Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uwm.edu!psuvax1!psuvm!jiw2
From: JIW2@psuvm.psu.edu (John Wagner)
Newsgroups: comp.sys.ibm.pc
Subject: Re: Floppy drive / controller woes
Message-ID: <90065.133116JIW2@psuvm.psu.edu>
Date: 6 Mar 90 18:31:16 GMT
References: <2039@aipna.ed.ac.uk>
Organization: Penn State University
Lines: 49


Hi folks

>I wonder if there's anybody out there who can help me with what looks  ke
>like a floppy drive controller problem.

>I have recently bought a second hand 720k 3.5" drive to be used as a seond
>drive (B:) on my AT clone (a Tandon Target running MS-DOS 3.3 - drive A
>is the usual 5.25", 1.2M). It seems to be working fine in that it readsand
>writes ok, but it looks as the directory contents are being buffered ina
>funny way.

>If I start up the machine and do DIR B:, I get the contents of drive B:as
>expected. If I then change the floppy in drive b:, and do DIR B: again,
>I still get the contents of the first floppy, even though the light on he drive
>goes on and it looks as if the machine is reading it. In fact I don't
>think it is: if I leave drive b: empty and do DIR B:, the light goes on and
>I still get the contents of the previous floppy. The only way to find ot the
>contents of drive b: is to do DIR A:, and then DIR B:. That seems to mae it
>forget the previous contents of b:.

>I vaguely remember that CP/M had a bug -sorry, feature- like that, in tat you
>had to tell it that you had changed floppy in the drive, or it wouldn't
>notice. But it's the first time I've seen this on a clone 8-{

>Can anyone give any advice? Is this a problem with the drive or the conroller?
>(the controller is on the motherboard, btw).

jlb@aipna.ed.ac.uk

John:

After installing a CompuAdd 1.44M floppy drive and controller in my
old XT Portable, I had exactly the same problem.  I posted this
to my local computer club BBS, and the reply I got from a local PC
repairman was to disable lead 34 between the drive and the controller
card, either by a jumper if there is one, or by putting a piece of tape
over the #34 contact on the drive.  I was told this is a very common
problem.

I can't vouch for this fix, since I decided to wait until the next time
I open up the PC to try it out.  If you try it, please post the results
to this newsgroup.  In the meantime, I have found that doing a CTRL/C
after changing disks forces a read of the disk's FAT, and DOS then knows
you have a new disk in the drive.  Pretty heavy-handed, but it works
as a stopgap measure.


                                       John Wagner