HP4396A power supply malfunction : SOLVED

[HP4396A]

Greetings !
I have a HP4396A network/spectrum analyzer, and yesterday the power supply failed.

The diagnostics tells me there is a problem in the post-regulator.
The symptoms have included a sour smell for some time and unstable RF level on the display that stabilizes as something warms up.
Now the RF level jumped outside the display area, so I suppose some OP-AMP voltage is wrong.
I have gathered that the power supply is fully undocumented and the post -regulator is quite complex,
with a lot of pins towards the backplane.
But in the best case I could find a physically damaged component and swap it for a working one.

I saw that "mark03" attempted to repair a AGILENT 4396B which seems to have the same/similar post-regulator board.
https://www.eevblog.com/forum/repair/agilent-4396b-spectrumnetworkimpedance-analyzer-repair/

To begin with it would be useful to know if the input voltages to the post-regulator are fine.
The measured voltages on the "10 pin rainbow cable" are as follows.
1: -18 VDC
2:  +17.8 VDC
3:  GND
4:  GND
5: +7.6 VDC
6: -7.9 VDC
7: +70.5 VDC
8: +25.6 VDC
9:  +2.5VDC, 3VAC
10: GND

So I was wondering if anyone could help me verify if those voltages are fine ?

[m k]

Seems that you must rip it apart a bit.

Try to find a source of that smell, pictures are not bad either.
My guess is that 2.5V regulation is not nice.

[chilternview]

Probably a capacitor gone in the power supply, open it up and inspect it, you should be able to find the one responsible.

If the +5V from the pre-regulator is down then the CPU will be non-functional including the LCD display logic. If not, you may get some error messages of the form 'POWER FAILED ON...'

Is the fan working? That is supplied from the post regulator. If it is, check if the shutdown LED is on.

Next, is the A1 5v red LED on? If not the 5v line is down and the problem could be in the A40 pre reg or the A50 DC-DC converter.

All of this is described in the 4396 service manual in some detail,  which you should be able to find online.

[HP4396A]

Managed to find the problem in the A2 post-regulator. At startup the voltages had volts of noise and over-voltage on them.
It was possible to measure the A2 post-regulator with the internal test programs quite conveniently.

At closer inspection the ground end of a voltage reference (zener I guess) was corroded. Some cleaning and soldering and all is well 

[m k]

Did you find any possible cause for that corrosion?

It's not necessary anything else than that exact solder joint.
Lead-tin is not necessary a stable compound in long run.

Other possibility is that some corroding elements are around.
Then leaking capacitors are the first suspects.