HP6034A TMS9981 CPU won't exit reset state.

Here's one that has me puzzled: I have a HP 6034A power supply that does
not come out of self-test when powered up. Internally I traced the problem
to the TI TMS9981 CPU itself
(Datasheet http://www.datasheetarchive.com/pdf/3434384.pdf)

On power up, the Interrupt code present on it's IC0 - IC2 pins is 000
(reset) it never exits reset. Tracing back through a few TTL glue chips I
found that a quad latch has no clock with which to pass D to Q. The clock
for the latch is derived from the TMS9981's pin 21 "03" clock output -
which appears to be floating at 2 volts. If I isolate pin 21 of the CPU,
it goes to a low state. It's not an open collector output AFAICT, so I
check the CLKIN and OSCOUT pins (33 and 34) where the crystal is located.
Looks fine at 9.2MHz, levels look good. Check +5 and +12 and GND - looks
good. So, must be a bad CPU, right? Nope. I place a "known good" CPU in -
same result, no "03" clock output on pin 21. I'm stumped or missing
something here... Anyone have any ideas? Even a "That makes no sense to me
either" would be appreciated.

Many thanks!

 

Is something holding up the "o3" o/p?

 

By "holding up" do you mean another device on the trace that is defective?
If so, I don't think so - as I said, I lifted and isolated the pin 21
clock output on the CPU chip (it's in a socket) and it's output went to
zero volts, but that's it. Near as I can tell from the datasheet, all that
could be required for the clock output would be GND, +5, +12, and an
oscillator which appears to be functioning. It makes no sense whatsoever.

Thanks Graham.

 

So with the IC pin lifted, does the socket trace still show +2V?

 

Yes, but in my experience that would be expected, as TTL inputs would do
that if nothing was driving them.