PDA

View Full Version : DD/HD floppies in DD/HD drives, and the mystery of the density hole.



khaz
September 20th, 2017, 08:42 AM
Hi there it's my first post, and I'd like to make it about the already-discussed-to-death differences between HD and DD floppies and how to use one instead of the other. Sorry.

I've been reading about the DD and HD floppies and their different coating, and also about DD and HD drives and how they write to discs. As I understand it, the overall rule is "don't mix and match": use DD media in a DD drive, and HD media in an HD drive. Else data can be written improperly and the disc can become unreliable over time. More specifically, a DD drive is not equipped to write and read an HD floppy and its specific coating. A DD floppy punched to be treated as HD doesn't have the coating to keep all that HD data properly either.

What's confusing me is that HD hole on HD floppies. It means that HD drives can make the distinction between a DD floppy and an HD floppy. HD drives are really DD/HD drives and can swap between configurations to read and write properly both formats. Is that assumption correct? A DD floppy used in such a drive would be in no more danger than an HD one, on the condition that it's formatted to its correct er... format. (MSDOS is awful at that, but let's not digress too much yet.)

The topic that interests me is when such a DD/HD drive is used in an old computer that was designed when HD didn't exist. Like 8bit computers, Amiga, etc. Even though the computer cannot read and write more than, let's say, 720k of data per floppy, the DD/HD drive we put on it is still fully functional. So it can read and write DD floppies without damaging them. Am I still correct? Now, I read plenty of time on various forums that covering the hole of an HD floppy to fool the drive was a good replacement for real DD floppies. The more I think about it the less believe that it's right. I believe that in order to use an HD floppy on an HD-equipped old computer, it's best not to cover that hole: covering it will have the tracks written with DD intensity, which isn't good on an HD coating. Keeping the hole free will have the drive use the correct intensity to write the tracks, regardless of the actual formatting the computer is asking. So you can have an HD floppy, in an HD drive, in a format understood by the computer, that won't go bad.

Another way to say it would be: as long as the correct head is used on the correct coating, the data will be written and read fine no matter the formatting used.
On an HD drive, the correct head is decided by the presence or absence of the density hole.

A broad statement like "Don't format an HD floppy as DD!" would be, while not incorrect, very broad and not taking into account the drive and the computer, and the density hole status.

OR I'm completely clueless and have no idea what I'm talking about. In that case, please laugh at me while pointing me to the right direction.

But I think it all comes down to wtf that hole does to the drive.

mbbrutman
September 20th, 2017, 08:53 AM
I'm not sure if you ran into this but I have a fairly decent write-up of the diskette types here:

http://www.brutman.com/PCjr/diskette_handling.html


You are essentially correct - use the correct media for the write current that will be used. The hole in the diskette was the primary way to specify the write current, but some machines did this wrong and some machines use the data rate as an implicit signal as to which write current to use.

Trixter
September 20th, 2017, 10:55 AM
I've always been curious why most 5.25" DSDD disks have a hub ring, and most 5.25" DSHD disks don't. Better clamping mechanisms in the DSHD drive, maybe?

I recall computing with an Apple II+ in the early 1980s with cheap disks that didn't have hub rings, and the hub would indeed start to get wrinkled over time. But no matter how bad it looked, I was usually able to use the disks. Completely different floppy encoding, drive mechanism, etc., I know, but I mention it in case the other reason DSHD disks don't have hub rings is because they were ultimately unnecessary?

krebizfan
September 20th, 2017, 11:28 AM
IIRC, early dual sided drives caused the disks to wrinkle. Drive manufacturers introduced better clamping mechanism and improved the read-write head design; disk manufacturers introduced hub rings; and Apple tried to design a drive using two single sided read-write heads with the Twiggy. The problem should not show with single sided disks so the Apple II 5.25" drives were safe.

Maxell 96 TPI disks did not have hub rings. Their explanation was that the hub ring itself would cause the disks to be slightly out of alignment. This means there was an actual difference between quad-density labeled disks (MD2-DD) and double density disks (MD2-D) despite the disk material being the same.

Extra holes in the shell will show the High Density and Extended Density disks, no hole was added when the 3.5" design went from its original Sony (500 Oersted) material to using material closer* to that used in 5.25" High Density disks. Just something to watch for if meeting a manual shutter equipped 3.5" disk.

* The standard development process suggests that an attempt was made to be the same but all the documentation indicates that it is a little bit different.

One more edit: The Dysan patents for the reinforcing rings are https://encrypted.google.com/patents/US4052750 and https://encrypted.google.com/patents/US4370689

Chuck(G)
September 20th, 2017, 12:04 PM
I've always been curious why most 5.25" DSDD disks have a hub ring, and most 5.25" DSHD disks don't. Better clamping mechanisms in the DSHD drive, maybe?

I recall computing with an Apple II+ in the early 1980s with cheap disks that didn't have hub rings, and the hub would indeed start to get wrinkled over time. But no matter how bad it looked, I was usually able to use the disks. Completely different floppy encoding, drive mechanism, etc., I know, but I mention it in case the other reason DSHD disks don't have hub rings is because they were ultimately unnecessary?

I thought that I discussed that on another thread. I was privy to the whole hub-ring thing back in 1977-78. Micropolis had just trotted out their single-sided 100 tpi drives.

One major difference (at least in 1977) between 8" drives and 5.25" drives was the introduction of the spindle motor-control signal. Up to that time, 8" drives generally used AC spindle motors, which turned the floppy continuously. So the natural thing was to put the motor control under program control and activate it only when reading or writing disks. I once asked about the reason for the motor control and never got an answer that I was happy with. Reducing disk wear didn't wash because the Micropolis drives that we used had a head-load mechanism (you can even find semi-modern Teac FD55s so equipped). Another explanation was that brushed DC motors were used for the spindle and those had a limited lifetime.

At any rate, we were using Dysan as our official supplier for floppies. Good stuff, generally. But every once in awhile, we'd turn up a floppy with a wrinkled hub annulus. At 100 tpi, this was disaster, as the floppy could now be clamped off-center. The engineers at Dysan figured out that the reason was that the floppies were being clamped in the drive with the motor off. This didn't allow them to self-center and, as a result, mangled the annulus.

A short-term fix from Dysan was a kit and a jig to apply adhesive reinforcing rings to existing floppies. That helped prevent the mangling, but still didn't provide a surefire way to register the disk with the spindle motor off. Micropolis recalled all of our drives and equipped them with a modification to detect when the drive door is being closed and activate the spindle for a short time to allow the floppy to register. After that, the problem went away. Because one could never be certain about older drives in the field, the hub rings persisted.

Yes, I do have plenty of Dysan labeled quad-denisty disks equipped with hub rings. The reason that HD floppies don't have them is because by the time HD drives came out, everyone knew the trick about spinning before clamping. The reason that RX50-labeled floppies don't have hub rings is because the RX50 already was suitably equipped.

3.5" drives also do this, but mostly so that the hub ring is correctly indexed.
-------------------------------------
As regards 3.5 HD vs. DD media, I've processed plenty of 3.5" HD floppies that were written as DD more than 20 years ago. The media are close enough that it shouldn't be much of an issue. As a matter of fact, early IBM 1.44M drives didn't use "media sense", but rather used "host determined" configuration (via pin 2) like 5.25" HD drives. That's something I keep an eye out for when handling old disk conversions--something that appears to be HD or DD in 3.5" ain't necessarily so.

SomeGuy
September 20th, 2017, 12:05 PM
am I still correct? Now, I read plenty of time on various forums that covering the hole of an HD floppy to fool the drive was a good replacement for real DD floppies. The more I think about it the less believe that it's right.

Practically speaking, however, covering a density notch on a 1.44mb disk to create a quick boot disk for a 720k drive machine or shuttling temporary data back and forth will usually *work*.

Similarly, back in the day people regularly used to buy bulk "white box" 720k 3.5" disks and punch a hole in them and use them as 1.44mb disks (or slightly more with special formatters :) )

Neither is pedantically correct, a great idea, or perfectly reliable. So don't cry about it when you loose some data.

Stone
September 20th, 2017, 12:21 PM
Neither is pedantically correct, a great idea, or perfectly reliable. So don't cry about it when you loose some data.I just pulled out and read/verified about ten 720k disks yesterday that I had manually punched and put data on 27 or 28 years ago. All of them tested perfectly. And, yes, I was slightly, but not terribly surprised.

khaz
September 20th, 2017, 02:17 PM
Practically speaking, however, covering a density notch on a 1.44mb disk to create a quick boot disk for a 720k drive machine or shuttling temporary data back and forth will usually *work*.

Similarly, back in the day people regularly used to buy bulk "white box" 720k 3.5" disks and punch a hole in them and use them as 1.44mb disks (or slightly more with special formatters :) )

Neither is pedantically correct, a great idea, or perfectly reliable. So don't cry about it when you loose some data.

Yeah but it would have worked even without covering the hole, and even better if I'm correct.

I feel like covering the density hole is like putting a spoon as a cork on a champagne bottle, or blowing in your cartridge to make them work better. Something that makes common folk sense but have no basis in reality.

Punching a hole makes sense, and the resulting unreliability makes also sense. Though if for archival the unreliability may be overstated, it seems that problems occur when you keep rewriting the disks.


I've been using this page to learn more about the compatibility between densities http://www.retrotechnology.com/herbs_stuff/guzis.html
There's a lot of information on what happens when reading/writing a DD disk as if it were HD, and an HD disk as if it were DD. But nothing about the density selection via the hole, unless I'm missing it.


I'm not sure if you ran into this but I have a fairly decent write-up of the diskette types here:

http://www.brutman.com/PCjr/diskette_handling.html


You are essentially correct - use the correct media for the write current that will be used. The hole in the diskette was the primary way to specify the write current, but some machines did this wrong and some machines use the data rate as an implicit signal as to which write current to use.

Thanks for the link! 5.25 are a mess. I'm glad I never had to deal with them.

So, covering the hole would get the wrong write current for the type of disk we're using, and some machines don't care for the hole and just use the wrong write current any way.

Unless there are machines that refuse to format to 720k a floppy with a hole, then there is no reason to cover it, ever, then.
Leave the hole intact, at worst the machine will see it as DD any way, and at best you'll get better reliability over time.

On Windows XP and beyond, you can only format a floppy to 1.44M regardless of the density hole status.
On MSDOS6.22, the default command FORMAT A: formats to 1.44M too. No warning message, you need to specify /F:720 to get the correct formatting.
Installing MSDOS6.22 on a floppy asks if you want to format the floppy to 720k or 1.44M, but again without checking the actual density setting.

krebizfan
September 20th, 2017, 05:18 PM
The most common complaints I heard about HD disks failing with DD formats were from Mac users, especially with 400K drives. Whether those Mac users had worse luck or earlier Sony drives did not like the difference between disks I don't know.

Trixter
September 21st, 2017, 10:42 AM
Chuck, thanks very much for the history, which completely answered my question. (I may write this up outside of this forum, as I feel it's worth repeating.)

konc
September 21st, 2017, 11:29 PM
...Micropolis recalled all of our drives and equipped them with a modification to detect when the drive door is being closed and activate the spindle for a short time to allow the floppy to register. After that, the problem went away. Because one could never be certain about older drives in the field, the hub rings persisted.

Amazing, this solved an old mystery for me

SomeGuy
September 30th, 2017, 07:57 AM
A short-term fix from Dysan was a kit and a jig to apply adhesive reinforcing rings to existing floppies.

Is this such a kit?
http://www.ebay.com/itm/Vintage-Data-Control-Flexible-Disk-/192320542770

It is kind of hard to determine the scale, and the seller doesn't seem to know what it is, but those look like hub rings in the bag.

Could be totally wrong, could easily be a part for any number of things, but I thought it was interesting and coincidental something like that might pop up.

Chuck(G)
September 30th, 2017, 09:07 AM
That looks like it. You can tell the age just from the typeface on the box lid.

khaz
October 17th, 2017, 04:14 AM
How does it work in a modern system?

I've been using Windows XP to make DD discs for other computers, but the software give me a lot of errors until I cover the hole and fake a 720k disc.

I've used samdisk and cpcdiskxp to transfer Amstrad CPC disc images to 3.5 floppies, and they both go crazy if I don't cover the hole.

I tried formatting discs under MSDOS 6.22, but it formats all my discs (be they HD or DD) as HD unless I give it the /720 command. There is no automatic detection of the type of disc inserted. At least for formatting I don't get any error. Using floppies formatted that way with the aforementioned software doesn't work better.

Is the advice for not covering the hole only for computers with a DD controller retrofitted with an HD drive?

(would it help if I used a different OS?)