Of course... I don't keep secrets from the classic computer crowd :-) The stuff below is all from memory, so I might be mistaken about something. Much of it was obvious to me from the printset for the controller and drive, though, and to be honest, if you can't figure out the details from said scheamtics I don't think you'll be able to design your own controller. Firstly remember I was only trying to move the head around and request status. I didn't actually try to read or write data. I do know that the data interface to the RL is a low-level one, essentially the bitstream to/from the head, so the controller hss to separate data and clock, work out what sector it is, etc. That bit I didn't attempt. The interface is all differential TTL-level-ish signals. I think I used 26LS31s as drivers and 26LS32s as receivers. The controller also has to supply the 'energiser' clock. This is a somewhat odd freqeuncy IIRC. To send a command, you send a 16 bit serial bitstream to the drive. Some of the bits are the 'cylinder offset' (how far to move the heads), another is the direction, yet another is the side select bit. IIRC, the order of bits is exactly as given in the programming information for the RL11 controller. To request status, you send it a command with a particular bit set. The drive then sends 16 bits back to the controller. What you get is, again, exactly as documented in the programming manuals. I will dig out the details of the strobe signals, etc to send/receive commands. That should be a start, anyhow -tony