• Please review our updated Terms and Rules here

Can't figure out 360K floppy drive

ElectroArc

Experienced Member
Joined
Apr 22, 2023
Messages
51
Location
Georgia, United States of America
Hello all,

I'm a little new here, so please excuse me if I've posted this in the wrong place.

If at all helpful, the PC specs are:
160MB PC100 RAM
128gb SSD via a sata to ide adapter
Pentium 3 600MHz
Windows 98

Recently I got a Fujitsu M2551A marked as untested, but took a chance because they're supposedly reliable and worth repairing. Putting it in my then Windows XP computer, all things seemed well. Until I wasn't able to read floppy's, or even format them.

To check on things further, I installed windows 98, copied over ImageDisk, and ran some diagnostics.

Testing RPM is a bit odd, as the diskette just spins for a few seconds with near 0 info on screen, before kicking me back to the main menu.

The alignment test is also really odd, as it will read nothing but "?" no matter what I seem to do.

To ensure parts were actually moving, and not seized up, I opened the shroud so I could see things running inside. At a quick glance there don't appear to be any mechanical problems or physical damage.

I'll attach a picture of the alignment test, and the drive.

20230525_175646.jpg20230525_175706.jpgmc-20230525_175738.jpg

Perhaps I have the jumpers configured wrong? I really can't find ANY documentation on this drive, so any help is greatly appreciated!
 
Welcome to the forums. I am not the expert here, but there are many here that are. Someone will give you some guidance.
 
First of all Windows XP does not work with 360k drives. It can use 1.2mb drives to read 360k disks, but not an actual 360k drive. There was a bug in the XP driver and they decided to just not fix it. Generally Win9x is preferred for this task as some disk tools require real-mode DOS.

Is your BIOS setup configured for a 360k drive? Are you trying to use it as A: or B:? If B:, then does your motherboard and BIOS actually support two floppy drives?

What kind of floppy cable are you using? On this machine the cable should have a twist in it and your drive, if set as "A" should be on the connector after the twist.

I see your jumper is set for DS1. On this computer that is correct for both A: and B: configurations. There is no terminator resistor installed, but that is usually not a problem.

Have you sanity checked this computer to make sure another drive, even a 1.44mb drive, works?

So when accessing the drive, check the following:

Does the drive activity light come on?
Does the drive motor spin?
Does the head seek track zero? (Push it back to the center of the drive and at some point it should move back to the outer track)
Does the head otherwise move freely?

The imagedisk alignment test will (if the drive is active and spinning) let you position the head over different tracks. If the disk is formatted and it can read it, it will show you what track it thinks it is positioned over.
 
And, for heaven's sake, run ImageDisk on a real-mode DOS prompt--i.e. shut down to an MSDOS prompt, not an MS-DOS window. The Win98 virtualization really screws with timing.
 
First of all Windows XP does not work with 360k drives. It can use 1.2mb drives to read 360k disks, but not an actual 360k drive. There was a bug in the XP driver and they decided to just not fix it.
What about using the win2k floppy driver in XP? Will it load?
 
Welcome to the forums. I am not the expert here, but there are many here that are. Someone will give you some guidance.
Thanks for the warm welcome!

First of all Windows XP does not work with 360k drives. It can use 1.2mb drives to read 360k disks, but not an actual 360k drive. There was a bug in the XP driver and they decided to just not fix it. Generally Win9x is preferred for this task as some disk tools require real-mode DOS.

Is your BIOS setup configured for a 360k drive? Are you trying to use it as A: or B:? If B:, then does your motherboard and BIOS actually support two floppy drives?

What kind of floppy cable are you using? On this machine the cable should have a twist in it and your drive, if set as "A" should be on the connector after the twist.

I see your jumper is set for DS1. On this computer that is correct for both A: and B: configurations. There is no terminator resistor installed, but that is usually not a problem.

Have you sanity checked this computer to make sure another drive, even a 1.44mb drive, works?

So when accessing the drive, check the following:

Does the drive activity light come on?
Does the drive motor spin?
Does the head seek track zero? (Push it back to the center of the drive and at some point it should move back to the outer track)
Does the head otherwise move freely?

The imagedisk alignment test will (if the drive is active and spinning) let you position the head over different tracks. If the disk is formatted and it can read it, it will show you what track it thinks it is positioned over.
So, from the top, I didn't know that. I switched to 98 because I needed ImageDisk, and that only ran in DOS environments, and XP was an NT based OS.

My BIOS has it configured as a 360k 5.25 drive. Any other setting, and the BIOS will urge me to change it. The BIOS also only has settings for one drive.

The cable I'm using is a CablesOnline universal floppy cable. It has a twist in it. Also, I don't believe the drive select jumpers could be wrong because I did a blind test changing them, and any other configuration would have the BIOS yelling at me that something was awry.

I can't really test any other drives at the moment. The computer came with a 1.44mb 3.5 FDD, however, I don't have any diskettes for it.

When alignment testing in ImageDisk, the activity light comes on, the drive motor spins, and the drive head moves freely to each position I call for. Testing for zeroing, I moved the head to the 30th track, and restarted the computer, and the head instantly moved to the 40th track, and then back to 0.

As far as I know, the diskette in the drive is formatted, but I have no way to test that with this drive not working. It's a disk that came with an IBM clone, and is labelled, which makes me believe that there must be at the very least corrupted data that ImageDisk should be able to read.

And, for heaven's sake, run ImageDisk on a real-mode DOS prompt--i.e. shut down to an MSDOS prompt, not an MS-DOS window. The Win98 virtualization really screws with timing.
Yes, I read that online, and I have been running it in actual DOS mode.
What @SomeGuy said plus remove the RDY jumper and see if that helps bring it to life. If the bios moves the head to track zero at boot that's a good sign.
I tried removing the jumper, and still no dice. ImageDisk still outputs all "?"
 
First of all Windows XP does not work with 360k drives. It can use 1.2mb drives to read 360k disks, but not an actual 360k drive. There was a bug in the XP driver and they decided to just not fix it. Generally Win9x is preferred for this task as some disk tools require real-mode DOS.

Is your BIOS setup configured for a 360k drive? Are you trying to use it as A: or B:? If B:, then does your motherboard and BIOS actually support two floppy drives?

What kind of floppy cable are you using? On this machine the cable should have a twist in it and your drive, if set as "A" should be on the connector after the twist.

I see your jumper is set for DS1. On this computer that is correct for both A: and B: configurations. There is no terminator resistor installed, but that is usually not a problem.

Have you sanity checked this computer to make sure another drive, even a 1.44mb drive, works?

So when accessing the drive, check the following:

Does the drive activity light come on?
Does the drive motor spin?
Does the head seek track zero? (Push it back to the center of the drive and at some point it should move back to the outer track)
Does the head otherwise move freely?

The imagedisk alignment test will (if the drive is active and spinning) let you position the head over different tracks. If the disk is formatted and it can read it, it will show you what track it thinks it is positioned over.
I've tried formatting a completely blank diskette using ImageDisk, and it seemed to have gone through. Testing again with alignment testing, however, still comes up with more "?"
 
I've tried formatting a completely blank diskette using ImageDisk, and it seemed to have gone through. Testing again with alignment testing, however, still comes up with more "?"
Scratch that. I've tried switching to H1 from H0, and now I'm getting readings. However, it is rather sporadic. Some cylinders will be completely unreadable, others will be partially readable, and some will play weird tones from my speaker. On cylinder 5, it has practically composed a new song for me. At times it sounds like Yankee Doodle lol. The other head still doesn't work though. This makes me question if maybe the heads aren't clean, but I already went over them with some rubbing alcohol, and no dirt or grime came up on the cotton swab, so I just assumed it was good. Is there something else at play here, or does it just need a really rough cleaning?
 
When the drive is powered down, how easily can you move the head? Does it move easily and smoothly? If not, then you may want to clean and re-grease the rails for it.

Also, you could try cleaning the heads with some isopropyl alcohol and a ear bud (just wipe them gently and carefully and make sure no cotton is left behind).
 
When the drive is powered down, how easily can you move the head? Does it move easily and smoothly? If not, then you may want to clean and re-grease the rails for it.

Also, you could try cleaning the heads with some isopropyl alcohol and a ear bud (just wipe them gently and carefully and make sure no cotton is left behind).
The head seemed a bit gummy, so I stopped at walmart and picked up some B'laster white lithium grease. I cleaned the rails with some Isopropyl (70% is what my local store had. Is that ok? everyone keeps saying 91%) and then applied the grease with a cotton swab. It moves rather smoothly, with a little resistance at first. I also tried cleaning the heads again for good measure, and nothing came off.

Also, PROGRESS!

So, in my infinite wisdom, I ordered some new HD diskettes, which is not the right format. Realizing that, I also ordered some new old stock Sony DSDD diskettes. However, I had more of the HD diskettes than DD diskettes.

I had read online that HD floppy disks, while not reliable, can still be read and written to by a DD drive. So, since I had more, I decided I would experiment with them. However, perhaps that was wrong. Swallowing my pride, I finally tried formatting one of those Sony's, and, voila, it worked!

There is, though, a crap ton of beeping when running the alignment test. Does this mean that the head is misaligned? I don't understand why a freshly formatted disk wouldn't be aligned with the head that just formatted it though.

I also keep getting this error while running the Test RPM, which is an improvement from absolutely nothing, but still not good:

0/0: Missing sectors
: 250k DD - 2 sectors of 512 bytes - G1:200 G2:255
Cannot lock onto disk - press ENTER for main menu

I believe I am formatting the disk correctly for this drive.
To be sure, my settings are:

Drive: A
Cylinders: 40
Sides: Two
Double-step: Off
R/W gap: Calculated
Format gap: Calculated
Format fill: E5
Full analysis: No
Interleave: 6:1
Retries: 5
Keep bad sector: No
500kbps -> : 500kbps
300kbps -> : 300kbps
250kbps -> : 250kbps


When formatting, I used the settings:

Sector/Track? 1-255: 9
Start sector? 1-255: 1
Data rate/type: 250 kbps MFM
Sector size: 512


Am I perhaps doing something else wrong?
 
Last edited:
The IMD format doesn't really verify--it's pretty much a write-only operation, so no, it didn't work with the HD floppies. You may possibly get a few early tracks to format successfully (depends a lot on the disk) if you degauss the disk completely. But I wouldn't depend on that working.
 
Imagedisk alignment test beeps when it finds a readable sector. The idea is one can work on the drive and not have to look at the screen. You should see numbers appear on the screen indicating what sector number and track it has found. (That said, I've never seen one of these Fujitsu drives with an alignment problem. The alignment test is just a good way to test functionality)

If a DOS format succeed, then the disk should be OK. Does an empty directory appear in dos? Can you run scandisk or chkdisk on it? If you want to test Imagedisk with that, try reading the newly formatted disk in to an IMD file.

I don't know why that speed test is giving that error. See if anything else in Imagedisk is failing to read the disk.

Yea, high density 5.25" 1.2mb disks generally can't be formatted low density 360k.
 
Might these symptoms be related to not having the pullup resistor sip installed. I would expect the drive to have open collector outputs on the muxed signals, index, read_data, etc. Unless the host side has pullups there is nothing the drive can do to generate a logic high output signal.
 
Sharp eyes, @jxm! Yes, that empty 7 pin blue socket near the connector is meant for a 150 ohm terminator. Basically, 150 ohms each from pins 2-7 bussed to pin 1.
Very important, particularly if this is the only floppy on the cable. If it's the second floppy, you probably don't need a terminator, so long as the other drive is properly terminated.
 
The IMD format doesn't really verify--it's pretty much a write-only operation, so no, it didn't work with the HD floppies. You may possibly get a few early tracks to format successfully (depends a lot on the disk) if you degauss the disk completely. But I wouldn't depend on that working.
Oddly enough, it was doing the opposite and only reading tracks further out. Whether that means they were readable from the formatting or not, though; I don't know.

Imagedisk alignment test beeps when it finds a readable sector. The idea is one can work on the drive and not have to look at the screen. You should see numbers appear on the screen indicating what sector number and track it has found. (That said, I've never seen one of these Fujitsu drives with an alignment problem. The alignment test is just a good way to test functionality)

If a DOS format succeed, then the disk should be OK. Does an empty directory appear in dos? Can you run scandisk or chkdisk on it? If you want to test Imagedisk with that, try reading the newly formatted disk in to an IMD file.

I don't know why that speed test is giving that error. See if anything else in Imagedisk is failing to read the disk.

Yea, high density 5.25" 1.2mb disks generally can't be formatted low density 360k.
I see. I just saw that the numbers kept fluctuating and changing so much that I thought something must be wrong. It's good that that may not be the case. I'll wait to test further however, as...

Might these symptoms be related to not having the pullup resistor sip installed. I would expect the drive to have open collector outputs on the muxed signals, index, read_data, etc. Unless the host side has pullups there is nothing the drive can do to generate a logic high output signal.
Sharp eyes, @jxm! Yes, that empty 7 pin blue socket near the connector is meant for a 150 ohm terminator. Basically, 150 ohms each from pins 2-7 bussed to pin 1.
Very important, particularly if this is the only floppy on the cable. If it's the second floppy, you probably don't need a terminator, so long as the other drive is properly terminated.
I was unable to find a 7 pin 150 ohm resistor anywhere, and near equivalents (e.g. 220 ohm) were gonna ship from Taiwan. So, with me being incredibly impatient, I opted to buy regular 150 ohm resistors from Amazon, and solder them together in a network like a normal 7 pin resistor instead. It'll work, but just look a lot uglier. With today being Saturday evening, they should arrive on Monday, but I won't be able to get them until Tuesday.

Soooo... I guess we'll wait hahaha.

If there's any other suggestions in the mean time, I'll try whatever you guys have for me.

Thanks!
 
Last edited:
You can buy an 8 or even 10 pin bussed 150 ohm resistor network (the value isn't too critical--you can, for example, get away with 1K ones) and just let the nether pins hang out. Won't hurt a thing, other than looks.
 
You can buy an 8 or even 10 pin bussed 150 ohm resistor network (the value isn't too critical--you can, for example, get away with 1K ones) and just let the nether pins hang out. Won't hurt a thing, other than looks.
Making my own is still the fastest option. And, well, I already bought them, so I might as well get some soldering experience in hahaha.

I believe the polarity of the network is based on that dot on the board, correct?

Because I don't see any markings that would mark where the common output would go other than the dot, but the dot also doesn't exactly line up with one of the inputs.

Maybe the marking "DSS" is where?
 

Attachments

  • 20230528_133824(1).jpg
    20230528_133824(1).jpg
    2.5 MB · Views: 18
Don't know about DSS, but notice the thicker trace coming from pin 1 of the SIP. That should be connected to the +5 supply line and is the "top" of the bussed network so to speak. So the dotted end of the SIP aligns with that pin 1.
 
Alright everyone, it's Tuesday, so that means it's game day.

I soldered the resistors together, and it was actually not a bad learning experience.

I stuck the array in the floppy drive, and tested all 7 connections between the array and the board, with all showing up as connected through my multimeter. I also test each individual resistor, and they're all within standard.

However, it still wont work :(

Alignment test and Test RPM both still show the error about only having 2, 3, or 4 sectors. It varies now.

Using Alignment test, though, I can see that there really only ever is 2, 3, or 4 sectors. For example, it says there are only 3 sectors, and then it only ever reads sectors 3, 8, and 9.

As a sanity test, I formatted a fresh floppy diskette with 9 sectors, and did the alignment test again. This time it detected 4 sectors; 2, 5, 8, and 9.

I don't know what could possibly be causing this.
 
Back
Top