Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

All questions about how to use KryoFlux go here.
exidyboy
Posts: 18
Joined: Wed Jul 24, 2019 2:11 pm

Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by exidyboy »

Hello again friends,

Thanks so much for your assistance last time with the 10 sector hard disks - some important Exidy Sorcerer software preserved as I noted in my final post.

This time hopefully something easier. I am trying to image around a dozen 77 track single sided 100TPI 5.25 inch disks for a later Exidy Sorcerer disk system. The disk system is the Exidy FDS (Floppy Disk Subsystem), developed by Exidy themselves, that did not require the S-100 Expansion Unit and S-100 controller card. Like the earlier system however it uses 100 TPI drives but in a soft sector format.

Note that I do have a 100TPI Micropolis 1053 drive mechanism so I am not trying to do some baroque reading of 100TPI disks in a 96TPI drive, in fact this is the actual drive that wrote the disks.

The issue I am having is that although the drive (soft of) responds to commands (ie seeks) and read one track but I am getting various errors:

1. I can't calibrate - it says "Error: Could not communicate with floppy drive" which is weird because the drive motor powers up, the head is loaded and the carriage seeks back and forth.

If I ignore the inability to calibrate and try and create a preservation stream I get an attempted read of the first track and then the capture aborts with "2 tracks: 0 good, 2 unknown" and the squares are grey.

Some things to note about this old school leadscrew drive:
a) it takes a long time to load the heads so perhaps that delay needs to be increased
b) the READY line is not where it is on later drives - does Kryoflux pay attention to the READY line at all?

Thanks so much and stay safe.

Michael

exidyboy
Posts: 18
Joined: Wed Jul 24, 2019 2:11 pm

Re: Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by exidyboy »

Some more info regarding the things I can think of that would make this drive different to a more modern drive:

1. I have tested a custom cable that correctly remaps the READY signal from the Micropolis drive to the conventional location and this has not changed the behaviour in Kryoflux-ui.jar

2. In terms of the drive specs:

Head loading time 75msec
Access time 30msec, track-to-track
Heading settling time 10msec

ZrX
Posts: 591
Joined: Tue Dec 06, 2011 9:09 pm

Re: Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by ZrX »

I'd say the ready signal is not used, but it could be some signal from the drive isn't activated until the head is loaded and settled and KF somehow times out.

If that's the case there's not much to do than to see if there are any ways to configure the drive somehow differently.

exidyboy
Posts: 18
Joined: Wed Jul 24, 2019 2:11 pm

Re: Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by exidyboy »

Thanks,

Can the timeout in the Kryoflux software be increased to allow for the use of drives that are slow to load the heads and that have a slow leadscrew mechanism for moving between tracks?

brightcaster
Posts: 249
Joined: Fri Nov 08, 2013 10:48 pm

Re: Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by brightcaster »

Are you really shure that your drive (which is a hard-sectored one) did once write the floppies (soft sectored as you said) that you are trying to read now? These old drives might behave weird if a floppy is inserted that has no sector holes...

Is it possible to calibrate with a hard sectored floppy inserted?

David

exidyboy
Posts: 18
Joined: Wed Jul 24, 2019 2:11 pm

Re: Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by exidyboy »

My understanding is that floppy drives themselves are neither hard-sectored nor soft-sectored?

But yes the drive mechanism wrote (most) of the discs originally and the drive mechanism can read all the discs. However they can't be read with the Kryoflux java client.

The drive can read the discs either when connected to a WD2793 to USB 'bridge' or to the WD1793-based ExidyFDS controller.

But obviously there is something about the drive that the Kryoflux client does not like. The most obvious way that I think it might behave differently than a more recent drive is the amount of time it takes to load the head initially and the track-to-track seek time which is a consequence of the leadscrew mechanism - it is very 'agricultural'.

It is also single-sided ie there is only head.

Thanks for your interest. I have multiple drive mechanisms and happy to do any tests that can be suggested.

brightcaster
Posts: 249
Joined: Fri Nov 08, 2013 10:48 pm

Re: Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by brightcaster »

You're right with the dumb drives, but didn't they use the index pulse to check if the drive is spinning or stucking? If the pulses don't come at the estimated times, they stop the motor as far as I remember (to avoid destroing the inserted floppy). That's why I thought testing with a hardsectored floppy should put some light on this issue....

David

ZrX
Posts: 591
Joined: Tue Dec 06, 2011 9:09 pm

Re: Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by ZrX »

The current hardsector supported DTC is also more strict about the index and gives up more quicker than the previous ones, but the would result it reporting no index signal detected instead of the no comms error.

exidyboy
Posts: 18
Joined: Wed Jul 24, 2019 2:11 pm

Re: Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by exidyboy »

These are soft-sector discs though - are you saying that by introducing hard sector supporting there has been a regression in soft sector imaging?

Anyway I had to do some Amiga discs tonight and noticed that if I tried to create a preservation stream and a .adf at the same time that I got the same problem ie the first track would be read then the process stalls - but I could create either format separately without any problem.

So I plugged the 5.25 back in and was able to image the entire disk as a preservation stream and as a lightly customised version of the standard MFM format *but only separately* - not at the same time. So twice the wear and tear on the discs.

So that is some progress. I dropped the preservation stream onto HxC floppy emulator and it did not show me any sectors but at least the heads can be driven across all 77 tracks now if I stick purely to a preservation stream.

ZrX
Posts: 591
Joined: Tue Dec 06, 2011 9:09 pm

Re: Imaging 100TPI 77 track Exidy Sorcerer disks on a Micropolis 100TPI

Post by ZrX »

What exactly was the command you used to image the disk in the first place?

And you can create imagefiles from the preservation files without re-reading the disk.

Post Reply