• Please review our updated Terms and Rules here

5.25" floppy disks with bad sectors - any solutions?

Mike, perhaps we define "overly verbose" differently. I have yet, in all my travels of the vast internet, to find an explanation of floppy disk drives and floppy disk media that completely, clearly and **in language that I will not then have to go and immediately research, sending me on a wild goose chase** gives me an understanding that prevents such misunderstandings as "DSDD = 1.2MB". I am by no means criticizing your input in this thread - I am explaining why I don't already "get it" and as a result why I still don't "get it" entirely.

That is fine but I think when we start throwing out the chemical composition of the magnetic layer or the ratio of binder to magnetic material we have probably gone too far. For the purposes of this thread it is enough to note that the material is different and thus requires different levels of magnetic force to manipulate.

I don't think I misled in the statement that Chuck was replying to and the elaboration is interesting, but the phrase "It's far more complicated than that, Mike" kind of irks me because it presumes that I don't know that it is more complicated and that I am leaving out something important.

If somebody doesn't stop at a sensible point, it will turn into a chemistry and industrial engineering class, which is what happens over at the Classiccmp mailing list. Am I to respond next with a related physics lesson? (Don't worry - I can't.)
 
That is fine but I think when we start throwing out the chemical composition of the magnetic layer or the ratio of binder to magnetic material we have probably gone too far. For the purposes of this thread it is enough to note that the material is different and thus requires different levels of magnetic force to manipulate.

I don't think I misled in the statement that Chuck was replying to and the elaboration is interesting, but the phrase "It's far more complicated than that, Mike" kind of irks me because it presumes that I don't know that it is more complicated and that I am leaving out something important.

If somebody doesn't stop at a sensible point, it will turn into a chemistry and industrial engineering class, which is what happens over at the Classiccmp mailing list. Am I to respond next with a related physics lesson? (Don't worry - I can't.)

Lol, alright, point taken (and agreed upon).
Basically, this whole business of tracks cylinders inches sides densities...it's all a jumbled mess in my head (and maybe that's how it is supposed to be, maybe this never made sense to anyone and trying to understand it is futile?)

It's probably a whole lot simpler than I make it. Similar story with how I (didn't) learn multiplication in grade school...I had a war with the teacher about how memorization is NOT learning, and that I wanted to learn how to do it properly, and in order to do that I should not be memorizing the results but instead memorizing the process.

As a result, I was never taught multiplication. Completely unrelated to the topic at hand, but might give someone a laugh, so there it is.
 
Lol, alright, point taken (and agreed upon).
Basically, this whole business of tracks cylinders inches sides densities...it's all a jumbled mess in my head (and maybe that's how it is supposed to be, maybe this never made sense to anyone and trying to understand it is futile?)

It's probably a whole lot simpler than I make it. Similar story with how I (didn't) learn multiplication in grade school...I had a war with the teacher about how memorization is NOT learning, and that I wanted to learn how to do it properly, and in order to do that I should not be memorizing the results but instead memorizing the process.

As a result, I was never taught multiplication. Completely unrelated to the topic at hand, but might give someone a laugh, so there it is.

My understanding is as follows:
48tpi DD = Double Density, 160KB/180KB/320KB/360KB
96tpi HD = High Density, 1.2MB

However, HD disks won't allways format correctly if formated as DD disks. In fact, most HD disks I have tried reported loads of bad sectors when I tried it. In the same manner, DD disks can't be formated as HD disks.

About your 96tpi DD disks, I think those may be a hybrid, capable of being formated as either HD or DD. Do those disks give you errors when you try to read them?

the format command you should be using goes as follows:
Code:
FORMAT B: /F:[size] /U
 
Ohh, alright. Thanks.
Now the question is, is it my HD drives not supporting 720k, or is it the DOS 7 format command not supporting 720k (or is it both?)

I try format b: /f:720 and it says not supported.

Your drive and bios may not support it. Unfortunately, a relatively uncommon format from a decade earlier will fall prey to the relentless effort to save fractions of a penny.

There were some 5.25" drives that did support both the 1.2 MB and 720kB formats with the ability to easily shift tpi and rpm as needed for disks. Specialized drivers would probably also be needed. I hope someone else can recommend methods that work.

I hesitate to point to the following web page since if the author is a poster here, the author is much better suited to answer questions:

http://www.oldskool.org/disk2fdi/525HDMOD.htm

But if you are lucky, you might be able to modify your drive to read the quad-density disks.
 
77 cylinders is...an 8 inch floppy?

While 77 cylinders is a standard 8" format, there exists an older 5.25" format, recorded at 100 tpi with 77 cylinders. Sometimes, it's called a "Micropolis" format, after the manufacturer that pioneered it, other manufacturers made the drives too. The Tandon TM-100-4M ("M" for Micropolis; a TM-100-4 is a 96 tpi drive) is one such drive. Commodore used the drives briefly.

Mike, forgive me for going into too much detail. I'd assumed that the reason for these discussions was to educate--why would one collect old gear if not to understand it? Maybe I don't understand this collecting business...
 
Chuck - I don't think I got irked by the info. I think I got irked that the implication was that I left out something important to the immediate discussion.

And the broader question is, where do we draw the line on what's an appropriate level of detail? This thread is by far and way more interesting that many of our 'my PC is broken' threads, but at what point does the extra become noise? ClassicCmp is a perfect example of people making this unreadable for the sake of completeness ..
 
Ohh, alright. Thanks.
Now the question is, is it my HD drives not supporting 720k, or is it the DOS 7 format command not supporting 720k (or is it both?)

I try format b: /f:720 and it says not supported.

What's going on here is that 720k 5.25" disks are not a standard format on PCs. The /f:720 option is intended for 3.5" disks, so FORMAT will refuse to make a 5.25" disk at that capacity. There are third-party programs that will format 5.25" 720k disks.

HD drives themselves are perfectly capable of using quad-density floppies, which are just 80-track disks at the DD bit rate. Tandy 2000s aside, QD was used by some CP/M machines as well as Commodore 8050 and 8250 drives.
 
What's going on here is that 720k 5.25" disks are not a standard format on PCs. The /f:720 option is intended for 3.5" disks, so FORMAT will refuse to make a 5.25" disk at that capacity. There are third-party programs that will format 5.25" 720k disks.

HD drives themselves are perfectly capable of using quad-density floppies, which are just 80-track disks at the DD bit rate. Tandy 2000s aside, QD was used by some CP/M machines as well as Commodore 8050 and 8250 drives.

Maybe that's why all the disks reports bad sectors when formated on a PC? What if all the disks has been formated in (and for) another system, and that disk-format confused the FORMAT program supplied by DOS?

Try to format one of the DD 48tpi ones using the command:
Code:
FORMAT B: /F:360 /U
 
HD drives themselves are perfectly capable of using quad-density floppies, which are just 80-track disks at the DD bit rate. Tandy 2000s aside, QD was used by some CP/M machines as well as Commodore 8050 and 8250 drives.

Didn't the DG One also use 720K 5.25"? (That's one portable you don't see much about here, BTW)
 
Kishy loves replies to read and reply to :)

My understanding is as follows:
48tpi DD = Double Density, 160KB/180KB/320KB/360KB
96tpi HD = High Density, 1.2MB

However, HD disks won't allways format correctly if formated as DD disks. In fact, most HD disks I have tried reported loads of bad sectors when I tried it. In the same manner, DD disks can't be formated as HD disks.

About your 96tpi DD disks, I think those may be a hybrid, capable of being formated as either HD or DD. Do those disks give you errors when you try to read them?

the format command you should be using goes as follows:
Code:
FORMAT B: /F:[size] /U

The "Cross-formatting" (HD to DD or DD to HD) kind of makes sense...the physical media, at least as I see it, doesn't correspond to the locations that the drive will try to access (yay or nay?)

The 96tpi DD disk (it looked like I had two but it was because they both had the same brand label on them, but only one is actually 96tpi DD) doesn't seem to want to work. I haven't tried putting anything on it and then trying to take it off, but logic says that would probably work fine because the same drive is doing the writing and reading, and it would only write to areas that it thinks are good.

Depending on what format I try to force on it, the "bad sectors" as detected by Scandisk will vary, but there are ALWAYS a lot of them. It's obvious that either a) the disk is actually damaged, b) the wrong format is being used resulting in those "fake" bad sectors or some sort of mix of both.

I'm not sure I want to try an unconditional format...if it is refusing to do it, wouldn't that possibly result in drive damage if the drive was instructed to, for example, seek to a location it doesn't know how to seek to (for example smash the r/w heads into something internally)? Obviously that's less catastrophic for a floppy drive which moves slowly compared to a hard drive, but still not good.

Your drive and bios may not support it. Unfortunately, a relatively uncommon format from a decade earlier will fall prey to the relentless effort to save fractions of a penny.

There were some 5.25" drives that did support both the 1.2 MB and 720kB formats with the ability to easily shift tpi and rpm as needed for disks. Specialized drivers would probably also be needed. I hope someone else can recommend methods that work.

I hesitate to point to the following web page since if the author is a poster here, the author is much better suited to answer questions:

http://www.oldskool.org/disk2fdi/525HDMOD.htm

But if you are lucky, you might be able to modify your drive to read the quad-density disks.

About modifying the drive(s), none of mine are Teac (Toshiba, Fujitsu, Panasonic, Canon/IBM, Chinon) so I'm a bit hesitant to start exploring that quite yet.

As one of the replies after yours states it looks like the format command /f:720 refers exclusively to 3.5" 720k disks, which are far more common it would seem (I actually have some of them, bought them new when I was younger...3 out of the set of 10 still work).

I wonder...I have a few Super-I/O cards, I wonder if the floppy controllers on any of them support 720k 5.25" formatting. Something to investigate in the future...those cards are never as simple to set up as they appear.

The motherboard I am using for this offers the following options for floppy drives in the BIOS:
-360K , 5.25 in.
-1.2M , 5.25 in.
-720K , 3.5 in.
-1.44M, 3.5 in.
(surprisingly, also this one:)
-2.88M, 3.5 in.

While 77 cylinders is a standard 8" format, there exists an older 5.25" format, recorded at 100 tpi with 77 cylinders. Sometimes, it's called a "Micropolis" format, after the manufacturer that pioneered it, other manufacturers made the drives too. The Tandon TM-100-4M ("M" for Micropolis; a TM-100-4 is a 96 tpi drive) is one such drive. Commodore used the drives briefly.

Mike, forgive me for going into too much detail. I'd assumed that the reason for these discussions was to educate--why would one collect old gear if not to understand it? Maybe I don't understand this collecting business...

Are there any specific advantages to the 5.25" 77 cylinder 100tpi format? I imagine it's useless without hardware support but interesting to know.

Collecting - if I may cut in between you two - can be for many reasons. I don't collect, I use. I have no specific attachment to older computers (and why would I, I was born in 1989) but I find them interesting...I like the challenge they can provide, the way they make me use my brain. I also enjoy dealing with intelligent, friendly, **interested** people when I have an issue or comment of some kind. I went to a computer store that was rumoured to have some older hardware and after the +1 hour trip was told they got rid of it all (but that I could come back in a few days and get some anyway?) rather inconsiderately and the person behind the counter clearly could not care less about the equipment he works on or the people he deals with. How could someone WORK WITH COMPUTERS and clearly hate them so much?

Stay away from the Second Infinite Byte in Windsor, Ontario, Canada - their staff are disrespectful, have a marginal amount of knowledge, and will trash talk you after you leave (as I witnessed when a customer left with their machine).

Chuck - I don't think I got irked by the info. I think I got irked that the implication was that I left out something important to the immediate discussion.

And the broader question is, where do we draw the line on what's an appropriate level of detail? This thread is by far and way more interesting that many of our 'my PC is broken' threads, but at what point does the extra become noise? ClassicCmp is a perfect example of people making this unreadable for the sake of completeness ..

Interesting is generally what I aim for, so it would appear I'm succeeding in that regard.

What's going on here is that 720k 5.25" disks are not a standard format on PCs. The /f:720 option is intended for 3.5" disks, so FORMAT will refuse to make a 5.25" disk at that capacity. There are third-party programs that will format 5.25" 720k disks.

HD drives themselves are perfectly capable of using quad-density floppies, which are just 80-track disks at the DD bit rate. Tandy 2000s aside, QD was used by some CP/M machines as well as Commodore 8050 and 8250 drives.

I figured it might be something about that (720k 3.5" disks), but was hoping maybe the format command could use the same switch for either type. Apparently not.

I've probably missed something in this thread, but are "QD" disks actually different physically, or is it simply the recording method that differs?
 
I'm not sure I want to try an unconditional format...if it is refusing to do it, wouldn't that possibly result in drive damage if the drive was instructed to, for example, seek to a location it doesn't know how to seek to (for example smash the r/w heads into something internally)? Obviously that's less catastrophic for a floppy drive which moves slowly compared to a hard drive, but still not good.

You are misunderstanding something there:

An unconditional format aren't dangerous to the drive at all. The only difference betwen unconditional format and normal format is that normal format tries to verify/write to the existing disk layout while an unconditional format will rewrite the disk layer before starting a normal format. Because of this, "un-format" information can't be saved to the disk, and in cause you had some imporiant files on the disk when it was "accidentally" formated, you won't be able to restore them.

Of course you are not trying to "accidentally" format any of the disks, so an unconditional format is totally safe in your cause.

(the R/W head of the floppy drive slides on a rail. When it reaches the end of this rail, the head can't possibly move any futher. And, when it comes to floppy disks, the heads are touching the surface of the disks all the time, soyou don't have to worry about disk carsh.)
 
Last edited:
You are misunderstanding something there:

An unconditional format aren't dangerous to the drive at all. The only difference betwen unconditional format and normal format is that normal format tries to verify/write to the existing disk layout while an unconditional format will rewrite the disk layer before starting a normal format. Because of this, "un-format" information can't be saved to the disk, and in cause you had some imporiant files on the disk when it was "accidentally" formated, you won't be able to restore them.

Of course you are not trying to "accidentally" format any of the disks, so an unconditional format is totally safe in your cause.

(the R/W head of the floppy drive slides on a rail. When it reaches the end of this rail, the head can't possibly move any futher. And, when it comes to floppy disks, the heads are touching the surface of the disks all the time, soyou don't have to worry about disk carsh.)

Oh, that's good to know then...thanks. I do know the head is on a rail, but I wasn't sure if it's possible to knock things out of alignment.

However, in this case, isn't it true that even an unconditional format won't format the disks to 720k, since the format command doesn't actually support 720k 5.25" disks?
 
As one of the replies after yours states it looks like the format command /f:720 refers exclusively to 3.5" 720k disks, which are far more common it would seem (I actually have some of them, bought them new when I was younger...3 out of the set of 10 still work).

3.5" DD disks were much more widely used than you might think. Aside from PCs, they would include the MSX, Mac, Apple II, Lisa, Amiga, Atari ST, Commodore 1581 drives, HP 150, and also things like music synthesizers, embroidery machines, and word processors.

I wonder...I have a few Super-I/O cards, I wonder if the floppy controllers on any of them support 720k 5.25" formatting. Something to investigate in the future...those cards are never as simple to set up as they appear.

Since QD disks use the DD bitrate, any PC controller can use them. You just need an 80-track drive.

I figured it might be something about that (720k 3.5" disks), but was hoping maybe the format command could use the same switch for either type. Apparently not.

FORMAT can only use the disk types supported by DOS, which do not include 5.25" 720k.

I've probably missed something in this thread, but are "QD" disks actually different physically, or is it simply the recording method that differs?

It's just the recording method; there's no physical difference with DD disks. Usually, QD disks were tested more rigorously because it was thought that 80-track formatting was more demanding on the disk. I actually have two boxes of QD disks that I use on my 5150 as 360k disks.

Most DD disks will format to 80 tracks with no problems, except perhaps for some no-name brands, which as Chuck mentioned, tend to have lower-quality media.
 
Oh, that's good to know then...thanks. I do know the head is on a rail, but I wasn't sure if it's possible to knock things out of alignment.

However, in this case, isn't it true that even an unconditional format won't format the disks to 720k, since the format command doesn't actually support 720k 5.25" disks?

There is almost nothing that can get the head out of alginment unless you drop the drive or physically alter the head manually.

Well, as I said, an Unconditional format is simply a "low-level" format of the floppy disk, and absolutely all data (including the controll data) is being rewritten. Since your disks has been formated a way they weren't supposed to, the magnetic data is all mixed up and an Unconditional format is actually the only way to fix them.

About 720KB 5.25" disks and the format utility, the format program isn't what's limiting your posibilities. when you format a disk while using the "/F:x" parameter, the Format utility checks with the BIOS to see if the drive actually IS what you specify. If the BIOS don't agree, you'll get an error. (I think. It may also check the drive itself without going through the BIOS.)

However, there is one way to override this (IIRC), and that is to manually specify the number of heads/cylinders(tracks-per-head)/sectors into the format utility. This should override the BIOS, and hopefully format it to what you want.

Just be aware that you can get into some problems by formating a 5.25" disk to 720KB. This wasn't/isn't usual for PC's, and many programs will either think that the disk is a 720KB 3.5" floppy or a 360KB 5.25" floppy.
 
I originally got a lot of information from Scott Mueller's Repair book. IIRC, it explains most of what I've seen here. Regarding physical writing of the disks, this Floppy Primer page explains, among other things, azimuth and radial alignment. I thought this might be of interest here.

I've also had a lot of educational fun playing with different formats. Programs such as the MS format command obviously have their own agenda, but there are other programs out there which allow you the freedom to make your own choices. My favourite is FreeForm, a menu-driven floppy disk formatter. It's a 91K download from Simtel.Net. Look for ffrm231a.zip. Almost any combination of formatting parameters can be chosen using that program.

Because of the menus, FreeForm is very educational (to me) but for a modern formatter I use, and recommend, the one from FreeDos. Look for FMT091v.ZIP here. It will let you specify tracks and sectors and, AFAIK it is written solely for the benefit of the users. It also has no licensing encumbrances.

Didn't MicroSoft use 1.7Mb distribution diskettes at one point? I seem to recall that one purpose was that you couldn't copy them using MS's own software. Of course those were 3.5" diskettes, but I'm sure the same trick has been used for 5.25s. You can certainly format 1.2s to 1.4, although I notice that the reliability goes down.
 
Last edited:
Are there any specific advantages to the 5.25" 77 cylinder 100tpi format? I imagine it's useless without hardware support but interesting to know.

None that I'm aware of; they were an early 5.25" format and there were comparatively few systems that used them. At Durango, we recorded GCR using a WD1781 controller and a whole bunch of glue ICs and got about 960K per 100 tpi 5.25" floppy. Had the 96 tpi drives been available, we would have happily used those instead--the Micropolis drives were glacially slow when it came to seeking.

There are some small differences in the connector assignments, but generally the drives are electrically compatible with the PC floppy interface.

Sort of an interesting evolutionary dead end, like 3.25" diskettes and Drivetec floppies.

Didn't MicroSoft use 1.7Mb distribution diskettes at one point? I seem to recall that the one purpose was that you couldn't copy them using MS's own software. Of course those were 3.5" diskettes, but I'm sure the same trick has been used for 5.25s. You can certainly format 1.2s to 1.4, although I notice that the reliability goes down.

Both IBM and Microsoft used "extreme" formats. Google "DMF" and "XDF" formats--there's plenty of information out there on them. Both IBM and MS offered to send you conventionally-formatted floppies if you yelled loud enough.


(I don't really collect either, but I have a ton of old stuff hanging around from having been in this racket for too long. Even if I could find the systems that I used when I was young, I couldn't afford to run them...)
 
There's pretty much too much stuff there for me to specifically address (I'm sure you all understand).

Basically, I don't have any more questions...you've been that thorough :)

About 1.7 DMF floppies - yes, MS used them for Win95 and possibly other releases on floppies. For the person today who may be trying to write those disks, I advise you to pick disks you won't care to lose...they will NOT be reliable when formatted back to 1.44 afterwards!
 
About 1.7 DMF floppies - yes, MS used them for Win95 and possibly other releases on floppies. For the person today who may be trying to write those disks, I advise you to pick disks you won't care to lose...they will NOT be reliable when formatted back to 1.44 afterwards!
I doubt that the media is actually damaged and I'm fairly certain that the disc would be perfectly good if you wiped it. Of course if you don't have anything to do that with then there might be a problem, I don't know about that.
 
I doubt that the media is actually damaged and I'm fairly certain that the disc would be perfectly good if you wiped it. Of course if you don't have anything to do that with then there might be a problem, I don't know about that.

I'm referring to starting with a 1.44mb floppy disk, formatting it to 1.7 DMF for holding files during an install, then formatting back to 1.44 afterwards. Every disk I used in that set (I installed Win95 via floppies once just for the experience...what a waste) was trash afterwards. Basically, you'd format the disk back, put a file on it, then pop it in a few minutes later and the file is gone (or heavily corrupted).
 
1.7 to 1.44 OK

1.7 to 1.44 OK

I just took an old Mackintosh 3.5" disk which, of course, wouldn't read on a DOS system. Then I formatted it to 1.7M and wrote a file to it and ran chkdsk. All was fine. Then I formatted it to 1.4M and wrote a file to it and ran chdsk. All is still fine! It is indeed possible to format to 1.7 and then go back to 1.44. This was all done in the drive and I did not need to remove the disk to wipe it.

The program I used was FreeForm v2.31. I tried reformatting with FreeDos Format v1.0 and it appeared to work but upon writing a file or running chkdsk it complained. So, it looks like it all depends upon the formatting program. :)

I have heard people complaining about inadequacies of the MS format program, but I am surprised that the FreeDos program has the same bug. However, I am not completely sure that the FreeDos program would not be able do it if given the correct parameters. It has several ways of using it. FreeForm is just a no-brainer that's why I used that.
 
Back
Top