• Please review our updated Terms and Rules here

Suddenly, 1.2MB drive won't read 360K

T

Tincanalley

Experienced Member
Joined
Aug 25, 2019
Messages
217
Location
Los Angeles, CA
This drive was working fine a day ago. I used it to boot 360K DOS disk to install a Hardcard. I even loaded files to the newly installed Hardcard from the drive off 360K floppies. Today I tried to install a game from 360K floppies and got a general drive read failure. I though the disk was bad, but when I tried any capacity other than 1.2, it failed. The second 1.2 drive works just fine with both capacities.

How does the drive know a 360K is in? I recently cleaned the heads and wonder if I messed something up in regards to disk identification. When I run IMD and test speed with a HD disk, it works just fine and displays speed. When I try the same test with a DD, it tries, but never determines speed and just drops back to previous menu. When I do the alignment test with a HD disk, it comes up error free on all tracks and heads. If I put in a DD, it has a question mark for track and can't perform test.

I've had this drive since '91 and never had an issue with it. Nothing has changed with the drive configuration, so it has to be something with the drive. Any suggestions where to look?
 
What drive brand and model do you have?
Normally, 5,25" 1.2MB drives use pin 2 to switch to high density mode. On a PC, this is +5V when HD, and GND if DD disks are in. There's no notch on the media to indicate density.

Since you've ruled out the cable and the controller by testing a different drive and media, and the offending drive continues to work with 1.2MB diskettes, the problem will most likely be with the logic board regarding this density signal.
Use a spare 360K floppy you don't mind erasing, and try to observe if the drive did not start changing RPM when trying to format this floppy. Watch the spindle, listen to the motor and compare that to when a 1.2MB floppy is in.
If it does seem to spin slower on 360K, some jumper might have been misplaced, or there's a problem with some logic gate. On a PC the drive needs to spin at 360 RPM constantly, regardless of density.

You can test this with IMD in alignment mode: put a 360K disk in, wait for the question marks, press P to set format parameters, set 9 sectors per track of 512 bytes, a data rate of 300 kbps MFM. Then press F to format and D to attempt to read the data. Afterwards, try the same process, but set the data rate to 250 kbps MFM. The disk should only be readable on the first try. If it reads on the second try, the RPM is being changed. If neither work, I suspect an issue with the internal handling of the HDSEL line.
 
Last edited:
Long shot but I think that DD vs HD disks differ when it comes to how the drive grips the disk around the center hole.

If the speed test just tests the drive without actually reading, there shouldn't be any difference between DD and HD.

Does it seem like 360k disks spin like they should?

A super long shot would be that sticky shed syndrome disks have left residue on the part of the drive that grips the disk, and it slips on your 360k disks but not on the 1.2M disks due to differences in the disk types. (IIRC one of them usually has some sort of reinforcement around the center hole, the other type usually doesn't).
 
MiaM said:
If the speed test just tests the drive without actually reading
Click to expand...

It does indeed read; there's a prompt to "insert a formatted diskette" before the test.
An ID field scan command is performed; and the time it took for the same sector to reappear (i.e. one disk revolution), is then averaged out.
This is because all PC floppy disk controllers implement the scan ID command, but only a handful provide means to inspect the /INDEX signal line itself. And, for a scan command to work, the controller must recognize the sectors on a disk, so the RPM test won't even execute on a drive that would otherwise read fine, but with RPMs grossly out of spec.
 
Well I realized I skipped a step in testing. I was so busy trying to figure out why it fails with reading 360s I never bothered to test formatting. I guess it never dawned on me since it reads 1.2s all day long, so how could it be some kind of alignment issue, blah, blah, blah.

I am able to format 360s in it, and read them as well. Can even use IMD Alignment test and RPM as well. Not a single error in the Alignment test. However, the disk is only readable in that drive. No other 1.2 or 360 will read it. So how can there be an alignment issue on the 360 side, but not the 1.2? Both densities have the same track zero.

Oh yeah, it's an Epson SD-600
 
All 5.25" 1.2MB drives used in PCs are 80 cylinder, i.e. with a density of 96 tracks per inch. 360K media on a PC were supposed to be used with 48 TPI (40-cylinder) drives, i.e. each track was physically wider.
So what essentially happens when you try to write 360K disks in a 1.2MB drive is, that it only records one half of the width of a track. And this caused problems, as you've noticed. It's best not to combine 40-cylinder formats in 80-cylinder drives, and if that's the only option, limit it to read operations only.

If the drive creates disks that only it can read, then it's an alignment problem, and that's finicky to perform without good equipment, especially on 80-cylinder drives.
Take a spare HD floppy, set 80 cylinders and use the Erase operation in IMD to destroy all the old sectors, then format it on the offending drive, with 15 SPT at 500kbps MFM (i.e. high density recording). Then take it out to a different 1.2M drive, and try to read it into an IMD image (the Alignment command "beeps" are sometimes not enough, unless you press the D key to read Data of each and every sector of a track).
My 2 cents is that the drive with the wrong alignment will also mess up 1.2M disks, not just 360K - if it's an alignment problem.
 
I'm not 100% sure about this, but I've read rumors about 1.2M drives that had double heads of sorts, one narrow and one wide, and switched heads depending on the density signal. If there is something weird going on with that type of switching, if that even is a thing, then it might explain your problem. Perhaps.
 
There is one other unlikely possibility. The 360K read/write mode needs to be at 300 kbps instead of the normal double density 250 kbps. The disk is spinning faster so the data needs to be written faster. One could test for that by writing a disk to capacity and seeing if tracks aren't complete. That would involve a very early and non-AT compliant floppy controller.

What is the system and floppy controller under consideration? Does the floppy drive still fail if it is swapped with the other 1.2 MB drive? Just to see if the connector was damaged.

Was the BIOS changed in adding the Hardcard?
 
Good point!

IIRC some drives were configurable, and the drive might be configured for use with some other system.

P.S. the speed is so slow that you might even be able to count rotations manually while a stop watch counts down for a minute, to get a rough RPM reading. Or you could use a simple voltage divider and feed the index signal to a line input of a modern computer and analyze the frequency of the pulses using for example Audacity.
(Btw I haven't tested this myself but somewhere I read that sound cards tend to read all the way down to DC on their inputs. Not that it matters in this case, the index pulses should still be fully visible, but still).
You might even be able to hold a microhone near the drive and record that and determine the rotation speed from the variations in sound when a disk spins. Maybe.
 
So many possibilities!

I tested the cable and controller by replacing the drive with a Teac FD-55GFR and it works for both disk types and IMD is able to test speed and alignment for both as well.

The BIOS is so basic it doesn't even know there's a hardcard installed. Nothing has changed on the system other than installing a WD MFM controller to test a couple drives. When nothing about it worked, I removed the controller and the system is back to previous configuration. All testing point directly to the drive as any other 1.2 I have works on the machine it its place.

I have a 5150 I can use to write to a 360K. So should I fill a disk with files and then put it in the 1.2 and what am I looking for? Just whether it reads it or not?
 
deanimator said:
All 5.25" 1.2MB drives used in PCs are 80 cylinder, i.e. with a density of 96 tracks per inch. 360K media on a PC were supposed to be used with 48 TPI (40-cylinder) drives, i.e. each track was physically wider.
So what essentially happens when you try to write 360K disks in a 1.2MB drive is, that it only records one half of the width of a track. And this caused problems, as you've noticed. It's best not to combine 40-cylinder formats in 80-cylinder drives, and if that's the only option, limit it to read operations only.

If the drive creates disks that only it can read, then it's an alignment problem, and that's finicky to perform without good equipment, especially on 80-cylinder drives.
Take a spare HD floppy, set 80 cylinders and use the Erase operation in IMD to destroy all the old sectors, then format it on the offending drive, with 15 SPT at 500kbps MFM (i.e. high density recording). Then take it out to a different 1.2M drive, and try to read it into an IMD image (the Alignment command "beeps" are sometimes not enough, unless you press the D key to read Data of each and every sector of a track).
My 2 cents is that the drive with the wrong alignment will also mess up 1.2M disks, not just 360K - if it's an alignment problem.
Click to expand...
Standard DOS format won't accomplish the same? If it's an alignment issue, wouldn't it have issues reading and writing a factory formatted disk?
 
The idea of a simple confirmation that the data rate is what it should be would be to write 360K disks in the 1.2 MB drive. If the data rate was too low, the end of a written track would overwrite the beginning of the track. Compare what was written to what was read. If there is a difference. the drive has a major problem.

DOS format will write the tracks wherever the drive says the tracks should be. If the drive is out of alignment, all the resulting tracks will be out of alignment. Note that floppy drives have considerable excess for the tracks so most drives can read disks from other drives even without perfect alignment.
 
You must log in or register to reply here.
Back
Top