TU58 yet one more time.

2016-05-29 20:10 GMT+02:00 Al Kossow <aek at bitsavers.org>: Yes. I am aware of that. But that also means that the head can be moved into the spot where the information is recorded. My quick reading of the QIC-117 spec give that the one control the drive with a quite weird interface consisting of the STEP, TRACK ZERO and INDEX lines. The READ DATA is supposed to be the actual flux transitions. Likewise is WRITE DATA. What is not yet clear to me is whether that the QIC-117 drive is rather dumb and leaves most work to the FDC or if it includes a lot of logic to handle the reading process. http://www.qic.org/html/standards/11x.x/qic117j.pdf The QIC36 drive that was used to recover the information stored on the Zilog S8000 tapes is completely stupid. The read data just reflected the flux transitions which AJ recorded using a logic analyzer so that it was possible to decode the MFM data stream in software. The tracks of course also differed since the QIC36 drive was 9 track while the DEI drive used for recording the S8000 tapes used four (fixed) heads. But nevertheless it was possible to recover the entire tape thanks to the circuitry that AJ devised to control the head position. Using a more modern SCSI drive it will not be possible to read these tapes since they only handle GCR encoding (as does QIC02 drives) Could this kind of operation be done with a QIC-117. I.e. is the drive stupid enough? Reading more of the spec may indicate that "segment" concept might be the culprit that makes it non-feasible. It could be that the drive keeps track of the segments somehow. Of course there is always the possibility of "hacking" the drive. But then it would help with schematics which is probably not available for these quite "modern" things. All input appreciated! That would be very nice. Send you a message off-list. /Mattis