Are you sure it is compatible with the IBM PC's format ?

That's what I've found on the Internet:

Double-sided double density (2D) 256 bytes/sector, MFM, 16 sectors/track, 40 tracks
256 x 16 x 40 x 2 = 327,680 bytes, commonly known as 320kB. This is a common format for domestically produced 8-bit computers, and the built-in drives of PC-80S31K and PC-8801mk2 also use this format.

Double-sided double density (2D) 512 bytes/sector, MFM, 9 sectors/track, 40 tracks
512 x 9 x 40 x 2 = 368,640 bytes, commonly known as 360kB. It was used on a PC.

Maybe it is the direction?

I cannot be sure if it is compatible with the IBM PC's format but it was advertised as 1.2MB, not 360kB.
I do have a SHARP X1 (8-bit) using a 1.2MB so I'll try swapping the drives.

Vendest wrote on 2023-10-14, 08:48:

I cannot be sure if it is compatible with the IBM PC's format but it was advertised as 1.2MB, not 360kB.
I do have a SHARP X1 (8-bit) using a 1.2MB so I'll try swapping the drives.

I mean Japanese Floppy Disk Drives used weird formats. I guess same was true also regarding 1.2MB ones.

It is a 96tpi 1.2Mb floppy drive, some model 380's cannot read 360k floppies because they are locked single speed 360 rpm, not dual speed. Make sure in bios you have it set as a 1.2Mb...

I have seen YD-380 but never a YD-380T.

I would have to check but I think I have a YD-380 that works on IBM PC compatible machines.

Vendest wrote on 2023-10-14, 04:59:

I moved some jumpers around, the DS0-3 selection and the 150 ohm terminations.

I'm assuming you have a standard floppy cable with a twist at the end? The drive needs to be set to DS1, and frankly the head solenoid should not engage unless the drive is selected (by the FDC), but possibly your jumper configuration is set to head load on motor on which would do that. Not really an issue for a PC but kinda pointless to have the solenoid at all if it is always engaged along with the motor.

Note, since you want to use this drive as B:, you have to plug it into the middle of the cable (before the twist), and on some BIOSes you need to have the A: drive if you want to use B: as well.

Thanks for the advice, I have tested the drive to the best of my knowledge.

2 things came clear during testing:
- The drive can only support 1.2Mb disk.
- The drive does not respond when set to drive A or 0.

1. IBM PC compatible
Tried with 2 known working cables, connector with twisted pair connected to the 1.44Mb A drive and connector in the middle connected to the B drive (YD-380T)
Checked all the jumpers and figured it out (I think) based on the documentation available for other drives and testing. Ended up with jumper DS1 for B drive, HS for the solenoid, jumper 5 for the LED.
Any other jumper were removed as there was either no or negative impacts.
Drive B: 1.2Mb set in bios
Boot in DOS 6.22 from drive A, trying to access drive B "Not ready". The comment seems to appear almost immediately while the drive is getting ready.
This is where I tweaked the jumpers, tried different floppy disks and cables.
I have no clues how to get past this issue.

2. Greaseweazle v4
The drive works fine only when set to drive 1 with jumper DS1, using a cable with a single connector and twisted pair.
The drive was able to format and write.

3. Shapr X1 - 8-bit computer
The computer comprises of 2 x 5,25 disk drive 0 and 1.
The drive is not detected when used as drive 0 (jumper DS0) but is detected when used as drive 1 (jumper DS1)
In 320Kb mode, the computer can only load the program header but not the full content as the drive is not 360kB compatible.
I need to further check in 1.2Mb mode now that I have the correct jumper settings.

FYI, on PC, the floppy drives are Always set for DS1. The twist in the cable is there for that reason. You shouldn't have to change the DS jumper. just leave it at 1.
Does the Sharp X1 have a twist in the cable ?

Vendest wrote on 2023-10-15, 07:16:

Thanks for the advice, I have tested the drive to the best of my knowledge. […]
Show full quote

Thanks for the advice, I have tested the drive to the best of my knowledge.

2 things came clear during testing:
- The drive can only support 1.2Mb disk.
- The drive does not respond when set to drive A or 0.

1. IBM PC compatible
Tried with 2 known working cables, connector with twisted pair connected to the 1.44Mb A drive and connector in the middle connected to the B drive (YD-380T)
Checked all the jumpers and figured it out (I think) based on the documentation available for other drives and testing. Ended up with jumper DS1 for B drive, HS for the solenoid, jumper 5 for the LED.
Any other jumper were removed as there was either no or negative impacts.
Drive B: 1.2Mb set in bios
Boot in DOS 6.22 from drive A, trying to access drive B "Not ready". The comment seems to appear almost immediately while the drive is getting ready.
This is where I tweaked the jumpers, tried different floppy disks and cables.
I have no clues how to get past this issue.

2. Greaseweazle v4
The drive works fine only when set to drive 1 with jumper DS1, using a cable with a single connector and twisted pair.
The drive was able to format and write.

3. Shapr X1 - 8-bit computer
The computer comprises of 2 x 5,25 disk drive 0 and 1.
The drive is not detected when used as drive 0 (jumper DS0) but is detected when used as drive 1 (jumper DS1)
In 320Kb mode, the computer can only load the program header but not the full content as the drive is not 360kB compatible.
I need to further check in 1.2Mb mode now that I have the correct jumper settings.

According to DOS Days (gallery at the foot of the page) - https://www.dosdays.co.uk/topics/floppy_drives.php - the 380T is merely a black version of the 380C

Also, this thread on VCF (actual info on links to Russian language forum) - https://forum.vcfed.org/index.php?threads/whe … assembly.58523/ - indicates anything other than 1.2Mb support requires modding the motor flex to allow 300 - 360rpm for other formats

weedeewee wrote on 2023-10-15, 07:45:

FYI, on PC, the floppy drives are Always set for DS1. The twist in the cable is there for that reason. You shouldn't have to change the DS jumper. just leave it at 1.
Does the Sharp X1 have a twist in the cable ?

Ok got it, it makes sense.
The Sharp X1 does not have a twist in the cable.

Some progress after further testing in Win98, the drive is now accessible.
But it is still not accessible in DOS 6.22 and DOS mode "not ready"

Posting a picture of the final jumper settings that works for me.

Have you tried leaving a disk in the drive and then booting the computer ?

Vendest wrote on 2023-10-15, 09:24:

Some progress after further testing in Win98, the drive is now accessible.
But it is still not accessible in DOS 6.22 and DOS mode "not ready"

I find it odd that you need to remove the pull-up/terminator pack jumper. Can you measure the resistance of the pull-up pack on that drive? Note, you need to measure it vs +5V line, not GND. In fact it would be best to make sure all the inputs to the drive are properly terminated.

In general the not ready error is either due to READY and DISK CHANGE singnals being output wrong, or some issue with track zero sensor but that would also trip Greaseweazle right away, as well as Win98. Since this setup kinda works with the terminator jumper removed I assume you have another drive (as A:) on the cable as well, and that one is properly terminated? Because the issue could be with the terminator pack, or a stuck output, on the other drive.

Try connecting this drive as A: at the end of the cable, and no other floppy drives. Reconfigure BIOS, possibly also close the jumper for termination. Does this work? If not set the drive to 360k in BIOS and try again, did anything change? It might not read floppies but does it try now or error out right away?

Deunan wrote on 2023-10-15, 11:56:

I find it odd that you need to remove the pull-up/terminator pack jumper. Can you measure the resistance of the pull-up pack on that drive? Note, you need to measure it vs +5V line, not GND. In fact it would be best to make sure all the inputs to the drive are properly terminated.

In general the not ready error is either due to READY and DISK CHANGE singnals being output wrong, or some issue with track zero sensor but that would also trip Greaseweazle right away, as well as Win98. Since this setup kinda works with the terminator jumper removed I assume you have another drive (as A:) on the cable as well, and that one is properly terminated? Because the issue could be with the terminator pack, or a stuck output, on the other drive.

Try connecting this drive as A: at the end of the cable, and no other floppy drives. Reconfigure BIOS, possibly also close the jumper for termination. Does this work? If not set the drive to 360k in BIOS and try again, did anything change? It might not read floppies but does it try now or error out right away?

Thanks for your help, I checked the pull-up resistor pack is 150 Ohm. I have a standard 1.44Mb 3.5" connected as A: and the 1.2Mb as B:.
I connected the 1.2Mb alone as A: with and without 150 Ohm termination but the drive fails on boot during POST.
The drive selection in bios does not have the 360kb option, however I tried it earlier on another motherboard with drive set a B: it failed during POST as well.

Maybe I should not leave the pins in a floating state.

Vendest wrote on 2023-10-15, 13:53:

Maybe I should not leave the pins in a floating state.

That you should not. If that drive is now a single on the cable, the jumper pack that connects the resistors to the I/O lines must be closed/installed. Do try a mobo with 360k option (and the resistors connected), that should tell you if the DC line on pin 34 is working properly, or is it perhaps configured to output RDY which will trip DOS. Greaseweazle does not care about disk change signal and Win98 might have some workarounds built-in for "weird" hardware so that's why it seems to work properly.

I suspected this is an older model (seeing how it has plenty of ICs instead of just one or two SMD ones) and might have 150 ohm pull-ups. You can read about the floppy drive bus termination on Wiki but the long story short is the older drives had a jumper or removable resistor pack with resistance of 150 ohm, and the new ones have permanently connected packs of about 1k (up to 10k). You can have only one 150 ohm pack for the entire cable, and preferably at the end of it, while the 1k ones can coexist and not overdrive the bus. Even 150 along 1k might not work properly.

Worse, I've read about some mobos and I/O cards that don't really work well with 150 ohm termination, don't have the required current drive to output correct L/H voltages on the bus. Could be whatver mobo or card you are trying to use with that drive is like that. But you do need to terminate the bus so try installing the jumper pack first, and if that doesn't seem to work leave it open but connect another another (newer) drive on the cable as well and have it be the termination. That will prevent you from trying out the drive in stand-alone configuration though, that's in case the other drive is actually somehow messing things up.

Deunan wrote on 2023-10-15, 14:15:

That you should not. If that drive is now a single on the cable, the jumper pack that connects the resistors to the I/O lines must be closed/installed. Do try a mobo with 360k option (and the resistors connected), that should tell you if the DC line on pin 34 is working properly, or is it perhaps configured to output RDY which will trip DOS. Greaseweazle does not care about disk change signal and Win98 might have some workarounds built-in for "weird" hardware so that's why it seems to work properly.

I suspected this is an older model (seeing how it has plenty of ICs instead of just one or two SMD ones) and might have 150 ohm pull-ups. You can read about the floppy drive bus termination on Wiki but the long story short is the older drives had a jumper or removable resistor pack with resistance of 150 ohm, and the new ones have permanently connected packs of about 1k (up to 10k). You can have only one 150 ohm pack for the entire cable, and preferably at the end of it, while the 1k ones can coexist and not overdrive the bus. Even 150 along 1k might not work properly.

Worse, I've read about some mobos and I/O cards that don't really work well with 150 ohm termination, don't have the required current drive to output correct L/H voltages on the bus. Could be whatver mobo or card you are trying to use with that drive is like that. But you do need to terminate the bus so try installing the jumper pack first, and if that doesn't seem to work leave it open but connect another another (newer) drive on the cable as well and have it be the termination. That will prevent you from trying out the drive in stand-alone configuration though, that's in case the other drive is actually somehow messing things up.

Thanks again for your advice and useful information, I have done additional testing based on your recommendations and more.

With the drive connected on another motherboard thru a cable with a single connector and twist, drive set as A: with and without the pull-up resistor pack and set as 360k then 1.2m in bios. The drive does not budge at all, fails during POST.
Then I tried with 3 different 3.5" drive as A and the 5.25" as B , with and without pull-ups. DOS reported "extended error 21" which seems to be similar to "not ready" error.
Also tried to boot DOS from 5.25" drive by enabling drive swap in Bios. The disk was not detected as bootable.

I then decided to check the drive with a dedicated DOS/Win3.11 unit that I use for gaming and disk imaging. The unit is equipped with a modern ISA FDC card supporting 4 drives (Quad-Flop) and currently working fine with 3 drives 1.44m, 1.2m and 360k.
The drive behaves the same in DOS "not ready" and Win3.11 "no disk detected"

Moving back to the Win98 unit, the drive set as B still works in Win98 with or without pull-ups. I must admit that that was the initial intent to have a 1.2m available in my Win98 unit, just that it bugs me that the drive will not be available in DOS mode.

The resistor pack is not socketed unfortunately but I could attempt to replace it with a 1k.

Vendest wrote on 2023-10-16, 15:46:

With the drive connected on another motherboard thru a cable with a single connector and twist, drive set as A: with and without the pull-up resistor pack and set as 360k then 1.2m in bios. The drive does not budge at all, fails during POST.

So, it fails POST both with and without the TD jumper pack? And that's both when set as 1.2M and as 360k? Is the error something like "FDC failure" or something else?
Can you test it (TD jumper pack installed, drive set as 360k, then 1.2M) with a floppy already inserted at power-on? Does it still fail the same way and is there any difference at all between 1.2M and 360k settings - like head movement or attempt to spin on one setting but not the other?

I have 3 theories but 2 of them are very unlikely as the drive does work in Win98 and with Greaseweazle. So it's not track 0 sensor issue and most likely not that INUSE signal is required but missing. By the way can you check where pin 4 of the floppy connector goes to, it's connected to the 150 ohm pull-up so must be used (though optional). Is it connected to the INU jumper perhaps?

Please check were pin 34 is connected. I think I see a missing jumper marked "R" nearby, is it permanently open or closed (copper trace connecting the pads)? Some chances are this particular drive was never meant to be used in a typical PC and there is no way to configure #34 to be DC instead of RDY, or perhaps it's not connected at all.

EDIT: BTW for a PC you should use this pinout rather than Shugart one: http://www.interfacebus.com/PC_Floppy_Drive_PinOut.html

Deunan wrote on 2023-10-16, 16:39:

So, it fails POST both with and without the TD jumper pack? And that's both when set as 1.2M and as 360k? Is the error something […]
Show full quote

So, it fails POST both with and without the TD jumper pack? And that's both when set as 1.2M and as 360k? Is the error something like "FDC failure" or something else?
Can you test it (TD jumper pack installed, drive set as 360k, then 1.2M) with a floppy already inserted at power-on? Does it still fail the same way and is there any difference at all between 1.2M and 360k settings - like head movement or attempt to spin on one setting but not the other?

I have 3 theories but 2 of them are very unlikely as the drive does work in Win98 and with Greaseweazle. So it's not track 0 sensor issue and most likely not that INUSE signal is required but missing. By the way can you check where pin 4 of the floppy connector goes to, it's connected to the 150 ohm pull-up so must be used (though optional). Is it connected to the INU jumper perhaps?

Please check were pin 34 is connected. I think I see a missing jumper marked "R" nearby, is it permanently open or closed (copper trace connecting the pads)? Some chances are this particular drive was never meant to be used in a typical PC and there is no way to configure #34 to be DC instead of RDY, or perhaps it's not connected at all.

EDIT: BTW for a PC you should use this pinout rather than Shugart one: http://www.interfacebus.com/PC_Floppy_Drive_PinOut.html

Motherboard used for test TMC TI5VG+

Pin 4 goes to TD1 pull-up, INU and HL.
Pin 34 connects to jumper/trace R which is closed and connects to the HD7438P logic chip.

The drive is not detected in 360k mode.

Hm. Was the A: test done as stand-alone (just this drive) at the end of the twisted cable? It's weird that it failed completly. But for now lets focus on what works.

I suspect the floppy drive error you get with drive set to 360k is because BIOS does a seek and detects more than 40 tracks. Look for a BIOS option that says "Floppy seek" or "Boot up floppy seek" and set it to disabled. This will prevent BIOS from moving floppy drive heads and discovering that "360k" drive is actually a HD one. If this can't be disabled then perhaps there is a way to continue booting even with the error (press F1 for example). You want to be able to boot to DOS to further test the drive set as 360k. You might need a DD/360k floppy or HD one formatted to 360k to be able to test - in fact two of those with different files written to both. Or just one file with different name, anything that would show a difference when the floppy is swapped and DIR command is run. And you don't want too many files, just a few, to make sure the floppy can be recognized just with the first track, to avoid any errors due to 40/80 track difference.

Deunan wrote on 2023-10-17, 10:14:

Hm. Was the A: test done as stand-alone (just this drive) at the end of the twisted cable? It's weird that it failed completly. But for now lets focus on what works.

I confirm drive A: test done as stand alone.
I have now tried the drive as per your suggestions:
- Seek floppy drive disabled in bios
- Set to 360k
- Tried with one 360k disk formatted in a 48 tpi drive and one 1.2M formatted as 360k in a 96 tpi drive.
After booting into DOS, it gives an error "General failure reading drive B"
It is a common failure I'd get on my DOS PC when a badly formatted disk is used in the drive.

I set the drive back to 1.2M again, then the usual error "Not ready reading drive B" was back.