Dying VaxStation 4000 VLC - help?

From: Mark Tapley <mtapley_at_swri.edu>
Date: Fri Nov 15 00:58:01 2002

All,
        More info:

>Linear Technology 1.5A adjustable voltage regulator, TO-220 package.

Has about the same voltages on its pins when the system is either running
or failed:
Vin 5.4V, Vout 3.5V, Adj(gnd) 2.1V

Should that last be 0.0? I'm referring all of these measurements to Chassis
ground, obtained by wedging a lead against the power supply housing.
Measurements power-off with the Ohmmeter say that last pin is about 75 Ohms
away from Chassis. Ironically, that's more than the readings for either Vin
or Vout.

>Low Power Low Offset Voltage Dual Comparator (LM193)

This guy is intimately linked to the failure.
V+ connected to 5.4 V (Red wires)
Gnd connected to ground (Black wires = chassis ground.)

Output A obeys the following rule:
        running fail
out A 5.4V 0V

One other interesting thing is that the first few times I touched "out A"
with the probe, it failed immediately. I actually *did* get to see the
needle swing a couple of times, as it tried to decide whether to run with
the probe on that lead. Eventually it decided to run even with the probe on
the output. The meter claims 20,000 ohms/Volt DC, so I did not expect to
crash the system by probing it. I tried Out B, but it didn't change while I
was watching and I didn't record what state it was in.

Thermal dependence is still flaky, but it did go rapidly and persistently
toes-up when I turned the hair-dryer + funnel combo on the LM193. I only
got to do one trial, though, before it got to the point that it wouldn't
run long enough to convince me the heat gun was doing the deed, so that
could have been coincidence. I tried a few other parts of the circuit
board, and got similar results out of a big chip near the SCSI connector.
Dunno if that's relevant.
                                                        - Mark
Received on Fri Nov 15 2002 - 00:58:01 GMT