I am trying to recover an old mini-machine that is using a Micropolis 1325 drive, and I am successfully reading parts of the drive. I am trying to recover as much as data from the drive, and potentially also try to boot up from the drive.

(The mini-machine in question is a Norsk Data ND-110/CX from around 1988)

I am able to read any track/cylinder on the drive as long as I am accessing head 0-3. As soon as I try to access head 4-7 i get "Address mismatch".

I thought first it could be signaling from the controller, but forcing the "head select bit 2" to GND (active low) doesn't improve the "Address mismatch for head 4-7. However, while forcing it to GND and trying to read head 0-3 now gives "Address mismatch".

Removed the GND signal from bit 2, and now I can read 0-3 (as before) but head 4-7

So the drive uses the head select bit 2, but heads 3-7 are still not behaving. Any ideas on what can cause this - and potential fixes - is greatly appreciated.

  • 3
    It sure sounds as though even though you are selecting heads 4-7 that these are not actually being selected properly. So you're requesting 4 but it's reading 0 and then reports an error. I suspect some sort of hardware issue going on here.
    – jwh20
    Jul 19, 2023 at 19:24
  • 1
    @jwh20 if the drive would read head 0 when you request head 4, it would still do so when HS2 is forced active, and trying to read head 0 with the bodged HS2 signal should still work. OP reports that you can no longer read head 0 with the bodged HS2 signal, though. This conclusion is true at least if the bodge is not wired to the controller. The controller must not see the forced HS2 signal for this experiment to be meaningful, as we don't know whether the expected head number is read from the signal on the control cable, or from some controller-internal latch. Jul 20, 2023 at 5:56
  • I agree that it looks like the wrong head is selected, causing mismatch with the head number in the address field on the harddisk. As for why forcing "head select bit 2" doesn't work, potentially you used the wrong pin, or GND doesn't work. First step would be to check levels on all head select bits during an actual read, and see if they match what you'd expect.
    – dirkt
    Jul 20, 2023 at 12:36
  • So clearly head 0-3 is not selected when I force "head select bit 2" to GND (active=1). This I deduce from the fact that I get "address mismatch" from the read of head 0-3, as I am guessing data is now flowing from head 4-7. There must be something that is flowing correctly, as i dont get parity error or timeout - so clearly "something" is read. I found the schematics for the drive, and if you have any idea on what I should test, please let me know - bitsavers.org/pdf/micropolis/Micropolis_1325_schematic.pdf
    – Ronny
    Jul 21, 2023 at 12:50


You must log in to answer this question.

Browse other questions tagged .