HP 7908 woes

From: Tony Duell <ard_at_p850ug1.demon.co.uk>
Date: Tue Jun 8 18:03:18 2004

> By pulling boards I was able to isolate the problem to
> the regulator board which contains three 12V supply
> circuits, one each for the tape, logic and drive

I assume it's one common fuse at the input to all 3 regulators.

> spindle. I was able to identify high power thyristors
> and op-amps on each 12V supply suggesting some sort of
> crowbar arrangement for overvoltage protection (no

Very likely. HP were good about providing crowbars on power rails (well,
apart from in the 9815 and 9825 calculator, but I digress). Your problem
could well be due to one of the crowbars firing.

> schematics available and this is a four-layer PCB
> which makes it hard to derive the actual circuit). I

Wimp ! (:-)). 4 layes is hardly difficult. Now 8 layers, with all the
signal tracks inside, that's a little harder (and I've done that and
survived).

> haven't checked the thyristors for shorts yet but I
> believe the problem is either a defective thyristor or
> some defect in one of the voltage comparators, maybe a
> dead zener or something.

Or a shorted pass transistor (I asusme this is a linear regulator, are
there any fat inductors on the board?). Any chips -- watch out for 10 lead
metal can devices in HP power supples. These are nearly always 723
regulators (and might be used in linear or switching circuits!). If the
chips have HP numbers (1820-xxxx, 1826-xxxx), post them and I'll look
them up for you.

A shorted pass transistor sounds likely, if only because I'd expect
current limiting in a supply like this. If the supply overcurrents
(possibly due to the crowbar firing), then the thing should remove the
base drive to the pass transistor, shutting the supply down. Of course if
the transistor is shorted, this doesn't do a lot, hence the blown fuse.

>
> I need to keep the fuse from blowing long enough to
> figure out why the overvoltage protection is
> triggering. What would be the risk in trying to defeat
> the overvoltage protection without any load connected

If it's a linear supply, there should be no danger at all. A switcher
could be 'interesting' under such conditions (you might blow other
components in the supply), but there are problems with it anyway, so it
probably wouldn't do that much harm in any case.

But I'd check the power transistors for dead shorts first.

> to the regulated 12V lines ? I assume I could do this
> by disconnecting the gate of each thyristor.

I'd just desolder the thyristors completely. And remove _all_ other
boards from the instrument, unplug the drive power cables, etc. Don't
trust that they don't use the 12V line, get them out of the way.

Pity I don't have this device, or I could probably be a lot more help...

-tony
Received on Tue Jun 08 2004 - 18:03:18 BST