PDA

View Full Version : Setting up IBM PC XT hard drives



CollinB1
July 11th, 2017, 10:32 AM
Hello,
I've got a PC XT with 2 Seagate 225 20MB hard drives.
A WD1002A-WX1 controller is installed in the slot second closest
to the CPU. Whenever I start up the computer, one drive's read light turns on
while the others doesn't.

39685

Are these data cables plugged in right?
Does it matter which connector they are plugged into.
I don't want to format the drives because I want to see what's on them.

In PC DOS 3.30, FDISK only says " there are no hard disks attached.
These hard drives have always been in the computer, but the cables came unplugged.

Also is the wider cable plugged in correctly?
Any help appreciated!

lutiana
July 11th, 2017, 10:45 AM
The two narrow cables look right (pin one is towards the back of the card, and indicated on the cable by the different colored edge), not sure about the wider cable as it is hard to see in the picture.

On the other end, the wider cable should daisy chain the two drives, you'll notice there is a twist on the cable towards the end, this is how the drive ID is set. J2 (top small connector) should go to drive 0, which if I recall correctly is the drive in the middle (before the twist) of the wider cable. J3 (the lower short connector) should go to drive 1, which should be on the end of the wider cable (after the twist). I could have the larger cable backwards, so try both ways.

Determining pin 1 on the wider cable is easy, if you look at the end of the cable, the twist is farther away from pin one that it is from pin 34. So if the twist is on the right side then pin 1 is on the left. Pin 1 on the card is towards the back of the card. So find pin one at the end of the cable and follow it back to the card and make sure its on the side of the connector that is towards the back of the card.

As for the DOS version, I'd try something a bit newer, maybe DOS 5. That way you can rule out the OS as a possible cause for not being able to read the drives. DOS 3.3 has some limits on partitions, file system types etc.

CollinB1
July 11th, 2017, 11:23 AM
Yes I have now confirmed that the cabling is correct,
But I tried them separately and the one which would turn
on the read light does not spin up.

The other one does not turn on the read light.

Are either of these problems fixable?

lutiana
July 11th, 2017, 11:30 AM
The one that does not spin up, try turning the PSU up, then quickly apply and remove power from the drive. This worked on a few drives I had that had a stuck motor, the sudden application of power enabled them to dislodge the stuck motor. Probably not the best thing to do to a drive, but could work.

Stone
July 11th, 2017, 11:44 AM
Are you sure the data cables (20 pin) are connected to the drives as they were originally? If they got switched after becoming unplugged neither drive would work. Try switching the data cables and see if that makes any difference.

MikeS
July 11th, 2017, 12:41 PM
Are you sure the data cables (20 pin) are connected to the drives as they were originally? If they got switched after becoming unplugged neither drive would work. Try switching the data cables and see if that makes any difference.Assuming the control cable is twisted (and an HD cable, not a floppy version) it shouldn't matter, should it?

But it sounds like there are other issues here.

Chuck(G)
July 11th, 2017, 12:54 PM
An ST225 should spin up with power applied and no signal cables attached. If it doesn't, it has problems that may be repairable.

Stone
July 11th, 2017, 01:05 PM
Assuming the control cable is twisted (and an HD cable, not a floppy version) it shouldn't matter, should it?Of course it matters...

The center connector on the control cable IS C: and the end connector IS D:. If you mismatch the data cables without reversing the control cable drive connections...

modem7
July 11th, 2017, 02:16 PM
The diagram at [here (http://minuszerodegrees.net/hdd/WDXT-GEN%20-%20single%20drive%20cabling.jpg)] is for the WDXT-GEN, but also applies to the WD1002A-WX1
If your control cable is twisted instead of straight-through (flat), then note the boxed comment in the diagram.

Another relevant diagram is at [here (http://minuszerodegrees.net/hdd/hard_drive_selection.png)].

CollinB1
July 11th, 2017, 05:07 PM
I started getting a 1701 error on boot up.

I took both hard drives out and I still get the error.

Apparently it has to do with a hard disk error.
It must be the controller.
There is no visible damage,
but is there a specific slot for it or can it go in any one?
I know the one closest to the cpu is reserved, but would it need to go there?

modem7
July 11th, 2017, 05:17 PM
I started getting a 1701 error on boot up.
I took both hard drives out and I still get the error.
Apparently it has to do with a hard disk error.
1701 is generated by code in the hard drive controller's BIOS ROM chip, and typically indicates a problem in the hard drive system. See [here (http://minuszerodegrees.net/5150_5160/hdd/1701_possibilities.htm)].
So, lots of possible causes.
It usually takes a while to appear after power on.

modem7
July 11th, 2017, 05:28 PM
I know the one closest to the cpu is reserved, but would it need to go there?
Any slot except slot 8.

modem7
July 11th, 2017, 05:30 PM
Is it possible, that even though the two drives have always been in the computer, that you are unaware of other things?
E.g. When the computer was in the possession of the previous owner, the hard drive controller failed, and that owner unsuccessfully tried a different make/model of controller (the WD1002A-WX1) without being knowledgeable enough to make that happen.
Have you personally seen the subject combination of hardware working?

If not, you should assume that you bought the components separately, and thus need to go through configuration (jumpers, etc.), low-level format, partitioning (FDISK), high-level formatting (FORMAT), etc.

CollinB1
July 11th, 2017, 05:52 PM
Yes it did take a while, but I had no idea about the previous owner because I found it.
I hooked them up when I first found it, and I never got the error,
But I could never get them working either.
I took it out once to look at the name, but I put it back in same slot.
Maybe bad connection in the slot?
I doubt anyone tried to repair it. It doesn't look like anyone
cared much about it :(
I'll try another slot.

All I know is that I didn't get the error at one point in time.

EDIT:
I moved it over a slot and cleaned connector and now I get no 1701 error.
But the one disk that turns on the read light does not spin up.
The disk that spins up does not turn on the read light.
Is there a way to fix this?

fatwizard
July 11th, 2017, 06:26 PM
The drive that isn't spinning up would cause a 1701 error all by itself. Another thing that will cause a 1701 is not having any drive connected. What kind of environment was the unit in when you found it? If it was in a damp location or other "less than ideal" space it will benefit from having all the connectors exercised to refresh the contacts. Using a pencil eraser on the card edge contacts has brought may non funtioning cards back to life. Having the card out and back in the slot, and working with the drive cable connectors is probably why you see an error where there was none before.

I think I would try getting the spinning drive working first. Leave the non-spinning drive for later because a non spinning drive is going to cause errors and make it harder to identify other issues.

After all these many years, it is pretty commom for drives like the ST225 to need a low level format to restore functionallity, but this ruins any chance of seeing what is currently on the drive.

modem7
July 11th, 2017, 06:59 PM
All I know is that I didn't get the error at one point in time.
A 1701 is generated by the ROM in the XT-class controller card. Therefore, possible causes of no 1701 (when one is expected) are, intermittent controller, and controller has poor connection to ISA slot.

Also to note is that the 1701 is triggered by different failure scenarios. So, a 'drive not ready' triggered 1701 might appear after one minute, but a 'drive ready, recalibrates, but a read of first sector fails' triggered 1701 might appear after two minutes.

modem7
July 11th, 2017, 07:02 PM
I think I would try getting the spinning drive working first. Leave the non-spinning drive for later because a non spinning drive is going to cause errors and make it harder to identify other issues.
Yes. No point in trying to get a dual-drive scenario working if a single-drive scenario cannot be made to work.

modem7
July 11th, 2017, 07:15 PM
Some 'MFM' drives, when they receive power' do a self test, and at some time during the self test, the drive's activity LED comes on.

I have only one functional ST-225 left. I have just now powered it on, with no control nor data cable, and I observed that the ST-225's LED never turned on. That to me suggests that the ST-225 either has no self-test, or that nothing in the self-test activates the LED.
That infers that the ST-225's LED only comes on during controller initiated operations.


But the one disk that turns on the read light does not spin up.
The disk that spins up does not turn on the read light.
If that was observed in a dual-drive configuration, then the lack of the LED coming on may be because the subject ST-225 is configured as the second drive, and during machine power up, the BIOS ROM in the controller only attempts operations on the first drive.

CollinB1
July 11th, 2017, 07:33 PM
I switched out the heads because I noticed the drive that spun had a cracked ribbon connector
between the head and the board.
Once I finished switching the 2, the second drive spun up!
I just thought that it was worth a shot switching the heads because
one wouldn't light up in single configuration
and the one that would had the bad cracked ribbon.

Switching the heads probably messed it up, right?
The head still move back and forth normally.

I guess it was just my last shot.

Also how do you perform a low level format?

fatwizard
July 11th, 2017, 08:37 PM
If you were very careful, and lucky, the drive might still function. I have found that old drives like the ST225 can be surprisingly tolerant of being opened to the room air for a short while. Hardly ideal, but sometimes it's the only thing between an old drive and the recycle bin.

The info on your card can be found here.

http://stason.org/TULARC/pc/hard-disk-floppy-controllers/U-Z/WESTERN-DIGITAL-CORPORATION-Two-MFM-ST506-412-driv-221.html

Confirm that the drive selection jumpers on the card are correct. They are different for each of the three revision of the ROM BIOS. If the label on the ROM is missing or can't be read, see if the jumpers correspond to one of the possible settings for the ST225 (which are Cylinders 615, heads 4, write precomp 300 or none). Don't be surprised if you can only select for 612 cylinders instead of 615.

Boot the system with a DOS diskette that has some DOS utilities on it including fdisk.com, format.com, and debug.com. At the DOS prompt type debug and hit return. At the - prompt type g=c800:5 and hit return. Follow the prompts after that. I believe your controller doesn't have the free format option, so the prompts will just be for which drive and (C or D or maybe 0 or 1) and the interleave (keep the default here for now).

Type Y and hit return (I don't have your specific controller card, but the prompts should be very similar). The drive light should come on and you might hear a steady ticking sound as the stepper moves the heads across the disks. It will take several minutes to finish, after which you fdisk a partition into place and then high level format.

CollinB1
July 12th, 2017, 03:23 AM
After replacing the heads, It finally recognized the drives!
But fdisk returned an error reading disk.
Is this because of the no low level format?

Stone
July 12th, 2017, 03:43 AM
Probably...

Time to do the LLF and find out.

MikeS
July 12th, 2017, 03:52 AM
Of course it matters...

The center connector on the control cable IS C: and the end connector IS D:. If you mismatch the data cables without reversing the control cable drive connections...

Duh; brain fart, somehow I read that we were talking about swapping drives.

But isn't it like a floppy, i.e. the end connector is C: and the center is D: ?

Stone
July 12th, 2017, 04:00 AM
Duh; brain fart, somehow I read that we were talking about swapping drives.

But isn't it like a floppy, i.e. the end connector is C: and the center is D: ?Nope... not if the ribbon has a twist.

And if it's twist-less -- then C: and D: are determined by the Drive Select jumpers on each HD.

MikeS
July 12th, 2017, 05:06 AM
Nope... not if the ribbon has a twist.

And if it's twist-less -- then C: and D: are determined by the Drive Select jumpers on each HD.Hmm.. so you're saying that the diagram Modem7 pointed to is incorrect? I guess it depends on how the drives are jumpered. I'm assuming the same as floppies, i.e. both drives as second drive; if you set them both to DS0 that would indeed probably reverse the positions, although I'd find that a little confusing and a little inconvenient if/when you add a second drive.

Stone
July 12th, 2017, 05:33 AM
Hmm.. so you're saying that the diagram Modem7 pointed to is incorrect?If you're referring to this (http://minuszerodegrees.net/hdd/WDXT-GEN%20-%20single%20drive%20cabling.jpg) diagram, no. It's correct, obviously. :-)

That's a 'twist-less' cable, exactly like I mentioned in my previous post.

fatwizard
July 12th, 2017, 07:32 AM
It does work like floppy drives. Both drives are jumpered as the second drive, but the twist in the cable makes it the first drive. So the drive at the end of the cable, after the twist, is C:

Stone
July 12th, 2017, 08:20 AM
It is not like floppy drives.

C: is before the twist.

I think this explains it quite well:

http://minuszerodegrees.net/5160/misc/5160_hard_drive_cabling_double.jpg

Chuck(G)
July 12th, 2017, 09:13 AM
Sheesh. :)

You're both correct. Take a look at a "twisted" control cable here (http://www.minuszerodegrees.net/misc/floppy_vs_hard.jpg).

Note that conductor positions 25-29 are swapped with the twist. Odd-numbered conductors are all ground, so they don't matter. The signal conductors that get swapped are 26 (DS0) and 28 (DS1) per the ringout here (https://allpinouts.org/pinouts/connectors/data_storage/st506-412-mfm-rll/).

So, if you want to put C: on the far end of the cable and D: in the middle of the control cable, you set both drives to DS1 and cable the data cables accordingly. If you want D: on the far end and C: in the middle, you set both drives to DS0 and swap the data cables to the reverse of the previous. Regardless, the most distant drive from the controller gets the terminator.

This is one of the reasons that I always used the "flat" control cable, even when I had the "twisted" ones available. C: set to DS0 and D: as DS1. Everything's obvious--and the same cables work with ESDI. What more could you want?

MikeS
July 12th, 2017, 09:56 AM
It is not like floppy drives.

C: is before the twist.

I think this explains it quite well:

http://minuszerodegrees.net/5160/misc/5160_hard_drive_cabling_double.jpg
I think you keep missing the point that it depends on how the drives are jumpered; If you set DS like a floppy they connect like a floppy; this explains it better (note that drive C: is always on the end where it belongs with the terminator):
http://minuszerodegrees.net/hdd/hard_drive_selection.png

MikeS
July 12th, 2017, 10:05 AM
...This is one of the reasons that I always used the "flat" control cable, even when I had the "twisted" ones available. C: set to DS0 and D: as DS1. Everything's obvious--and the same cables work with ESDI. What more could you want?
I'm with you, Chuck, especially since most of the time (like the last few days) I'm working with dual HDs on S100 systems following the DS0/1 straight cable 'standard'.

But when there is a twisted cable I prefer to put the terminated C: drive at the end of the cable where it belongs, properly jumpered of course, and then just put the second drive in the middle if there is one, without having to move the terminator etc.

Stone
July 12th, 2017, 10:43 AM
I'm with you, Chuck, especially since most of the time (like the last few days) I'm working with dual HDs on S100 systems following the DS0/1 straight cable 'standard'.We're discussing an IBM 5160, aren't we?

Why would you introduce S100 information?

CollinB1
July 12th, 2017, 11:29 AM
I did the debug command in pc dos 3.3 and got bad command or filename.
I also had to switch out the platters because they were really scratched on one.

So is it a problem that fdisk returns error reading from drive?

It finally recognizes the drive and I don't get the 1701.

Stone
July 12th, 2017, 11:35 AM
I did the debug command in pc dos 3.3 and got bad command or filename.Then you need to locate your copy of debug.

MikeS
July 12th, 2017, 11:41 AM
We're discussing an IBM 5160, aren't we?

Why would you introduce S100 information?
Just agreeing with Chuck that I also prefer the 'standard' untwisted cable setup in PCs (including the 5160), which is the same as in S100 and most other systems using ST512 interface drives (which is the cabling I just happen to have been working with for the past few days while testing a number of drives).

That way it doesn't matter which drive goes where on the cable as long as the jumpers are set correctly (and the data cables are connected correctly, which takes us full circle back to the beginning ;-) )

CollinB1
July 12th, 2017, 11:47 AM
Sorry nevermind.
It was on the utilities disk.

Now I've got a more concerning problem

Whenever I run debug, it either freezes or gives me
I/O SYSTEM ERROR
40000
CONTINUE(Y/N)?
<y><enter>
I/O SYSTEM ERROR
00000
CONTINUE(Y/N)?

Is this the card having an i/o problem?

Stone
July 12th, 2017, 11:50 AM
Just agreeing with Chuck that I also prefer the 'standard' untwisted cable setup in PCs (including the 5160), which is the same as in S100 and most other systems using ST512 interface drives (which is the cabling I just happen to have been working with for the past few days while testing a number of drives).

That way it doesn't matter which drive goes where on the cable as long as the jumpers are set correctly (and the data cables are connected correctly, which takes us full circle back to the beginning ;-) )I only do IBM and have gotten dozens of HDs from varied sources and since all of them have been set to operate with the twist I have seen fit to not change the jumpers on the drives for the sake of simplicity. :smile: KISS

Stone
July 12th, 2017, 11:51 AM
Sorry nevermind.
It was on the utilities disk.

Now I've got a more concerning problem

Whenever I run debug, it either freezes or gives me
I/O SYSTEM ERROR
40000
CONTINUE(Y/N)?
<y><enter>
I/O SYSTEM ERROR
00000
CONTINUE(Y/N)?

Is this the card having an i/o problem?I'd say it's more probable the drives themselves have the problem. They're likely hosed.

MikeS
July 12th, 2017, 12:05 PM
I only do IBM and have gotten dozens of HDs from varied sources and since all of them have been set to operate with the twist I have seen fit to not change the jumpers on the drives for the sake of simplicity. :smile: KISS
No need to change anything; just sayin' that your way is not the only way and that some of us see an advantage to doing it the other way. I've certainly seen PCs using both straight and twisted cables.

Chuck(G)
July 12th, 2017, 12:13 PM
Sorry nevermind.
It was on the utilities disk.

Now I've got a more concerning problem

Whenever I run debug, it either freezes or gives me
I/O SYSTEM ERROR
40000
CONTINUE(Y/N)?
<y><enter>
I/O SYSTEM ERROR
00000
CONTINUE(Y/N)?

Is this the card having an i/o problem?

I wonder if your copy of DEBUG is hosed. You should at least get the DEBUG "-" prompt.

Stone
July 12th, 2017, 12:42 PM
No need to change anything; just sayin' that your way is not the only way and that some of us see an advantage to doing it the other way. I've certainly seen PCs using both straight and twisted cables.As have I...

Chuck(G)
July 12th, 2017, 02:49 PM
The twist in floppy cables makes sense--and is necessary, as the PC family separates out motor enable lines for each drive. Normal (S100 CP/M, whatever) convention was to use a flat cable with means that when one drive motor was activated, all drive motors were. I suspect that the 63W power supply on the original 5150 was a bit too skimpy to support 2 floppies with the motors running, as well as a PC full of expansion cards.

But it's stayed with us all along.

For ST412-type drives, the twist swaps only the drive selects--so not so important.