RE: FD 2000 Teac mechanism

This kind of sounds like an issue that was in the vice emulator.

It seems the floppy disk controller has a time out for recalibration on the FD-2000. The amount is 77 pulses (or tracks), but of course the disk can have 80 tracks. So what would happen was the controller would time out, but the ROM code expected it to, and would reissue the recalibrate command. The thing is the ROM code expected a particular return value from the disk controller on the first failure. I had to hard code this in the vice emulator or else the drive would behave like the floppy was removed and subsequently all IO to the disk failed. Perhaps the return value is not directly derived from the controller itself but also the mechanism in some way.
Here is the comment I placed in the code:
/* CMD FD2000's mechanism also sets bit 5 and 6 on recalibration timeout, it is hard coded in ROM to check for $71 on result (_at_ $878a) and try again */
/* The FD4000 uses 85 pulses, so this shouldn't timeout */
Maybe a good test would be to put the same mechanism in an FD-4000 and see if the issue occurs when using high density disks. In that situation the timeout would never happen. Of course I’m not sure how it would like that drive mechanism as opposed to an extended density one.
Since this problem I stated occurs during recalibration over 77 tracks, are the issues happening when the drive is moving between the partition table and the beginning of the first partition (like a native one since the header and BAM are at the beginning of the disk)? Does it happen if the current working partition starts after track 4? I would suggest creating a dummy native partition at the beginning of the disk and avoiding it during these tests.

Roberto

From: Terry Raymond <traymond160_at_gmail.com>
Sent: Tuesday, June 01, 2021 12:30 AM
To: cbm-hackers_at_musoftware.de
Subject: FD 2000 Teac mechanism

Hi concerning the CMD FD-2000 I have one I first bought from CMD
the TEAC mechanism needs some modifications to work properly.
I found the Disk change and that seems to work.
In all I found 3 noticeable changes on the TEAC PCB so I tried this and it kind of works
but sometimes it acts like the disk is not even formatted.
I tried the same files on my still working FD-2000 and those same files copied fine.

So has anyone here ever figured out the hacks for this if so could you please send me
viewable pictures please.

Terry Raymond
traymond160_at_gmail.com<mailto:traymond160_at_gmail.com>