Is that Windows 9x on the PC monitor ?! Beware, Windows 9x is a floppy killer! 😟

https://www.os2museum.com/wp/the-ihc-damage/

Windows 3.1x and WinImage 3 might be safer for making/restoring disk images and file transfer. Or Windows NT/2k/XP.

Jo22 wrote on 2023-04-24, 09:59:

Is that Windows 9x on the PC monitor ?! Beware, Windows 9x is a floppy killer! 😟

https://www.os2museum.com/wp/the-ihc-damage/

Windows 3.1x and WinImage 3 might be safer for making/restoring disk images and file transfer. Or Windows NT/2k/XP.

Thanks for the heads up..
It is windows 98 on the monitor...but I drop to DOS to do the format..
Hopefully that is safe ?
rgds

Vipersan wrote on 2023-04-24, 09:11:

I remember punching an extra hole and using the other side of the disk for commodore drives...so theoretically both sides of a S […]
Show full quote

I remember punching an extra hole and using the other side of the disk for commodore drives...so theoretically both sides of a SS DD disk could be used in a PC effectively turning a SS DD a DS DD without modification.
I have tried this and it does appear to work.
How reliably is debateable...only time will tell.
🤣

That was only necessary on the Commodore-drives because they have only one head. PC drives have two heads and already use both sides. (The Commodore 1571 also has two heads)

HanSolo wrote on 2023-04-24, 10:14:

Vipersan wrote on 2023-04-24, 09:11:

I remember punching an extra hole and using the other side of the disk for commodore drives...so theoretically both sides of a S […]
Show full quote

I remember punching an extra hole and using the other side of the disk for commodore drives...so theoretically both sides of a SS DD disk could be used in a PC effectively turning a SS DD a DS DD without modification.
I have tried this and it does appear to work.
How reliably is debateable...only time will tell.
🤣

That was only necessary on the Commodore-drives because they have only one head. PC drives have two heads and already use both sides. (The Commodore 1571 also has two heads)

Understood..
The point I was getting at is that although a Floppy is actually labeled SS DD it doesn't mean it cant be used in a PC which requires both sides..
Chances are that both sides are useable ...and both sides coated even though the disk claims single sided.
That said...maybe only the Single side is 100% reliabilty guaranteed when new.
cheers

Sorry, I misunderstood what you actually meant.
Yes, disks are coated on both sides but might not be certified. However, I never found a definition for which side is #1 and which is #2. The C=1541 accesses the underside and not the one with the label as one might think 😀

Jo22 wrote on 2023-04-24, 09:59:

Is that Windows 9x on the PC monitor ?! Beware, Windows 9x is a floppy killer! 😟

https://www.os2museum.com/wp/the-ihc-damage/

Windows 3.1x and WinImage 3 might be safer for making/restoring disk images and file transfer. Or Windows NT/2k/XP.

The bane of floppy disk preservationists with kryoflux or Greaseweazle...such a shame.
This tool helps with restoring OEM name and revert some other Win9x changes in image files: https://github.com/Digitoxin1/DiskImageTool
Sectors will still show up as modified, though. No way around that.

Boohyaka wrote on 2023-04-24, 12:04:

Jo22 wrote on 2023-04-24, 09:59:

Is that Windows 9x on the PC monitor ?! Beware, Windows 9x is a floppy killer! 😟

https://www.os2museum.com/wp/the-ihc-damage/

Windows 3.1x and WinImage 3 might be safer for making/restoring disk images and file transfer. Or Windows NT/2k/XP.

The bane of floppy disk preservationists with kryoflux or Greaseweazle...such a shame.

Aye. Kryoflux would be neat indeed. 😎
For most things an normal image file is sufficient, though, I suppose.
(Unless copy-protection comes into play.)

It provides a logical, digital copy of a diskette (w/ sectors and FAT), much better than merely saving a file/directory structure (the old method many PC people used for years) . .

Kryoflux format is more like a wave file, an attempt at storing "analogue" values in digitized form. It can also store defective sectors, as used by some copy protection schemes. And "data" between tracks.

Unfortunately, there's no Kryoflux software available for DOS or Windows 3.x, OS/2 or Amiga so far, while *.IMAs (or *.ADFs, respectively) can be created/restored on those vintage platforms.

All in all, using Kryoflux is like using a 78 shellac record player instead of a CD player or MiniDisc, essentially.
It will pickup all those little hissing sounds, the humming of the motor, the synchronization error, etc.

The second one remembers me of CatWeasel controller, which was available for ISA sloths.
That controller could read Amiga floppies and other formats under DOS, I vaguely remember.
It also contained an Atari joystick port and a SID chip, AFAIK.
There also was a PCI version, I vaguely remember.

Boohyaka wrote on 2023-04-24, 12:04:

This tool helps with restoring OEM name and revert some other Win9x changes in image files: https://github.com/Digitoxin1/DiskImageTool
Sectors will still show up as modified, though. No way around that.

Thanks for the link for the tool! 😃👍

Yes indeed Jo22..
grabbed a copy myself..
I'll check it out at some point.
I do have a dedicated PC for archiving.
..and use Kryoflux ...and various tools for flux imaging.
Hardware wise I have the Kryoflux ...a Supercard pro...a Greaseweazle...combined with various drives...
I've archived a lot of floppies...but never found anything truely rare.
Guess I'm stuck in a timewarp...cos I still love floppies.

Jo22, not fully following you there in regards to KryoFlux and images? KryoFlux (and Greaseweazle for that matter) do record low level raw magnetic flux transitions, but also create working image files straight from the DTC software (downloadable for free off the KF website). It is the most accurate way to actually image any kind of floppy (so to be clear, not only to grab magnetic raws, but also produce working image files).

For illustration to preserve my DOS originals I select the kryoflux output+the appropriate image profile so I dump the raws and also produce an .IMA of the floppy in one go.

I mentioned it in response to your link about Win9x changes, because that's a known issue for people trying to preserve original floppies, ideally you want them as "original" as possible which means free from any change.
That OEM boot sector stupidity from Microsoft is really annoying because nobody was really aware (not would have cared, most probably) back in the days, and simply browsing a (non write-protected) floppy on any Win9x computer at any point of time and it was forever modified.

As I do own a lot of Sierra originals that I'm imaging with my Kryoflux, there's another somewhat similar annoyance: when installing a game from floppy 1, and picked options during setup, it would write resource.cfg to the floppy and complain if your disk was write-protected, and that even if you installed the game on hard-disk. Which means all disk 1's from used games are very likely to have modified sectors.

Kryoflux is able to detect when a track was modified. Here's an example from one of my Sierra games, an otherwise pristine 5.25" 360kb disk 1:

track00.0 shows as modified, that's Win9x OEM bullshit.
the other track showing as orange is the resource.cfg file created by setup.
all green tracks are read perfectly and haven't been tampered with since being written for the first time.

Obviously this is very much some preservation perfectionism, and does not matter that much in the big picture. A slightly modified dump is probably fine, and infinitely better than no dump at all.

I was recently gifted a Teac FD-55GR.

It is working but seriously misaligned.

Luckilly I have a working one to compare with.

I have little to no experience with IMD..but am convinced this is the tool to use...to attempt to get it working.

I have taken a few photos showing track 0 and track 10 ..head 0 and head 1...running a known good HD 5.25" floppy.

Alignment test..

The speed test shows both drives are ok speed wise...but the head alignment is obviously wrong on the bad drive...obvious even to me with my limited understanding from the photos.

Obviously both heads are good...but aligning it will be a challenge.
I have nothing to lose ..so why not try.
Anyone know of any youtube videos specifically aimed at alignment ?
The bad drive is seriously messed up.
Good drive >>

and the bad drive...

So from what I have observed so far..
The drive starts ok - ish at track zero by pushing the heads manually back to start...the selecting an arbitrary track number..say 50...
The heads move to 50..
Then using the track zero option the heads move back to start position ..but never get there..
coming to rest at track 4 -ish.
Not zero as requested.
something preventing this.
Could be sticking on the sled rails...or possibly the stepper motor faulty ?
I admit I am a little confused here...as it is possible to push the head manually back to zero from 4...
..so it is free to move ???
I guess the track zero sensor could have moved thus it reads track 4 as zero...?

Boohyaka wrote on 2023-04-25, 11:21:

Jo22, not fully following you there in regards to KryoFlux and images?

Long story short, I meant to confirm that Kryoflux is a fine luxury for saving disk images.
However, its perfection is rarely needed in practice. Not that it is overkill per se, just very sophisticated.

Emulators and virtual floppy drives rarely require a raw image, nor do they have the processing power to handle them in real-time.

They usually default to using sector-by-sector formats, anyway, which equals what WinImage, Teledisk, Disc Copy Fast etc use.

I assume that 90% of the users who save random floppy image don't have access to a Kryoflux, anyway.
Rather, they're users using vintage hardware w/ floppy drives or a modern PC w/ 1,44 MB USB floppy drive.

So I think it's important that they're being supported by the community first - through advices, help at trouble shooting etc. 🙂

Edit: To be precise, who I'm thinking of are all those possible volunteers out there who have precious floppies stored in their attic, garage etc.

Asking them to get hold of an 1,44 USB drive and download/run WinImage to make an image is possible.
Those drives are cheap and can be found in various places, maybe at the workplace, even. WinImage is easy to use, too, even for laymen/laywomen.

The resulting IMA is very small (below 2MB), can be easily mailed (10MB limit on free accounts) or attached to a forum post (5MB limit) .

However, asking them to buy a Kryoflux and a vintage floppy drive first is a bit too much.
Asking them to send over the floppy disk(s) for dumbing, is, too. It causes unnecessary stress.
In the end, we may loose volunteers if we demand that people always use Kryoflux for accuracy.

Edit:

Boohyaka wrote on 2023-04-25, 11:21:

[..] As I do own a lot of Sierra originals [..]

Understood. Though for Sierra games like KQ3, there used to be Sierra Unprotect (SUP).
And that utility itself is nearly a vintage classic in its own reign by now.
No absolute need to adhere to keep using a prestine executable that does funny stuff.

Edited.

Has the PCB on the underside been removed at all? Maybe that's not your problem since you can get pick an arbitrary track and get there successfully, but maybe it's a Track 0 problem. Actually, what you've just described where you're manually moving the head to 0 then seeking to 50, should mean it's a Track 0 problem.

The underside PCB with the molex / floppy connector should be what the TP0 sensor is attached to and it slides backwards and forwards in the screw holes. That's a factory calibrated fitting I think.
If it has been removed - that's a mistake I made, it'll need to be adjusted using a known good disk. Beware that getting TP0 adjusted to Track 0 is relatively easy, but getting the alignment good enough to work over all of the tracks (is it 80?) can be tricky.

Additionally, looks like you may already know it but in the alignment test of IMD, press H to switch the heads and see if they're in the same place. If H0 gets a good result but H1 gets a bad result then it's not a Track 0 sensor problem and you would need to adjust the top head vs the lower head, which takes a while. If H0 and H1 get the same results, do not touch the adjustment screws on the top head, that's really not a fun job to get back into alignment.

Soooo....close I can taste it..
This drive has been seriously messed with...
Used the pcb underside screws to adjust for track zero...only to discover track 1 was no longer producing an output...so adjusted the heads topside.
Then back to adjust tack zero underneath...then back to upper side to adjust the heads...and I have eventually after much flipping top side to bottom side...got it to read the directory on the disk.
Chances are this is not stable or good enough yet...but sooooo close.
...a lot of variables....
A big shout out to IMD for pointing me in the right direction.
..and yes Thermalwrong..both adjustments have been messed with imo.
Head 0 and Head 1 were NOT lined up at all...and the metal screening can from over the top head is missing..( an indicator of interference ? )
I really wish I owned a factory alignment disk......
Perhaps IMD could make one on the good working TEAC

Looks like track zero is now correct...and head 0 ...which I assume is the lower head reads fine up to track 69 wheras the upper head (head 1 ?)looses alignment above track 29...so more work to do.
certainly a very critical setup.
This aint easy but maybe I'll nail it with persistance.

I'm using a blank floppy formatted on the good TEAC...but watching the track numbers and data returned on the bad TEAC...
looking for a consistant (30) which I can now achieve up to track 66.
I guess this is pretty damned close as the good TEAC playing the floppy it formatted drops data above track 70.
Getting closer may not really be practical without access to an alignment disk...and use of a scope.
This approach is definately hit and miss...but you can get pretty close and I guess acceptable results by eyeballing it....and a lot a patience.

Vipersan wrote on 2023-04-26, 23:16:

I guess this is pretty damned close as the good TEAC playing the floppy it formatted drops data above track 70.

That's not normal. You've got something wrong with the "good" setup if that's the case

Vipersan wrote on 2023-04-26, 23:16:

Getting closer may not really be practical without access to an alignment disk...and use of a scope.
This approach is definately hit and miss...but you can get pretty close and I guess acceptable results by eyeballing it....and a lot a patience.

Sorry to hear it causes so much trouble.
My first thought was using a scope for monitoring the tracks, too.
Ironically, this would be a real use case for a Kryoflux - dublicating an alignment floppy.
I'm not sure, though, if someone ever dumped such a maintenance floppy before.
If so, one of the Kryoflux owners could send you a physical copy if it.