The Sinclair QL Forum

Hi,

Sorry for confusing the issue, I dud not realise that really old disk drives were being used.

dex wrote: Same as on SCSI.
Remove on the thru device, keep them in place on the last device.

Thank you Dex for your explanation.
So in conclusion, I made dozens and dozens of tests with and without terminations, non-twisted and twisted cables, all possible format variations and media types, Trumpcard and Goldcard and more. And... no way.
At this point we can say that QL FDD controllers (at least Miracle's ones) can not handle 1.2MB 5.25" floppy drives.
Anyway, it is always possible to manage Qdos 5.25" disks when the drive is connected to PC/Qemulator and recover data.

The disk controller in the Gold and Super Gold card should definitely be able to drive a 1.2Mb drive.

an HD 3.5" drive is 80 tracks, 18 sectors per track, and 2 sides

an HD 5.25" drive is 80 tracks, 15 sectors per track and 2 sides.

The Trump card floppy controller - 1771 - cannot reach the HD densities with it's clock rate of 8MHz



Attached is the data sheet for the controller on the gold and super gold cards.

Pr0f wrote:The disk controller in the Gold and Super Gold card should definitely be able to drive a 1.2Mb drive.

Well, definitely. The (S)GC floppy controller chip is the same that used to be in the IBM AT.

Cristian wrote:QL FDD controllers (at least Miracle's ones) can not handle 1.2MB 5.25" floppy drives.

1.2 MB 5.25" (high density) FDD has 360 RPM (same as 8"), while 3,5 " FDD (in IBM PC - Japanese computers as Sharp X68000 or PC-98 being an exception) and 360/720 K (double/quad density) FDDs have 300 RPM.
If the controller does not expect higher RPM, cannot record whole track.
I have never tried it, but some 1.2 MB drives have switch to 300 RPM for compatibility purposes.

Maybe it's down to the low level QL floppy driver. It knows about 720K, 1.4M, and 3.2M disks. Perhaps it gets confused when it sees a 1.2M.

That might explain why when Qemulator formats the floppy, it gives 720K. It can't return 1.2M because that's not one of the recognized sizes. So it goes back to the next one down.

It just so happens that 360/300 = 1.2 = 18/15 - so that would imply that the sectors per track takes care of the speed difference.

There was another thread on here with format issues - turned out to be density detection on pin 2 of the cable - I wonder if it is something like this ?

viewtopic.php?f=2&t=3015&hilit=flp_density

dex wrote:some 1.2 MB drives have switch to 300 RPM for compatibility purposes.

Thanks, this information is interesting. Actually there are some 1.2MB drives with the 300 RPM switch or jumper. Unfortunately the M2553K-83B does not have it. I couldn't find a specific datasheet and, for what I can see, there is nothing similar to a 300RPM indication on the board. A clever hack might exist, or might not. Who knows