• Please review our updated Terms and Rules here

720kB and 1.44MB drive compatibility riddle with 720kB floppy disks

1

1200XL M.U.L.E.

Experienced Member
Joined
May 9, 2021
Messages
84
My Epson Equity 1+ has a 720kB floppy drive made by Mitsubishi. I bought it from eBay.

My PS/2 has a 1.44MB floppy drive. I'm not sure who made it or if it is the original drive.

I can format a 720kB floppy disk and read and write to it on the Equity 1+. No problems.

I can read files from that same floppy on the PS/2 with its 1.44MB drive. The same floppy is also readable by my USB based 1.44MB floppy on my Windows 10 PC.

Now here is where things get sticky and where I need some help.

If I write anything to that 720kB floppy on my PS/2 or the USB drive then that floppy become unreadable in my Epson Equity 1+.

Any 720kB floppy formatted in the PS/2 or with the USB based drive on my Windows 10 PC is unreadable by the Equity 1+. I am using genuine, name-brand 3.5" DSDD disks, like Verbatim, 3M, etc and the behavior is consistent.

Is this one-way kind of compatibility to be expected or is there something wrong with the Mitsubishi drive in my Equity 1+?

Thanks!
 
Epson has their Product Support Bulletins online still. Those don't have the answer to this question but I suggest downloading it if you haven't already. https://files.support.epson.com/pdf/e1____/e1____ps.pdf

Can the PS/2 or USB drive read a 720 KB floppy written by the USB or PS/2 drives? Just an obvious check. What OSes are in use? Can the Epson read a diskette formatted or written to by either the USB or PS/2 drives but not the Epson?

I suspect based on the symptoms that the Mitsubishi drive is slightly out of alignment compared to the other drives. Realigning a drive is hard. The solution used to be to get a replacement drive. Note that if one is lucky, the drives might only have issues when modifying a disk while if all formats and writes for a disk are done on one drive, the other drive can still read it.
 
I suspect based on the symptoms that the Mitsubishi drive is slightly out of alignment compared to the other drives.
Click to expand...
This. Fits exactly to what you describe.
 
Supposedly the IBM PS/2 1.44 drives ignore the density hole in the floppy disk. Could this cause the IBM drive to write to an already formatted 720K floppy using the wrong DENSEL? Or does DOS or the controller set that correctly using /REDWC (pin 2 on a normal drive) via software, even though the drive doesn't have the hardware to to detect the density hole?
 
As for the PS/2, most of the line employed "host density select"; the LSI FDCs (see the Intel 82077 datasheet, section 7.0 for an example) could be configured as "media sense" or "host select". 1.2MB 5.25" disks, for example are always "host select" as there is no distinguishing physical indication of the density. However, almost all clones and third-party PCs use "media sense" for 3.5" media.

The IBM PS/2 thing, in my opinion was a grave error that we still are dealing with. When HD 3.5" media first came onto the scene, it was perhaps 4x more costly than the DD disks. PS/2 users quickly discovered that pennies could be saved by simply formatting the DD disks as HD. Of course, the disks couldn't be read in "media sense" drives on non-PS/2 systems without punching or boring a hole in the drive jacket to inform the drive that this was high-density. For a brief time, there were kits offered to do just this.

I still occasionally receive the PS/2 DD disk formatted as HD--time has not been kind to these. It isn't just the coercivity of the coating that matters (something that's often erroneously cited as evidence of no difference in the media), but it's also the oxide particle size and coating thickness. Back in the day, I consulted the 3M NML people on this subject and they weren't very optimistic about the survival rate of DD media formatted as HD. In fact, I've got one oddball setup where the HD media formatted as DD outperforms the native DD stuff.

The whole mess continued into the ED 2.88M arena later. Some PS/2 models allowed for formatting HD media as ED. I can't point to a single specimen of that stuff that's survived the test of time. ED disks are usually coated with a barium ferrite emulsion, not iron oxide for a very good reason. The price of ED media never really came down to the level of the HD stuff, so it died the death it deserved. Back in the day, a box of 10 floppies would set you back $50 (in non-inflated currency).

Fast forward to today and cheap USB drives. Many vendors take the attitude of "Who the heck still uses 720K media?" and remove that capability (and the sensing of DD media) from their drives, aiming for the rock-bottom price of handling HD floppies. In fact, if you were to buy a carton of no-name Chinese (usually black slimline) USB drives, you'd find that several different models of various manufacturers were used in the same lot. I recall picking up three of the things at the same time from the same vendor and finding out (via lsusb) that one had a NEC drive, the second had a Teac and the third had a Sony. The outside of each drive bore the same label.

To IBM's credit (and perhaps chagrin), the later PS/1 line all used the media-sense convention.
 
Hmm, I'd want to try a few more tests. Try reading the unreadable disks back in on the Equity using a tool like ImageDisk, TeleDisk, or Disk2img. Find out if only the written sectors have become unreadable, or if there is more going on here.

Also, test the drive rotation speed of each drive with ImageDisk (Well, ImageDisk won't work on the USB floppy drive).

What version of DOS are you running on the Equity? Some early OEM versions can do odd things. I'd recommend using a "vanilla" MS-DOS 3.3, 5.0, or 6.22 for testing.

But to answer the last question, as long as that USB drive supports 720k, then all three of those drives should be able to exchange floppy disks just fine.
 
My two cents is on alignment issues; the tracks written by the old drive are offset so they’re almost a half-step off from the other two drives; said drives can catch enough of the track written by the Epson’s drive to read them, but when they write back it’s only half-covering the previous data area so when you carry it back to the Epson it sees a mix of the old and new data.

An interesting experiment if you have the means to bulk erase a disk so it’s genuinely and truly blank of old data would be to format the disk 720k in one of the other drives and see if the Epson can read that. With no old data to confuse it my guess would be it would have a chance. But then if it wrote to it the other two machines would likely get confused.
 
I have seen drives that couldn't write successfully but it seems unlikely that two different drives would have similar write problems.

I rather doubt the drives would be writing high density data to a double density formatted disk without having to do a reformat. Having two drives write tracks offset enough that the result can't be read is a common enough problem.
 
Wow, lots of replies! Thanks, everyone! :)

Let me try to answer the questions here.

krebizfan The PS/2 and Win10 machine can read and write each other's disks regardless where they were formatted. I can format on the PS/2, write on the Win10 with the USB drive, and read it all back on PS/2. The reverse works too. I can format on the Win10 with the USB drive, write a bunch of files on the PS/2, and read it all back on the Win10 machine.

The PS/2 is running MS-DOS 6.22 and the Equity 1+ is running MS-DOS 3.30.

Chuck(G) Thanks for posting all that history. That kind of makes me suspect 1.44MB drives in general. Like you said, all to save a few pennies.

SomeGuy Epson has a diagnostic tool that checks the rotational speed of the drive. The tool registered 299.9 RPM. I will try to look for the disk imaging tools you mentioned. My Win10 machine has an unregistered version of WinImage. Would that program work here?

So ... even with all that, it looks like the bottom line here is saying the Mitsubishi drive is out of alignment. If a different 720kB floppy drive works perfectly here then I know the Mitsubishi drive is busted. However, if I see the same behavior with another 720kB drive then am I looking at a broader format compatibility issue?
 
Chuck(G) said:
As for the PS/2, most of the line employed "host density select"; the LSI FDCs (see the Intel 82077 datasheet, section 7.0 for an example) could be configured as "media sense" or "host select". 1.2MB 5.25" disks, for example are always "host select" as there is no distinguishing physical indication of the density. However, almost all clones and third-party PCs use "media sense" for 3.5" media.

The IBM PS/2 thing, in my opinion was a grave error that we still are dealing with. When HD 3.5" media first came onto the scene, it was perhaps 4x more costly than the DD disks. PS/2 users quickly discovered that pennies could be saved by simply formatting the DD disks as HD. Of course, the disks couldn't be read in "media sense" drives on non-PS/2 systems without punching or boring a hole in the drive jacket to inform the drive that this was high-density. For a brief time, there were kits offered to do just this.

I still occasionally receive the PS/2 DD disk formatted as HD--time has not been kind to these. It isn't just the coercivity of the coating that matters (something that's often erroneously cited as evidence of no difference in the media), but it's also the oxide particle size and coating thickness. Back in the day, I consulted the 3M NML people on this subject and they weren't very optimistic about the survival rate of DD media formatted as HD. In fact, I've got one oddball setup where the HD media formatted as DD outperforms the native DD stuff.

The whole mess continued into the ED 2.88M arena later. Some PS/2 models allowed for formatting HD media as ED. I can't point to a single specimen of that stuff that's survived the test of time. ED disks are usually coated with a barium ferrite emulsion, not iron oxide for a very good reason. The price of ED media never really came down to the level of the HD stuff, so it died the death it deserved. Back in the day, a box of 10 floppies would set you back $50 (in non-inflated currency).

Fast forward to today and cheap USB drives. Many vendors take the attitude of "Who the heck still uses 720K media?" and remove that capability (and the sensing of DD media) from their drives, aiming for the rock-bottom price of handling HD floppies. In fact, if you were to buy a carton of no-name Chinese (usually black slimline) USB drives, you'd find that several different models of various manufacturers were used in the same lot. I recall picking up three of the things at the same time from the same vendor and finding out (via lsusb) that one had a NEC drive, the second had a Teac and the third had a Sony. The outside of each drive bore the same label.

To IBM's credit (and perhaps chagrin), the later PS/1 line all used the media-sense convention.
Click to expand...
Not to mention the demand for cobalt.
 
I had the same issue with my Tandy 1400LT. The A drive was slightly out of alignment, leading to odd compatibility issues with the other systems, including the USB drive attached to my Win 10 system. Swapping it with the B drive in the same system solved the problem, and I didn't need two floppy drives so that's as far as I got with the troubleshooting.
 
Chuck(G) said:
Not following you. 2.88M ED media is BaFe₁₂O₁₉--just like the stripe on your credit card. At least that's what my box of 3M ED diskettes claims.
Click to expand...
I am talking about HD floppies. Edit: Though this is a good time to mention that cobalt is probably far common than barium, but...
 
The problem lies in the way that the heads were designed. 720 K floppies, have a larger head than a 1.44. That's how you can get more data on the disk, by using a smaller head area, which can access the higher density tracks on 1.44 media (aka HD media).

What a 1.44 drive DOS formatting does when formatting the disk, is use either the single or double density stepping of the head. You can actually override the density selector control by using an option switch with the FORMAT command. The low density will step the head position motor twice as much as the high density which will step the motor half as much. Then, it verifies the format to see if it can retain the selected format.

Remember, that the disc was only certified for 720 tracks, not the 1.44 high density tracks. Obviously, you can TRY and fit a 1.44 on a 720. If the 720K media is of a high enough quality, then you can actually force it to think that it has a 1.44 high density disk.

This has NOTHING to do with cobalt vs whatever. Those folks who wanna talk about what metals are used on a plastic medium, are blowing smoke up your dress, because they don't really know the answer. But they WANT to sound important, so I'll give them that much...they SOUND important.

Truth is, we didn't care about what metal was on the disk back in the day. It was either a high density or a low density disk. It was all about how dense the material was placed on the plastic. Period. Those folks are probably thinking of chromium tapes versus iron oxide tapes, which is a totally different technology. Spinning versus dragging technology. First thing that comes to mind is the word "goober".

What you are seeing is the 1.44 drive, which has a smaller head, only writing on HALF of the 720 disc track. This is NOT a drive alignment problem! This is because the 1.44 head is half of the width of the 720 and it cannot possibly write and cover the entire 720 track. That is the nature of the beast.

When you write on a 720 with a 1.44 drive and then take it back to a 720 drive, the 720 head is reading two sets of data, the "old" (from the 720 head) and the "new" (from the 1.44 "half-track" head) because the 1.44, half sized head, only writes on half of the track, not the entire thing. Therefore, the Equity cannot possibly read the information without getting errors.

This has NOTHING to do with drive alignment and curse those people who say that it is. They simply do not understand the technology involved. I was a tech in the 1980s and 90s, IBM certified. I serviced a lot of these drives, and had to field this question MANY times. Yes, you can TRY to realign the heads on a 1.44 to try and match up with the 720 or vice versa. I made a lot of money from people that insisted I do this for them. However, the problem still persists and if you change drives, it's gonna come back and haunt you. So much so, that it was in our service agreement that we were not responsible for realigned drives, specifically when it came to the performance with other drives.

You're really just complicating things by realigning the drive. The drive is actually perfectly aligned. The 1.44 head is only writing on half of the 720 track, BY DESIGN! When you realign a drive to try and fix this problem, all you're doing is misaligning the 1.44 to try and cover the entire 720 "wide" track. Again, because the 1.44 head is half the width of the 720, at best you're going to get marginal results.

You HAVE to use a 720 floppy drive to read and write reliably on 720 floppy. There's no way around this, and the same problem exists between a 360K and 1.2M 5 1/4 inch floppy drive.

The main reason the disc is complely unreadable on the Equity, is because the disc's FAT table has become corrupted, again because you have only written on HALF of the 720 track. Usually, if you run a check disk on the floppy, using the 1.44 drive, it will come back and tell you that there are serious errors on the disk that cannot be corrected.

<mic drop>
 
Last edited:
Sometimes you have to wonder.

Consider the Teac FD235HF (most common Teac HD drive that I can think of). It can step 80 tracks at 135 tracks/inch or a track spacing of 0.1875mm.
The stated recording width is here stated as 0.115mm after guard erase (you need this to separate the tracks and avoid inter-track noise). So the total size of the guard band is 0.0725 mm.
Now, let's go to the granddaddy of 720K 3.5" drives, the Sony 0A-32D, a full-height drive with a 26-pin interface. OEM manual here Note that the track recording width is given as 0.125mm with a guard band width of 0.063mm.
In short the recorded track width is slightly wider (0.0095mm or 9.5µM). I suspect that if I run across a spec sheet for the FD-235F (720K), the recorded width will match that of the FD-235HF.

So what's the difference? (I've gone over this multiple times and have documentation from the 3M National Media Laboratory to back me up. Namely, coating thickness, particle size and less importantly, coercivity. The same situation applies to the ED 2.88M recording method, save that a different formulation is used to obtain coercivity in the area of 1200 œ and a particle and coating size that works with perpendicular recording. Track width remains the same.

I can see where the difference between 48 and 96 tpi 5.25" drives gets conflated with the 3.5" drive picture, but they're not the same, aside from the magnetic coercivity of the two media.

If someone can offer hard evidence otherwise, I'd be delighted to review it.
 
You must log in or register to reply here.
Back
Top