[Simh] PDP7 DECtapes

Christian Brunschen christian at brunschen.com
Sun Jun 7 07:06:05 EDT 2015


I wonder if it would be worth putting together something like the Kryoflux
<http://www.kryoflux.com/?page=kf_features> for tape drives (the Kryoflux a
device that connects directly to the floppy drive instead of through a
floppy controller, and records the timings of the magnetic transitions at a
low level, so that they can be used to extract data, or duplicate a floppy,
in almost any format).

// Christian

On 7 June 2015 at 03:47, Bob Supnik <bob at supnik.org> wrote:

> Yeah, PDP-7 (and earlier) DECtapes are different. They use different
> checksum algorithms, for one thing, and some early PDP-4 and -7 DECtape
> controllers used four-word headers instead of five. The checksum is
> probably key; only five "four word header" DECtape controllers escaped the
> factory.
>
> /* Checksum routine */
>
> int32 dt_csum (UNIT *uptr, int32 blk)
> {
> int32 *fbuf = (int32 *) uptr->filebuf;
> int32 ba = blk * DTU_BSIZE (uptr);
> int32 i, csum, wrd;
>
> #if defined (TC02)                                      /* TC02/TC15 */
> csum = 077;                                             /* init csum */
> for (i = 0; i < DTU_BSIZE (uptr); i++) {                /* loop thru buf */
>     wrd = fbuf[ba + i] ^ DMASK;                         /* get ~word */
>     csum = csum ^ (wrd >> 12) ^ (wrd >> 6) ^ wrd;
>     }
> return (csum & 077);
> #else                                                   /* Type 550 */
> csum = 0777777;
> for (i = 0; i < DTU_BSIZE (uptr); i++) {                /* loop thru buf */
>     wrd = fbuf[ba + i];                                 /* get word */
>     csum = csum + wrd;                                  /* 1's comp add */
>     if (csum > DMASK)
>         csum = (csum + 1) & DMASK;
>     }
> return (csum ^ DMASK);                                  /* 1's comp res */
> #endif
> }
>
> This is documented in the 18b PDP DECtape controller, based on sources I
> have for PDP-7 DECtape code from the late 60s; but for some reason, I never
> published a paper on it.
>
> This is one reason why recovering DECsys required reading the DECtape on
> the last working PDP-7 in the world.
>
> /Bob
> _______________________________________________
> Simh mailing list
> Simh at trailing-edge.com
> http://mailman.trailing-edge.com/mailman/listinfo/simh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.trailing-edge.com/pipermail/simh/attachments/20150607/2969dff9/attachment.html>


More information about the Simh mailing list