Fixing an RL02

Tonight I wanted to at least see what was wrong with the RL02 which lit its FAULT light when a disk pack was inserted. The Technical Manual lists several possible causes for a FAULT indication:

  • Drive select error
  • Seek time out error
  • Write current in heads during sector time error
  • Loss of system clock
  • Write protect error
  • Write data error
  • Spin error

It's not until you delve deeper into the manual (namely the description of the Multipurpose or MP register) that these phrases start to make any sense. As the drive was at least attempting to spin up, but took it's time to indicate a fault, my hunch was that it was either a spin error or maybe a seek time out. The former occurs if the spindle either doesn't get up to the required speed in the allotted time or ends up going too fast. The latter happens if the heads don't find the track they're looking for in the required time; I assumed this was a possibility as the drive heads pop out to check for the outer track when a disk is loaded.

As the TU58 set-up was playing up, I wasn't looking forward to booting XXDP to run the diagostics. Fortunately, the manual provides a few toggle-in programs, one of which can be used to examine the drive's status. In fact, you don't even need to use a program as the procedure involves nothing more than reading and writing addresses:

  • Deposit a 3 into the DAR at 774404 (13 if you want to reset the drive)
  • Deposit a 4 into the CSR at 774400 (404, 1004 or 1404 for units 1, 2 or 3)
  • Wait for the operation to complete
  • Read off the contents of the MPR at 774406

This is easily done with a few choice ODT commands. The resulting MPR read 004210, the "4" (bit 11) indicating that it was indeed a spin error that was casuing the fault. As the drive sounded pretty much the same as the other, which was working fine, I suspected that the drive was incorrectly sensing the spindle's speed. The speed is sensed my a transducer in front of the spindle (it's the bit with the gold-coloured cap and long grey wire heading off into the head positioner's enclosure - you can see it on this picture).

A multimeter told me the wires were still good. Although there's only one wire exposed at the transducer's end, the other two pins on the connector are electrically connected to each other inside the sensor itself. The proper way to test the transducer requires an oscilliscope, which I don't have. What I do have is a known-good transducer, so I set to transplanting this from the working drive to the faulty one. Then it was moment-of-truth time... I powered up the system and the LOAD light came on. Nothing new so far, but I apparently hadn't caused any other problems routing around in the drive's innards. In went a disk pack and I pressed the LOAD button. The light went out, the drive spun up and... plink! On came the READY indicator! Not bad going for a DEC newbie, even if I do say so myself.

I put the suspect transducer back into the other drive and, sure enough, it FAULTed exactly as the other did. Damning evidence indeed. So now I need to find a new transducer. Thankfully these are a lot easier to transport than a complete replacement RL02!

Next: XXDP Revisited >>