) on and near track 0.

So, what does 'Seek error during burn-in or recal' imply? The drive will not be in a burn-in mode, something that Miniscribe would use at the factory. So the error message becomes 'Seek error during recal'. A 'recal' is taking the heads back to track 0, expecting the track 0 sensor to activate. Guesses: Maybe short seeking is working, but not 'long' seeking. Or maybe short seeking is working near track 0, but there are seek problems (short and long) significantly distant to track 0.

Perhaps try re-seating the various connectors on the drive.

falter said:
Dumb question.. how long should an LLF take on an 40mb drive like this?

It's been going over 25 minutes.. and it seems like it's bringing up the same bad sectors over again. Like "bad cyl = 01A7, hd = 02, sec = 0B". it'll hit something like that, it'll carry on ticking away, and then it'll hit it again 5 minutes later. seems to be the same four or five that keep coming up. Is it possible it's caught in a loop?
Click to expand...
The norm for a 20 MB sized MFM drive is just a couple of minutes, but that is for what could be called a 'basic' low-level format..
Some controllers do various forms of verification, and some of those may make multiple write-then-read passes (each time, using different test data).

An alternate low-level format method would be to reconfigure the ACB-2002 for a drive selection that is not "ALL OTHERS", then use SpeedStor software to attempt the low-level format. SpeedStor uses the term 'initialize' instead of low-level format.

falter said:
Is it possible it's caught in a loop?
Click to expand...
A bug is a possibility, say, one that occurs when there are lots of detected bad sectors. The firmware authors/testers many have only tested against drives with few bad sectors.
 
I got it!

Borrowing from Adrian's Digital Basement, I took the Miniscribe and tested out the movement of the stepper motor and it was quite stiff. I added some 3 in 1 oil, just a little bit, and then worked it in... noticed it was spinning much more freely, and then put it back together and voila! It worked! Booted right up to DOS!

This Miniscribe has a very unique stepper sound.. like the drive has been chain smoking for 40 years. Very raspy. But yeah, all good.

Now to figure out why the copro died. Are copros supposed to get hot?
 
Thank you. I wonder why copros get so hot but not the CPU.

I plugged it back in.. computer works fine..Autocad loaded up (it depends on having a copro).

Really glad this machine is all working again with its original hardware.
 
You must log in or register to reply here.
Back
Top