If your PC has a USB port, then just use USB drives. It should work -- for example on my "daily driver" PC, I have a four-in-one card gadget. I can't think of why the PC would care whether the drives are floppies or otherwise. --- Vintage Computer Festival <vcf(a)siconic.com> wrote: ------------------------------------------------------------------------------ Computers http://marketplace.vintage.org ===== Tell your friends about the Computer Collector Newsletter! -- It's free and we'll never send spam or share your email address -- Publishing every Monday(-ish), ask about writing for us -- Mainframes to videogames, hardware and software, we cover it all -- W: http://news.computercollector.com E: news(a)computercollector.com -- 653 readers and counting!

Ahhh... Pardon? DOS doesn't support USB ports. *********** REPLY SEPARATOR *********** On 16-Jan-05 at 08:33 Computer Collector Newsletter wrote: ------------------------------------------------------------------------------ -=-=-=-=-=-=-=-=-=-=-=- Bruce Lane, Owner & Head Hardware Heavy, Blue Feather Technologies -- http://www.bluefeathertech.com kyrrin (at) bluefeathertech do/t c=o=m "If Salvador Dali had owned a computer, would it have been equipped with surreal ports?"

From: "Vintage Computer Festival" <vcf(a)siconic.com> > > Is it possible to get 4 floppy drives onto a PC?

The original IBM-PC's floppy controller made provisions for four floppy drives. There are control lines to select up to four drives on the single interface. They put a 37-pin D-shape connector on the bracket of early IBM floppy controller cards so you could plug in an expansion case with two external floppy drives. I only did this once, as an experiment, to see if I could run my old WWIV BBS on floppy-drives only, to reduce the noise of the machine. (It worked, but was much slower.) I'm pretty sure the 'fix' isn't to plug in an old IBM controller, though, unless you're already talking about a vintage restoration project. But the third and fourth select lines should be available on some controllers.

Unless you need to be able to have 2 disks of the same size active at the same time, you don't need four drives.

The whole impetus behind this was that I was trying to read an old 360K (might even be 320K or even 180K, I'm not sure) disk and while I could read the directory, none of the files larger than say a cluster could be read. I attributed this to the drive not being a "true" 360K drive.

Fred seems to be the Guru with regards to floppy formats, and perhaps I misinterpreted what I read, but I recall him bringing up a scenario in which a disk formatted as 360K on a newer drive would not be readable on an "actual" (i.e older) 360K drive, or the other way around, or something.

The whole impetus behind this was that I was trying to read an old 360K (might even be 320K or even 180K, I'm not sure) disk and while I could read the directory, none of the files larger than say a cluster could be read. I attributed this to the drive not being a "true" 360K drive.

There is no problem reading a true 40-cylinder disk on an 80 cylinder drive (provided the software knows to double-step it). The problem comes if you write to a 40 cylinder disk on an 80 cylinder drive and then try to read it on a true 40 cylinder drive. I remember writing a FAQ about this a couple of years back.... -tony

I'll stick my 5.25" MFM (PC, essentially) disk FAQ at the end of this message. You (or anyone else) is welcome ot make use of it in a more comprehensive FAQ provided you give me an acknowledgement. -tony ----------------- 5.25" floppy disk Frequently Asked Questions -------------------------------------------- 2000 A. R. Duell with comments and additions from Fred Cisin. Please feel free to distribute this document (on web sites, ftp sites, mailing lists, etc) provided this notice is intact. This doecument was written as a response to questions on the classic computer mailing list regarding the use of various types of 5.25" floppy disks in various types of drives. It attempts to explain what combinations work and why. For the moment I am only considering soft-sectored drives... 1) What types of (soft sectored) 5.25" drives are there. There are 5 common types : 48 tpi single sided, double density. These have 40 cylinders (and 40 tracks). On a PC they'd store 180K bytes, however on other systems, their capactiy could be anything from 160K to 200K. A few very early drives (Shugart SA400, for example), and systems designed to use them, only used 35 of the 40 tracks, but kept the same 48tpi track pitch 48tpi double sided, double density. These have 40 cylinders (80 tracks total, one for each cylinder on each side of the disk). This is the common PC 360K drive. When used on other types of machine, the capacity is in general twice that of the single-sided drive, or somewhere between 320K and 400K. 96 tpi, single sided, double density. These have 80 cylinders (and being single-sided, 80 tracks). These are not common on PCs, and in fact most versions of PC MS-DOS don't support them, but if used there would store 360K. The DEC RX50 is a (double) drive of this type. The capacity of these units is the same as that of the 48 tpi double sided drives (not too suprising, as they have the same number of tracks). 96 tpi, double sided, double density. Again, they have 80 cylinders (and thus 160 tracks). On a PC, they'd store 720K, although they're not commonly found on PCs, although the IBM PX=JX used them. Theyre are common on non-PC compatible MS-DOS machines, CP/M machines, Acorn BBC micros, and early-ish unix workstations, where they store between 640K and 800K depending on the formatting. 96 tpi double density drives are sometimes called 'quad density' units. 96 tpi, double sided, high density. This is the PC 1.2Mbyte drive, and is not commonly used elsewhere. These drives have several differences compared to the double density version, although the PC controller hides some of these. This drive is, in fact, somewhat similar to the 8" drive -- it uses the same data rate and even the same spindle speed (360 rpm, as against 300rpm for all the other types of drive). It thus can often be used as a substitute for an 8" drive in older machines, if connected up with a special cable. 2) What are the real differenced between the various types? The difference between single and double sided drives is obvious -- the double sided drive has an extra head (mounted on top of the disk) and a switching circuit to select between the 2 heads. In passing at this point I'll mention that single-sided drives record on the bottom (non-label side] of the disk and that this is 'side 0' on double sided units The differences between 48 tpi and 96 tpi double density drives are again fairly obvious. The head positioner (mechanism that moves the read/write head) is designed to move the head only 1/96" per step rather than 1/48". The actual head in a 96tpi is narrower (radially) than the one in a 48tpi drive so that it writes a narrower track on the disk (so that adjacent tracks don't overlap at the closer spacing). The high density drive has several differences wrt the 96 tpi double density unit. Firstly, the spindle motor rotates at 360rpm (at least in high density mode, see below) rather than the 300rpm that all other drives rotate at. Also the 'write current' (the electric current passed through the head coil to write on the disk) is higher in high density mode so as to be able to write on the special high density disks. These have a higher coercivity than normal double density disks. There is a signal on the interface connector of high density drives (at least the properly-designed ones) that, when asserted, reduces the write current to the value used with normal double density disks. In this mode, the drive will reliably work as a 96 tpi double density unit. In some drives, asserting this signal will slow the spindle motor down to 300rpm, in others, it continues to turn at 360rpm and the controller has to handle a data rate of 6/5 times times the standard 250kbps double density rate (=300kbps). The IBM PC/AT disk controller is capable of doing this. 3) What types of disks exist? All 5.25" disks that I have ever seen are coated with the magnetic oxide on both sides. A double sided disk means that both sides have been tested and shown to be reliable for storing data. A single sided disk may be one where the top side has failed this test (and the bottom side is good) or one which simply hasn't been tested on both sides. It is often claimed that single sided disks are those that have certainly failed the test on the top side, but in fact, the reliability of the disk manufacturing process was such that very few disks failed, and that therefore there simply wouldn't be enough made to sell as single-sided ones. Most single-sided disks are therefore good on both sides. I (ARD) have used a number of DEC RX50 disks that have been bulk-erased as double-sided 96 tpi disks. Since the RX50 is a single-sided 96tpi drive, there would be no reason for these disks to work reliably as double sided ones, but I have yet to have one fail. '80 track' -- 96 tpi -- double density disks do seem to be different from '40 track' -- 48 tpi ones. I suspect, without proof, that the 96 tpi ones are lower 'noise' which is important for the narrower tracks used on such drives. The original 5.25" disks were designed to be used in 48tpi drives, since that's all that there was at the time. Once 96 tpi drives became popular, many manufacturers starting making all their disks suitable for use in such drives (it was cheaper for them to have one production line) and sold them as 'universal' disks, suitable for use in 48 or 96 tpi drives, single or double sided. However, once the IBM PC and PC/AT became the only common computers to have 5.25" drives, many manufacturers went back to making 48 tpi disks only, since that was the only double density drive in common use. Therefore many modern double density (known as '360K disks') are _not_ reliable in 96tpi drives. High density disks are different. Period. The magnetic media has a different coercivity (600 oerstedt for high density disks, 300 oerstedt for all other 5.25" disks), and it can only be used in the high density drive _at the high density_. 4) What sorts of blank disks can be used in what drives? Double sided disks can always be used in single sided drives. The fact that the unused side is also perfectly good for storing data doesn't matter, of course. Some people modified the disk jacket of double sided disks by cutting an extra write protect notch and index hole (for those systems that used it, which is basically everything apart from Apple and Commodore) in it. Such disks could be used either way up in the drives so that both sides of the magnetic disk could be used, and were often known as 'flippy' disks. Some manufacturers even sold disks with the extra notch and hole pre-cut at the factory, and I've even seen the service manual for a Siemens drive with 2 write protect sensors and 2 index sensors so that a normal, unmodified disk could be used either way up. Note that doing this means that the data on one side of the disk is recorded 'backwards' (the disk is effectively spinning in the opposite direction), and thus the result cannot be read in a 2 head drive simply by selecting the other head. Even with a 2 head drive you still have to flip the disk over manually. 96 tpi double density disks can be used in 48 tpi double density drives. Again, the disk is 'better' than it needs to be, but that doesn't matter. This means that these disks can be used as follows : disk works in 96 tpi DS : 96 tpi DS, 96 tpi SS, 48 tpi DS, 48 tpi SS 96 tpi SS : 96 tpi SS, 48 tpi SS 48 tpi DS : 48 tpi DS, 48 tpi SS 48 tpi SS : 48 tpi SS For that reason, many manufacturers sold 96 tpi double sided disks as 'Universal' disks. They could be used in all types of (double density) drives. High density disks are special. They can _only_ be used in high density drives at the high density format. Similarly, high density drives will only reliably work in high density mode on such disks. But if the appropriate signal is asserted, then the high denisty drive behaves like a 96tpi double sided double density unit, and can use double density disks. 5) What combinations may work under some circumstances? Single sided disks may work in double sided drives. Firstly, some systems (many systems?) allow you to format them as single-sided, for which they are (obviously) suitable. And in many cases the 'other' side of the disk is perfectly good and the disk can be formatted as double sided. 48 tpi disks may be good enough to work in 96tpi drives. My experience suggests this is not reliable, though. 6) What about data interchange? What (already recorded) disks can be read in what sorts of drives? Let's deal with the obvious cases first. A double sided drive can read a single sided disk. The upper head is simply not used. Similarly, a double sided drive can write to an already-used single sided disk. Another obvious case is that the high density disks can only be used in high density drives. The less obvious case is the 48 tpi versus 96tpi issue. The drives were designed so that the centre line of alternate 96 tpi cylinders is the same distance from the spindle as the centre line of each 48 tpi cylinder. Thus 48 tpi disks can be read in 96 tpi drives if the drive 'double steps'. Some drives can do this in hardware (there may be a switch marked 40/80 on the drive casing), some operating systems can handle this. Since a high density drive can be 'turned into' a double density drive by asserting that signal I keep mentioning, a high denisty drive can also reliably read 48 tpi disks. 96 tpi drives writing to 48 tpi disks is a cause of many problems, which deserves its own section. Erasing a file involves writing to the disk directory, of course, and thus counts as writing to the disk. 7) What's all this about writing to 48 tpi disks in 96 tpi drives? This is perhaps the biggest cause of problems with 5.25" disks. People write a file to a 48tpi disk using a 96 tpi drive and find that the result is readable on 96 tpi drives but not on 48 tpi drives. Remember that the 96 tpi drive has a narrower head than the 48 tpi drive, so it writes a narrower track to the disk. A blank disk is one that has just been manufactured, or that has been bulk-erased. Reformatting a disk in a normal disk drive _does not_ produce a blank disk in this sense -- it writes the track pattern to the disk, with whatever width of head the drive doing the formatting uses. Suppose we take a totally blank disk and format it on a 48 tpi drive. This writes 40 tracks on each side of the disk. They may be 'empty' in the sense that they contain no user data, but they're still recorded. Then we write to it on a double-stepping 96 tpi drive. The narrower head overwrittes the middle band of some tracks, but the edges remain unchanged. A 96 tpi drive can read that perfectly well. Its narrow head only 'sees' the 'new' data down the middle of each track. But a 48 tpi drive with its wider head, sees both the old and new data. The result is a mess that the controller can't decode. So the disk is not readable on a 48 tpi drive. A similar argument shows that the same problem occurs if you take a blank disk, format it on a double-stepping 96 tpi drive, write some files to it there, write to it with a 48 tpi drive and then write to it with the double-stepping 96 tpi drive again. The result is not readable on a 48 tpi drive. And if you take a disk that has been formatted and used on a 48tpi drive and then reformat it on a double-stepping 96 tpi drive, the result may well not be readable on a 48tpi drive. Each track will have the newly-formatted narrow track down the centre and the remains of the old wider track along the outsides. If you must format a disk in a double-stepping 96 tpi drive, then either use a brand new disk (and not a preformatted one, of course), or bulk-erase the disk. In general, if you take a blank disk, format it on a double stepping 96 tpi drive and write to it there only, the result is readable both 48 tpi and 96 tpi drives. The narrower tracks generally do provide enough signal for the wider head on the 48 tpi drive provided there is nothing in the 'blank' spaces between the tracks. The simple rule is : * If you ever write to a disk in a 96 tpi drive that has already be * * written to (including formatting) in a 48 tpi drive then the result * * may well not be readable in 48 tpi drives. *

On Mon, 17 Jan 2005, Randy McLaughlin wrote: Micropolis did soft sectored 100tpi. (and also made 48tpi) Very nice drive, with a somewhat slow helical lead screw positioner. The Tandon TM-100-4M was also 100tpi ("M" stood for "micropolis"?) When Tony originally wrote that FAQ, we agreed that the 100tpi was obscure enough that it didn't need to be mentioned in the FAQ (along with things like the AMLYN drive) There have certainly been some strange additional variations! -- Grumpy Ol' Fred cisin(a)xenosoft.com

I'm having problems remembering off the top of my head, but didn't Commodore use 100tpi drives rather than 96tpi drives for the 8050/8250 IEEE-488 drives?

I'll stick my 5.25" MFM (PC, essentially) disk FAQ at the end of this message. You (or anyone else) is welcome ot make use of it in a more comprehensive FAQ provided you give me an acknowledgement.

I have a 5.25/3.5" combo drive that handles all four of those formats (and 3.5" 360K) just fine. The gorgeous part is that it occupies a single 5.25" bay.

I'd like to make a box with four floppy types: 5.25" 360 and 1.2, and 3.5" 720 and 1.44. I seem to recall someone here did this.

I imagine I can use the integrated floppy controller and then add another board to provide the additional controller. I'll check the DOS manual to see if it says anything as Nico indicates.

You don't really need a 3.5" 720K drive. Unlike the 360K/1.2M situation with 5.25-inch drives, the 1440K 3.5" drives can reliably read and write 720K diskettes.

IIRC, there was a driver supplied in some (but not all) versions of DOS called DRIVER.SYS which had to be installed using a "DEVICE=" command in CONFIG.SYS in order to support extra floppies. I don't recall whether it allowed specifying a base IO address for a secondary controller. They may have removed some functionality in later DOS releases.

To add to the fun,... SOME drives had a media sensor that would adjust the write current for HD diskettes, others required that come from a signal from the controller. The early PS/2 drives did NOT have a sensor. They would "successfully" format a 720K diskette to 1.4M. (Reliability was less than perfect due to 600 Oersted v 750 ) But then, when that disk was put into a drive that had a sensor, the drive would have problems with it.

Yep. DOS had only one quirk, though: If you wanted to format a 720K disk in a 1440K drive, you had to specifically tell DOS to format as 720K. Otherwise, it assumes you're going for 1.4, attempts it, and fails. You can choose /f:720 with later versions of DOS; use "/t:80 /n:9" for 3.x.

I used to run four (5.25") drives with no problems. My NEC Multispeed laptop also had a provision to connect it to a PC's external drive connector and you could then access the Multispeed's two 720k 3.5" disks as the 3rd and 4th drive on the PC. I made a cable to connect the Multispeed to the PC but I loaned it and all my NEC Multispeed manuals to one of the members of this list and he never returned them. In both cases, I had to add a line or two in the Config.sys file to make DOS recognize the 3rd and 4th drives.

In a REAL IBM PC? You bet! The original IBM PC Floppy drive controller had a 37 pin connector on the back panel. You could connect two drives to the internal connector and two to the external connector. Many of the thrid party FD controller cards also had the external drive port. Yes DOS will recognize the 3rd and 4th drives as C: and D: if no hard drive is installed. I don't remember what drive letters it assigns if a HD is present but I THINK they will still be C: and D:.