[time-nuts] HP 5328 PSU nightmare... Or stupid engineer, you decide...
masondg44 at comcast.net
Thu Sep 10 02:58:45 UTC 2009
> I'm not familiar with this particular instrument, but a standard
> technique for linear power supplies is to hook it up to a variac. This
> lets you turn down the line voltage so you can do some measurements
> without smoking the system.
> Douglas Wire - PUPCo Studios wrote:
>> Good day everyone and thank you all for hosting this wonderful community
>> and allowing me to participate. I have several HP5328 with the ?really-
>> nice? newer 10811-xxxxx Oscillators in them. I have found while I have
>> used the good old gold trace reliable HP instruments all of my life,
>> units have been especially difficult. The first unit the 4500uF
>> electrolytic?s went bad and produced essentially a dead short; an easy
>> enough repair for me to not only track down in minutes, but it only takes
>> straight bit screwdriver to fix in seconds!
>> Now our second unit has been giving me fits and while I would agree 100%
>> with one of the posts I saw here about how well HP did not only with
>> schematics, but also the wonderful troubleshooting flow charts usually
>> repairs on any of their old units a breeze. Sadly I have a unit here that
>> is giving us fits! It is a PSU issue and not related to the Motherboard
>> any of the cards as I tested it with everything unhooked/ unsoldered and
>> still got the same result. It is quite similar to what we see when we get
>> an old HP unit that has a fried cap and is darn near creating a short to
>> ground, but alas I simply cannot find the problem (I am sure it is
>> me in the face is and I just can?t see it?) What I am seeing is super
>> high current flow through the R1 (I believe, but HP?s every unit I have
>> ever serviced had.47? resistor, NOT a 22-? as is stated in the
>> schematic?) that leads to F1. The troubleshooting is complicated by the
>> fact that unless I want to smoke that heavy duty, relatively close
>> tolerance resistor, I cannot even check voltages anywhere for it will
>> the fuse or if I put a slow blow to try and catch some measurements in a
>> second or two, well that is not very feasible either.
>> If I had to guess, I would say it has either a cap that has fried,
>> chance of a transformer issue, or the way this thing reacts, pretty well
>> effective dead short somewhere, but I will be damned if I can find the
>> problem anywhere. I replaced the bad and 4500uF caps as well as the
>> rectifier, wondering if part of it had blown with no change in its
>> One cannot follow the flow cart to much of anything other than boxes that
>> say look for a short, but so many areas one tests even on a perfectly
>> working unit come clear down near the zero ? point even when they are
>> operating correctly.
>> I apologize if 1) this is not a clear email that anyone can easily
>> understand and 2) I almost feel embarrassed to ask anyone for advice from
>> their practical experience, for I feel as If I should easily be able to
>> to the bottom of this in a matter of minutes with the wonderful data HP
>> provides us all for these old workhorses.
>> So if anyone has run into a problem such as this in the past where
>> the flow chart only yields No, No, No -> check for shorts and has any
>> advice for how I might logically proceed, or what in fact you have found
>> out in dealing with a similar problem, it would be of great help, as we
>> need this in-service ASAP, but I guess I just cannot see the forest for
>> tress in front of me or something here? Any advise, suggestions would be
>> greatly appreciated.
>> Warm regards,
>> Douglas M. Wire, GED, FNA,
>> PUPCo Studios, PUPCo Research Group
Start by making sure that the line voltage switch is set correctly.
Barring that as a cause of the problem, you will have to start removing
components that are fed by the +12V supply until the overcurrent is gone.
First choice would be CR1 (12V Zener), then ICs U1, U2, U3, U4 and U5. When
bad component is removed, the current should drop back to normal, or
slightly less. That might be a lot of unsoldering, but unless you want to
start cutting traces on the PCB, it's the only way to proceed.
If you have a good ESR meter, it can be used as a low ohms meter to locate a
short or low resistance.
You might remove R1 and install a higher value, higher wattage resistor so
that you avoid further damage while you make measurements. Measure the raw
supply voltage (+25V); make sure that it's the right voltage and low ripple.
Measure the voltage drop across R1 in an operating instrument, so that you
have a reference of what to look for in the bad guy. From the voltage drop,
you can calculate the current going through it. Use a sensitive voltmeter
(5 digits or better; you want to see millivolt drops along the traces) to
trace the voltage drops along the circuit traces to find where the majority
of the current is going.
Good luck with the troubleshooting.
masondg44 at comcast dot net
More information about the time-nuts