Troubleshooting an IBM 5155 - (portable personal computer)

[Tony Duell]

> > > Qume drive working, but don't know where to go next. I 
> > guess I could swap
> > > the boards between the two units, but that'll only tell me if it is
> > > electronic or electro-mechanical. 
> > > 
> > > Other than that, any suggestions for figuring out if it is 
> > alignment, broken
> > > sensor, or something electrical? 
> > 
> > The alignment adjustments, in general, are what are known as 
> > 'interchangeability adjustments'. In other words, if they're 
> > a bit off, 
> > the drive will be able to format, read, and write its own disks, but 
> > those disks will not be useable on other drives, and vice versa.
> > 
> > So what you do first is to hook the drive up as B: and try 
> > formatting a 
> > blank disk in it. Note that if, for some reason, the machine 
> > thinks it 
> > only has one drive, it'll map A: _and_ B: to that drive. 
> > You'll get some 
> > obvious messages if this has happened, in which case the problem is 
> > certainly not just alignment.
> 
> Well, I did this and learned something VERY important. The disk only
> formatted as 180K. So, it doesn't see that it is a double sided drive. The

In wich case there's no point (yet) in trying to do any form of 
alignment. There is a real fault.

> same diskette put back in the working drive formatted to 360K. So, how do
> these drives tell if they're dealing with SS vs. DS meadia? Does it just
> look for a signal from the second head? Is there another sensor that
> indicates "yep - double sided". This is how it works with 8" drives. The
> index hole is in a physically different spot.

Unlike the 8" drives, there is no signle/double sided signal on the 5.25" 
drive interface connector, and there's no sensor for anything like that 
(I looked at the schematics this morning, they look pretty standard, 
there are 3 sensors for index. write protect, and track 0). 

My guess is that either : 

a) The side select signal does nothing due to a logic fault in the drive, 
and that therefore when it tries to read the second side, it gets the 
data from the first side. A true signle-sided drive would do this.

b) There is no useful signal from the side-1 head (this is the head on 
the top surface of the disk, of course), either because the head is 
dirty/defective, or a damaged connection back to the logic board, or an 
electronic fault on the board. I am not sure what the software would do 
under those conditions -- by rights it should give errors for every 
sector on side-1, but it's possible it then assumes the media is really 
single-sided, and formats it accordingly. 

I think i will have to grab the schematics and see if I can talk you 
through the circuitry.

> in. Any tips based on the schematic would be of help. Since all I have at
> the moment is a DMM and a homemade logic probe, my tools are lacking to
> fully solve this problem.

Well, although _now_ I have some very nice test gear, I debugged my first 
homebrew computer (This was when homebrewing meant soldering chips to 
stripboard and wiring everything up, not pluging a video card into a 
motherboard) with an anlaogue multimeter and an LED+resitor 'logic 
probe'. So it certainly _can_ be done...

-tony