Looking for someone with a single sided STF

[npomarede]

Hello

I'd like to see what happens at the FDC level when trying to read side 2 of a floppy on a single sided drive (what values are returned in that case)
Is there someone that could do some quick tests by running a test program (on an STF with an internal single sided drive) ?

Nicolas

[npomarede]

Ping
No one with a single sided STF ?

[Hippy Dave]

I'll pull out my old Atari ST 520 (after 15+ years) it has 2 external single sided floppy drives. Hopefully it runs, and I could do tests.

[Hippy Dave]

I dusted it off and it don't boot up exactly right. Monitor works like new though. Could be a flakey power supply or corroded IC pins. I'll work on it in my spare time.

[charles]

THE DRIVES ONLY HAVE ONE PHYSICALL READ/WRITE HEAD , WHERE THE DOUBLE SIDED HAS TWO .

I GUESS IF YOU CAN READ ASM 68K THE INTERNALS ABACUS HAS THE TOS DISASSEMBLED IN THE BACK ..

CHARLIE

[Hippy Dave]

Reseated the ICs and now it boots from floppy, mouse works too!
Gimme there test program.

[npomarede]

THE DRIVES ONLY HAVE ONE PHYSICALL READ/WRITE HEAD , WHERE THE DOUBLE SIDED HAS TWO .

I GUESS IF YOU CAN READ ASM 68K THE INTERNALS ABACUS HAS THE TOS DISASSEMBLED IN THE BACK ..

CHARLIE

I guess I can read asm
But the problem is not that : even if a drive is single sided, you can tell it to read on the 2nd side. Of course, you won't get a correct track, but you should receive some bits. The question is : "are those bits randoms or all "0" or all "1" ?"

[npomarede]

Reseated the ICs and now it boots from floppy, mouse works too!
Gimme there test program.

Great !

For now, I don't think I need to provide a custom program to test this, using Panzer by Dr CoolZic should be OK to read on this drive :
http://www.atari-forum.com/viewtopic.ph ... 25#p237793
Just put the PRG and the RSC on a single sided floppy, then run panzer.prg

In options/general, you can choose drive B
Then start commands / Read Track (with a floppy in drive B)
If you choose "side 0", it should work.
If you choose "side 1", command should complete, but the buffer should not contain valid data. Can you tell what you get ? Random bytes ? If you do several read track, do you always get the same data ?
After each read track (for 5-6 attempts), can you save the track buffer (using buffer / write buffer) ? I'd like to send those track buffers to kryoflux guys to check if there's a known pattern.

Thanks

[Hippy Dave]

When I run PANZER.PRG it complains that it cannot find PANZER.RSC on my real Atari 520 ST
I noticed that the .rar archive has PANZER.PRG and panzer.rsc
I 've spent 2hrs using (1)mtools and (2)mounting and (3)renaming and (4)formatting and
(5)determining that other programs work on real Atari and (666)looking at a hex dump of PANZER.PRG
Maybe my .rar program is brittle.
Maybe PANZER.PRG can be improved.
Maybe you can email me a genuine tarball with a working PANZER

[npomarede]

I sent it to you by mail (older version) ; when trying it on my STF, I had no error about the RSC file, but I will check if I converted names to upper cases or not.

Nicolas

[Steven Seagal]

On a single side drive like the poor SF354, side changes are just ignored and the drive sends data from/to the only side it knows. Steem assumes that so, of course, it's correct.

[npomarede]

On a single side drive like the poor SF354, side changes are just ignored and the drive sends data from/to the only side it knows. Steem assumes that so, of course, it's correct.

I don't think that's the correct behaviour. All programs I tested that chek for single/double sided drive do a read sector on side 1 and check if this gave a RNF error in the status register.
If they always read side 0 as you suggest, they would never get RNF, yet they manage to detect a drive is single sided; so reading side 1 on a single sided drive will certainly return some data (random ones I guess).
Hence the test I'd like to make on a single sided drive.

[Hippy Dave]

Reseated the ICs and now it boots from floppy, mouse works too!
Gimme there test program.

Great !

For now, I don't think I need to provide a custom program to test this, using Panzer by Dr CoolZic should be OK to read on this drive :
http://www.atari-forum.com/viewtopic.ph ... 25#p237793
Just put the PRG and the RSC on a single sided floppy, then run panzer.prg

In options/general, you can choose drive B
Then start commands / Read Track (with a floppy in drive B)
If you choose "side 0", it should work.
If you choose "side 1", command should complete, but the buffer should not contain valid data. Can you tell what you get ? Random bytes ? If you do several read track, do you always get the same data ?
After each read track (for 5-6 attempts), can you save the track buffer (using buffer / write buffer) ? I'd like to send those track buffers to kryoflux guys to check if there's a known pattern.

Thanks

I have a TOS 1.00 Atari ST 520 with no RF and no internal floppy drive.
I have two external single sided SF354 floppy drives.

The only way to have PANZER.PRG and panzer.rsc work is to have both
drives connected, and use Panzer in the second drive (drive B:)

Data read from Track Zero, Side One (side 1 is the Top side with No head)
appears to be random. See enclosed data.

This data was collected with no floppy in drive A:
Panzer was in drive B:
Track Zero, Side One of Drive B: was read to get the data.

[npomarede]

Thanks for the time you spent on this, those indeed look to be random pattern, even if we see lots of "00" bytes and often bytes with only 3-4 bits set to 1.
I will send this to Kryoflux for analysis.
At least, this makes my single sided drive support in Hatari correct and is coherent with the fact that the FDC will return Record Not Found when trying to access a sector on 2nd side.

As for the fact Panzer worked only from drive B:, maybe Dr Coolzic will read this thread and can have an explanation/fix for it.

Nicolas

[Steven Seagal]

If it's a fact it's a fact, but it seems more trouble to support this with plugins (pasti, capsimg).

[npomarede]

If it's a fact it's a fact, but it seems more trouble to support this with plugins (pasti, capsimg).

I already discussed this with IFW, so it should be available in next capsimg lib.
As for pasti, you're indeed stuck without the sources of the lib. But I tested Hatari's STX support with several games in single sided mode and the results are the correct ones (see this thread http://www.atari-forum.com/viewtopic.php?f=3&t=26403 for examples of games/demos that benefit from an optional second side)

Nicolas

[Steven Seagal]

Alright, I'm not gonna try to hack it with pasti.dll. I'll wait that ijor improves it

[IFW]

Good luck with that
In the meantime, the WDC emulator in the IPF library now supports single sided drives returning random garbage for side 1.

[ijor]

Yes, reading the second side on a single sided drive will return garbage. Many games depend on this feature, mostly by design, a few by mistake.

This is, of course, a drive feature, not strictly an FDC or computer feature. It even depends on the exact wiring of the ST connector adapter.

Sorry, don't have Pasti sources with me at this time. Replying just from what I remember. Just here by chance, not sure when I'll be back, sorry.

Hi Nicolas, hi Steven, hi Istvan. It is great to see you still improving this stuff! Take care all of you

[AtariZoll]

Surprise of the day ! What day, a year ! I guess, next one will be in 2018

P.S. all best Jorge. Watch SyFy on this week Fridey night

[npomarede]

Hi Ijor
I guess you have little time, so thanks for posting anyway.
Hopefully, Hatari's implementation of Pasti doesn't have too many bugs, but your insight is of course welcome.

Nicolas

[IFW]

Yes, good luck

[Steven Seagal]

Hi ijor
Who knows, there could still be a new build of pasti.dll after all.