19 Jul
2013
19 Jul
'13
10:15 a.m.
On 2013 Jul 19, at 8:03 AM, Dave wrote:
I have a 9830A that I haven't powered up in a while. Previously,
it would take a few seconds to get to the lazy-T prompt, but now on
powerup, the screen remains blank. I got Tony Duell's phenomenal
schematics, and was all set to start debugging, starting at the
clock, but I checked the user manual first, and following the
suggestion in the setup section, I hit "STOP", which brought me to
the prompt, and from there, it appears to be functioning
normally. However, the machine powers up this way every time--
never to a prompt.
Does anyone have an idea why the machine is behaving this way? I'd
love to have it power up to a prompt.
As Dave suggests, one place to start would be to check the power-on-
reset circuits. The main POR circuit for the 9830 is a 74121
monostable on the power supply board. The signal it generates is
available on pin 3 of the little 12-pin service connector situated
between boards 12 and 13 on the backplane. The signal is supposed to
go, or stay, low for around 100mS around power-up, after which it
goes high, releasing the processor to run.
There are actually two POR circuits in the 9830 though. The second
one is in the tape drive circuitry, on board 62. It may be that if
the tape drive POR is sluggish, the CPU is hanging on the tape drive
during the initialisation process.
Original HP schematics and some theory-of-op for the power supply and
tape drive are in the 9830 service manual from the HP Computer Museum.
"9830ACalculator-ServiceManual-09830-90030-140pages-Jan76.pdf" from:
http://www.hpmuseum.net/exhibit.php?hwdoc=55
There's actually a little timing graph on the power supply schematic
(pg 129) for the power-on signal.
Also, does anybody happen to have schematics for the
infotek
bitslice processor? I don't have one, but I'd be very interested
in trying to acquire or recreate one.
I have access to a set of the infotek CPU boards through fellow list
member Rob here and would some day like to reverse engineer them.
I have been reverse engineering the 9830 recently (a couple things
left to clean up in the schematic), and Rob and I are currently
working on a project regarding the 9830/9865 tapes. Between all that
and other projects I don't know when I'll get to the infotek boards.
I saw the boards briefly and IIRC they were pretty dense multilayer
boards, so RE'ing could be fun. It would be nice if the originals
schematics showed up but I'll amazed if they do, might be an idea to
try contacting the infotek principals to see if they have docs still.