• Please review our updated Terms and Rules here

IBM ps/2 model 50

Is this drive a WD-325? Does anyone know the heads/cylinders/precomp for it? I may try the wdfmt tool on it.
 
Can one specify the hard disk type manually somehow? I saw in the table there was a 3 head drive with more cylinders, but I could partition it so it doesn't use those cylinders.
 
Could try spinrite, should work ... Chances are the drive is toast. Sounds like a head died. Upside, st-225's arent that hard to find. (y)

It most likely wasnt parked before shipping.
 
Would something like an on track disk manager allow me to change the geometry to 3 heads instead of 4?

The format completed; 3.5MB bad out of 20MB. Scandisk is running much faster, but it is running into sectors it had to reread over and over again, but aren't reporting errors.
 
I wrote a PS/2 CMOS tool here that can read, validate the checksum, write, and fix the checksum. Source for BC31 includded.

It took a bit to figure out the checksum algorithm as I couldn't find it online, but I assembled the even/odd bios together, disassembled it, and I got lucky and I found the section f000:2fac kicks things off and calls f000:24fe which also calls f000:2514 where the scrambling occurs:

Code: 
uint16_t checksum()
{
  reg16 a, d;
  int i1;
  uint8_t c1;

  d.w.x=0xffff;
  a.b.h=0x00;

  for (i1=0x10; i1<0x32; i1++)
    {
      //get cmos byte
      a.b.l = read_cmos(i1);

      //xor dh,al
      d.b.h ^= a.b.l;

      //mov al,dh
      a.b.l = d.b.h;

      //rol ax,c1
      a.w.x = a.w.x << 4;

      //xor dx,ax
      d.w.x ^= a.w.x;

      //rol ax,1
      a.w.x = a.w.x << 1;

      //xchg dh,dl
      c1 = d.b.h;
      d.b.h = d.b.l;
      d.b.l = c1;

      //xor dx,ax
      d.w.x ^= a.w.x;

      //ror ax,c1
      a.w.x = a.w.x >> 4;

      //and al,e0
      a.b.l &= 0xe0;

      //xor dx,ax
      d.w.x ^= a.w.x;

      //ror ax,1
      a.w.x = a.w.x >> 1;

      //xor dh,al
      d.b.h^=a.b.l;
    }
  return d.w.x;
}

1655067640342.png
 
The controller is what does the geometry on MFM, I don't think you can just tell it not to use a head. Could be wrong but I don't think thats how it works. It doesn't work like IDE drives. So On Track will be useless in this situation.

You will need to find a copy of spinrite. Thats the best utility I know of. Sadly I cannot share commerical software on here.... *cough* Internet Archive *cough*.
 
Probably so. Does anyone know how to rewrite a test cylinder? Is there a way to do that? I've been messing with the drive (nothing to lose) and the test cylinder is now lost. The IBM reference disk won't even try to LL format it now because the test cylinder or some information making it match the drive number is missing.

SSTOR will begin to LL format it, but within 15 cylinders from the top, it will have a problem where the arm will move all the way to 0 and it will fail.
 
Point of advice. Only do formatting and testing while the drive is warm. If you have any chance of getting this drive to work, start again another day when you can let the drive spin up and run at least 4-5 hours before you do any testing or formatting.
 
Semi-off-topic, anyone know how likely it is that a ‘171’ error in a PS/2 Model 56SLC is just a bad battery? There’s one on eBay:

www.ebay.com

Vintage IBM Personal System/2 PS/2 Model 25 Computer type 8556 | eBay

<p dir="ltr" style="margin-top:0; margin-bottom:0;">For parts as is all it shows is the black screen with numbers </p> <p dir="ltr" style="margin-top:0; margin-bottom:0;">Vintage IBM Personal System/2 PS/2 Model 25 Computer type 8556.</p>
www.ebay.com

Nice thing about these later PS/2s is they use SCSI drives, which while still slightly a pain are far more replaceable than the drives for the original 50/55/60/65/70/80 series…
 
With nothing to lose I opened it up yesterday and the platters did not look great. I disassembled it enough to get them out and gave them a wash in the sink with some soap which seemed to clean off the junk on them. I tried to take a little IPA to the heads ever so carefully too, but the heads wanted to stick together (magnetized?). Anyway, I got it back together and it is unhappy that it can't find the test cylinder. Maybe the firmware of the drive/controller demands that, but I was thinking that I could just rewrite a low level format and get it going if there was any chance in that. I think it is done for. Sometimes it boots with a 1780, sometimes a 1790, sometimes no error at all. I can see it at startup do some seeking out to what I am guessing is that "test cylinder" and then return to 0 again. The couple of times I try to LL format it using speedstor, it would start at 604 and go downward to maybe 580 or so because something happens and it decides to slowly seek to 0. I am guessing it thinks it needs to rehome/calibrate for some reason, but why it would decide it needs to do that in the middle of the LL format sstor is doing I have no idea.
 
Well, in any event since we can have a funeral for the drive, I have an ESDI controller you can have for free + shipping if you want it. Might be a cheaper route for you. Also have a kingston memory upgrade card if interested.
 
Thanks twolazy that is kind of you! Let me see if I can find a drive first. I would love to get some sort of HD going for this, but it sounds like it can be challenging!
 
You must log in or register to reply here.
Back
Top