staffan@phos.serum.kodak.com (Kenneth Staffan (x37507)) (03/15/91)
I often use floppies to port stuff between my 3b1 at home and PC-AT at work.
I always format the floppies at home. When I write the disk at home & read
it at work, it _always_ works. I have a problem going the other way, though.
9 out of 10 times (or more, it's very frustrating) when I write the disk on
the PC and try to read it on the 3b1, I get a "can't read file allocation
table" error. If I take the floppy back to the PC, it reads it fine. This
happens whether I use a freshly formatted floppy, or one which I just used
for a 3b1->PC transfer. Anybody have any suggestions?
Thanks,
Ken
---
I know it's a big company, but I'm not really nobody@kodak.com, try
staffan@serum.kodak.com -or- kes7613@ma.rit.edudt@yenta.alb.nm.us (David B. Thomas) (03/15/91)
staffan@phos.serum.kodak.com (Kenneth Staffan (x37507)) writes: > [ his unix pc can't always read a perfectly good msdos disk ] metoo. If anyone has suggestions, please post. It's aggravating. little david -- Bottom of stack = 0x40000 Stack pointer = 0x3fffe Don't push it!
mvadh@cbnews.att.com (andrew.d.hay) (03/15/91)
In article <1991Mar14.212645.10354@ssd.kodak.com> staffan@phos.serum.kodak.com (Kenneth Staffan (x37507)) writes:
"I often use floppies to port stuff between my 3b1 at home and PC-AT at work.
"I always format the floppies at home. When I write the disk at home & read
"it at work, it _always_ works. I have a problem going the other way, though.
"9 out of 10 times (or more, it's very frustrating) when I write the disk on
"the PC and try to read it on the 3b1, I get a "can't read file allocation
"table" error. If I take the floppy back to the PC, it reads it fine. This
"happens whether I use a freshly formatted floppy, or one which I just used
"for a 3b1->PC transfer. Anybody have any suggestions?
the at has a 1.2M drive, right? this is a common problem in
transferring disks between 360K and 1.2M drives; the low density drive
writes wide tracks which the high density drive can read, but the high
density drive writes narrow tracks the low density drive can't read.
--
Andrew Hay +------------------------------------------------------+
Ragged Individualist | JAAAAAAANE! |
AT&T-BL Ward Hill MA | HOW DO YOU STOP THIS CRAZY THING? |
a.d.hay@att.com +------------------------------------------------------+dold@mitisft.Convergent.COM (Clarence Dold) (03/16/91)
in article <1991Mar14.212645.10354@ssd.kodak.com>, staffan@phos.serum.kodak.com (Kenneth Staffan (x37507)) says: Subject: Re: Do you use MS-DOS format floppies? > it at work, it _always_ works. I have a problem going the other way, though. This has been noted here also. From previous scenarios involving Tandy PC with 360K floopies, I am lead to believe that the quality of the 360K drive is the culprit. If you have access to another 360K PC drive, see if it will interchange properly. If it does, try substituting that drive into the 3B1. I would be interested in the results. The old drives in the 3B1 seem to give some truth to the statement in the MSDOS manual that claims 1.2 -> 360 transfers are unreliable. -- --- Clarence A Dold - dold@tsmiti.Convergent.COM ...pyramid!ctnews!tsmiti!dold
jwbirdsa@amc-gw.amc.com (James Birdsall) (03/16/91)
In article <1991Mar14.212645.10354@ssd.kodak.com> staffan@phos.serum.kodak.com (Kenneth Staffan (x37507)) writes: >I often use floppies to port stuff between my 3b1 at home and PC-AT at work. >I always format the floppies at home. When I write the disk at home & read >it at work, it _always_ works. I have a problem going the other way, though. >9 out of 10 times (or more, it's very frustrating) when I write the disk on >the PC and try to read it on the 3b1, I get a "can't read file allocation >table" error. If I take the floppy back to the PC, it reads it fine. This >happens whether I use a freshly formatted floppy, or one which I just used >for a 3b1->PC transfer. Anybody have any suggestions? Since you have a PC-AT, I suspect that your floppy drive is high density (1.2M). These drives can write low-density (360K) floppies, yes, but there is a catch: the head is only half as wide, and the write current (and hence field strength on the disk) is much lower. The head width causes problems like so: if you have written to the disk with a 360K drive, there is data across the whole track. When the high density drive writes, it writes a strip down the center of the track, leaving the old data around the edges. This old data can potentially overwhelm the desired signal and you get read errors. OLD_DATA_OLD_DATA_OLD_DATA_OLD_DATA \ width of 1.2M head { NEW_DATA_NEW_DATA_NEW_DATA_NEW_DATA | width of OLD_DATA_OLD_DATA_OLD_DATA_OLD_DATA / 360K head Even with a freshly formatted floppy, the head width can cause problems by corrupting the file allocation table. HOW TO GET AROUND THIS: for transfers from the PC/AT to the 3B1, use a virgin disk (or one that you have bulk erased) that was formatted to 360K on the PC/AT. Write to this disk ONLY with the high-density drive, and there will be no old stray data to cause trouble. The low-density drive should be able to read this disk, unless the field strength is extremely low, but I've never seen that happen in practice. For transfers from the 3B1 to the PC/AT, use a disk formatted to 360K on the low-density drive. This disk does not have to be virgin, since the low-density head is wide enough to completely overwrite anything that was there before. Write to this disk ONLY with the low-density drive, since writing to it with the high-density drive will cause the read errors that started this discussion in the first place. As long as you abide by the rules on what drive to use for writing, either of the drives can be used to read any disk. This topic has been discussed for years in the MS-DOS world; the preceding techniques have proven about as reliable and foolproof as anything can be. -- James W. Birdsall WORK: jwbirdsa@amc.com {uunet,uw-coco}!amc-gw!jwbirdsa HOME: {uunet,uw-coco}!amc-gw!picarefy!jwbirdsa OTHER: 71261.1731@compuserve.com ========== "Think of an animal that's small and fuzzy." "Mold." -- RM ========= =========== "For it is the doom of men that they forget." -- Merlin ===========
dt@yenta.alb.nm.us (David B. Thomas) (03/16/91)
mvadh@cbnews.att.com (andrew.d.hay) writes: >In article <1991Mar14.212645.10354@ssd.kodak.com> staffan@phos.serum.kodak.com (Kenneth Staffan (x37507)) writes: >" [ often cannot read perfectly good MSDOS disks with 3b1 ] >the at has a 1.2M drive, right? > [ suggests the high-low density issue is the problem ] No, that's not it in my case. I have taken virgin floppies, formatted them on a low density drive on a dos box, and copied files to them. On moving them to the 3b1, some read with mtools, and some do not. Those that do not read can still read with dd(1) from beginning to end, so the 3b1 can read the data, it just can't understand it. I think mtools needs a little improvement (or I need a newer version). little david -- Bottom of stack = 0x40000 Stack pointer = 0x3fffe Don't push it!
floyd@ims.alaska.edu (Floyd Davidson) (03/16/91)
In article <6222@amc-gw.amc.com> jwbirdsa@europa.amc.com (James Birdsall) writes: >In article <1991Mar14.212645.10354@ssd.kodak.com> staffan@phos.serum.kodak.com (Kenneth Staffan (x37507)) writes: [...] >>9 out of 10 times (or more, it's very frustrating) when I write the disk on >>the PC and try to read it on the 3b1, I get a "can't read file allocation >>table" error. If I take the floppy back to the PC, it reads it fine. This >>happens whether I use a freshly formatted floppy, or one which I just used >>for a 3b1->PC transfer. Anybody have any suggestions? > > Since you have a PC-AT, I suspect that your floppy drive is high density >(1.2M). These drives can write low-density (360K) floppies, yes, but there >is a catch: the head is only half as wide, and the write current (and hence >field strength on the disk) is much lower. > > The head width causes problems like so: if you have written to the disk >with a 360K drive, there is data across the whole track. When the high >density drive writes, it writes a strip down the center of the track, >leaving the old data around the edges. This old data can potentially >overwhelm the desired signal and you get read errors. > [...] >Even with a freshly formatted floppy, the head width can cause problems >by corrupting the file allocation table. > > HOW TO GET AROUND THIS: for transfers from the PC/AT to the 3B1, use a >virgin disk (or one that you have bulk erased) that was formatted to 360K >on the PC/AT. Write to this disk ONLY with the high-density drive, and there To sum this up, if you even one time write the disk with the wide track head of a low density drive, you can never expect to successfully write it with a high density drive. (The high density drive will be able to read what it wrote, but the low density drive will get errors.) What I wanted to add to this has to do with bulk erasing. If the disk has ever been written to with a low density drive it must be bulk erased to be used by a high density drive reliably. Even though the high density drive will probably read what it wrote, the head alignment becomes very critical, and with only a slight mis-alignment even the hd drive will be reading errors from the wide track written by the ld drive. Commercial bulk erasers probably do a good job. I don't know because I haven't used one since zapping 2" video tape with one in '66... What I do use is two LARGE speaker magnets. I put one on each side of the disk and move them around the entire disk. I do one at a time. These magnets are about 4" in diameter and almost 1" thick. Anything less, even only one magnet, won't do it well enough! The fact that it takes a huge magnet to really zap the disk is a bit surprising. Even a small magnet will destroy enough data that the disk is useless, but for this it needs to be *totally* zapped. The way I discover how much it takes is interesting. I had a Kaypro-4 in 1983, which would write single or double sided disks, and automatically detected which it was when reading them. The difference was detected by trying to read the second side. If it didn't get an error it was a double sided disk, if it got the error then it was single sided. Now take a previously double sided disk and try to erase it well enough that it could be formatted again as single sided... It has to be well enough zapped that absolutely nothing shows up on the second side, or else the formatting goes just fine and you end up with a double sided disk formatted for one side only. A genuine single sided drive can use it, but a double sided drive can't. So you write to it with a Kaypro-2 (SSDD) and it can read it fine. But the Kaypro-4 couldn't use it at all (DSDD drives). One large magnet would almost always erase the disk well enough, but one out of ten or so wouldn't be useable. With two magnets it never failed. If you want to mix and match high density and low density disks, or 48 tpi and 96 tpi disks, get a couple of the biggest magnets you can find. Give every disk a good once over before re-formatting, and plainly mark which kind of drive it was formatted on. Then do not ever write to it with a different type of drive. Making sure that no part of the disk is written to just about requires that it be write protected whenever it is inserted into a drive of a different type than the formatting. Floyd -- Floyd L. Davidson | floyd@ims.alaska.edu | Alascom, Inc. pays me Salcha, AK 99714 | Univ. of Alaska | but not for opinions.
emanuele@overlf.UUCP (Mark A. Emanuele) (03/17/91)
In article <1991Mar15.001223.18393@yenta.alb.nm.us>, dt@yenta.alb.nm.us (David B. Thomas) writes: > staffan@phos.serum.kodak.com (Kenneth Staffan (x37507)) writes: > > > [ his unix pc can't always read a perfectly good msdos disk ] > > metoo. If anyone has suggestions, please post. It's aggravating. > The problem may be because the unix-pc formats ms-dos floppies to 8 sectors/track and your IBM is formatting 9/tk. There is an option in dos format to give you 8 sectors / track. This SHOULD solve the problem. -- Mark A. Emanuele V.P. Engineering Overleaf, Inc. 218 Summit Ave Fords, NJ 08863 (908) 738-8486 emanuele@overlf.UUCP
tom@afthree.as.arizona.edu (Thomas J. Trebisky) (03/20/91)
In article <1991Mar14.212645.10354@ssd.kodak.com> staffan@phos.serum.kodak.com (Kenneth Staffan (x37507)) writes: >I often use floppies to port stuff between my 3b1 at home and PC-AT at work. >I always format the floppies at home. When I write the disk at home & read >it at work, it _always_ works. I have a problem going the other way, though. >9 out of 10 times (or more, it's very frustrating) when I write the disk on >the PC and try to read it on the 3b1, I get a "can't read file allocation >table" error. If I take the floppy back to the PC, it reads it fine. This >happens whether I use a freshly formatted floppy, or one which I just used >for a 3b1->PC transfer. Anybody have any suggestions? I am glad you asked for suggestions, because I am not sure I have a definitive answer - but I can make plenty of suggestions. First a question - since you say an AT, I wonder if you are writing on a HD drive, there are some known troubles going from the HD drive to "normal, i.e. 360K drives" even between PC's. I will assume this in all that follows. OK, here is my guess - the 1.2Meg (HD) drives on the AT have thin (narrow) heads and 96 tracks per inch. The drives on most XT's (and the unix PC) are 48 tpi drives with fatter heads. The unix-pc writes a wide swath of data and the AT reads right down the center and no problem. However the AT writes a narrow swath of data, and the head is not wide enough to erase the wide swath (written at format time on the unix-pc -- formatting does fill the sectors with data), so when you take the floppy to the unix-pc, it's wide head sees both the data in the center of the track, and the slop on the sides and bad things happen. The only problem with all this is the following -- I have tried this every possible way and cannot replicate the problem. Even formatting on the 360K and then writing on the 1.2M drive works every time. (I have an AT running DOS 4.1 with both 1.2M and 360K drives -- and have tried all the possible combinations of which drive I format on and which I write the data on, and I can always read them on the UNIXPC (but I am using the UA window driven MSDOS floppy reader). Aha -- maybe here is the rub?? When I use mtools I get all manner of chaos -- floating exception core dump is most common, and this traces back to a divide by zero in init.c in init(), after reading sector zero -- the boot block -- divides by zero calculating: tracks = WORD(psect) / (heads * sectors); (Why should there be a boot block - this is not a bootable disk). My conclusion is that mtools has some bugs. Maybe DOS4.1 has changes in the disk format mtools cannot deal with? (Mtools seems fine to format and write files on the floppy -- the AT reads them fine -- the only wrinkle here was writting a file not in the current directory to an empty floppy produced nothing on the floppy.) So my question to the original poster is: What version DOS are you running on your AT? What floppy hardware are you using on your AT? What software are you using on the unixpc to read the floppies? I am tempted to dive in and tackle an mtools fixup -- any comments? -- Tom Trebisky ttrebisky@as.arizona.edu (Internet) Steward Observatory University of Arizona Tucson, Arizona
ward@unix386.Convergent.COM (Ward Griffiths) (03/21/91)
The one thing that I have seen that ALWAYS works, that I
have been using since the Tandy 2000 showed up with 720k
drives, to avoid the problem with transfers between HD/LD
drives, is to bulk erase the disk first and format and
write it from scratch on the machine I'm transferring the
data from. That way, there is no leftover wide data with
narrow data down the middle. At Radio Shack, you can get
a bulk tape eraser for about $15 that will save endless
headaches. For a couple more dollars, there is the video
version that (besides being good for reusing video tapes)
is rugged enough to clear up all those leftover problems
with a glitched DC600A. Just keep the tape eraser away
from your hard disk. Having the computer in the den and
the eraser in the kitchen has always worked for me.
--
Ward Griffiths, Unisys NCG aka Convergent Technologies The people that make Unisys' official opinions get paid more. A LOT more.
=========================================================================== To Hell with "Only One Earth"! Try "At Least One Solar System"!
"... the device every conqueror, yes every altruistic liberator should be required to wear on his shield ... is a little girl and her kitten, at ground zero." Captain Sir Dominic Flandry