Date   : Mon, 12 Nov 2007 07:09:26 +0000
From   : mike@... (Mike Tomlinson)
Subject: Disc interface problem [was Manufacturing fault]

In article <d8834d20711100453h3398aecbic69021e88132f64e@...>,
Alex Taylor <zeem.uk@...> writes

>> Check the NMI interrupt link - S9, IIRC.
>
>The machine's at work at the moment, but from memory the link has been
>snipped open. Is that right?

I can't actually remember, but judging from the circuit diagram, yes.
Check S2 (Econet NMI) as well as this also drives NMI.  What does NMI on
the 6502 (pin 6) do when you issue a *. ?

Also check that the drive itself is working!  Does the light illuminate?
How long have you waited when the Beeb hangs to see if an error
eventually appears?

Have you checked every single chip in the disc interface against the
installation instructions to see if they are all present and correct,
and that the links are correct?

What issue is the PCB?  I seem to remember changes are needed to issue 3
(and maybe issue 4) boards to make the disc interface work.

>I've been probing around the disc connector while a drive is plugged
>in, and while I type commands. For example, the DS0 line is high until
>I type *., at which point it drops low.

Sounds good.

> Typing *.1, it stays high.

Sounds OK.

>However the DS1 line seems to be permanently low, regardless of what I
>type. The motor on line ('load head' in the service manual) seemed to
>be doing something, but now seems to be stuck low too.

My first line of attack would be to change out ic79.  This needs to be a
7438 (not 74LS38), and note that pins 6 and 11 of ic79 also drive an
input on ic83 and ic84.  Are those chips present and correct?

> I've tried
>swapping out IC79 that drives these lines, but to no avail.

...  you've already done that.  Next step would be to substitute the
8271, and in view of your discovery of a bent pin on the 8271 socket,
I'd take out the chip and check every single pin for continuity to its
destination (i.e. check pin 1 connects to ic79 pin 1, etc.)  Use a very
thin piece of wire wrapped around one of your meter probes to insert
into the socket pins; you don't want to force the probe in and wreck the
socket.

>Next thing I want to look at is what the pins on the 8271 itself are
>doing. One of the things I'm struggling with at the moment is which
>direction the disc interface signals work, relative to the motherboard
>connector.

Use the circuit diagram.  If pins on the disc connector are connected to
outputs of chips on the BBC, the BBC must be driving those pins.  A
glance at the cct diagram shows the inputs from the drive to the Beeb
are read data, write protect and Trk 0.

> I.e., does the drive itself put 5v on to the wires, which
>the Beeb then pulls low? (I'm learning this as I go along, and some of
>the theory still escapes me!)

The disc drive signals are active low. The terminator in the drive will
pull most signals high (check the drive does have a terminator!)  There
are pullups in the Beeb for the inputs - R49 for read data, R22 for
write protect, and R23 for Track 0.

-- 
(\__/)   Bunny says NO to Windows Vista!
(='.'=)  http://www.cs.auckland.ac.nz/~pgut001/pubs/vista_cost.html
(")_(")