Recalcitrant Fluke 335A outputs only negative 1.24v

[enut11]

The story so far:
I had just finished repairing a premature voltage trip problem on my 335A

https://www.eevblog.com/forum/repair/fluke-335d-trips-on-higher-voltages/

and was rejoicing on having a working Voltage Standard. Then, during some routine testing I placed an ammeter across the output (dialed at 4v). Normally, the output current would be limited to whatever is set on the current limit pot, up to a max of 60mA. I heard a vibration noise so I immediately disconnected the ammeter and placed a voltmeter across the output. It read approx minus 1.24v and would not respond to any setting of the output dials. Further investigation showed a blown 1/8A fuse at the back of the unit but when I replaced it, it immediately blew again.

I am currently going through the troubleshooting procedure on the manual and researching data on the Forum and the net. I am also systematically replacing all old electros with new ones. Meanwhile, any suggestions to speed things along would be appreciated.

Latest version of my 332/335 service stand attached. I added removable side cheeks of 16mm chip board so the 335 unit can be supported vertically either face up or face down.
enut11

[Cubdriver]

Nice stand!  It is a heavy beast, isn't it?

At this point it's way past my bedtime and I'll not be much help regarding troubleshooting, but 1/8A sounds like a line fuse (though a quick glance at the gazillion sheets of schematics I've printed out but not yet taped together doesn't readily reveal its location in the circuit) - I'd start by checking for a shorted rectifier(s) - if something went awry in the current limiting circuitry, one of them could have been overloaded.

Good luck with it!

-Pat

[enut11]

Found the location of the blown fuse - F1 on the High Voltage section. Lots of circuit to troubleshoot after it...

Mine had a 1/8A fuse installed. The schematic and specs call for 1/4A slow blow although I don't think using the larger rated fuse will help as I had the 335 working nicely with the 1/8A for some time before the latest problem set in.
enut11

[Cubdriver]

I just dug deeper and saw that.  I noted mine had a 1/4A, and figured it was line related when you said yours was half of that.  It's beyond the HV bridge rectifier, so that lessens the possibility of it being a shorted rectifier, though I'd still check them.

I suppose I'd then go on to look for shorts in the pass element transistors.

-Pat

[Dr. Frank]

That doesn't sound so good.


Check all semiconductors of series pass element A7A1.
Then also check the pre-regulator board, if all semiconductors are still fine, and if the small signal transistors have sufficient beta.

I assume, that the HV pre-regulation is not working correctly, or is not steered correctly, so that the discharge reed-relays K1A or K2A are energized and shortens the HV output over the 2k power resistor R19.
If the pre-regulator A7A2 is damaged, specially the power zener diode CR14, or Q1, the diode bridge will be shortened and HV will be generated full power, which the series pass element then tries to clamp down.

I had that latter problem..

For a quick check, you may also remove (pull out) the preregulator A7A2 board first, and apply an external voltage from a PSU, current limited, between -output, and at the F1 fuse, at the node K1A/K2A /R10.
The voltage should be at least 10V higher, than the programmed output voltage, for checking the correct output regulation.

That's safe, as no HV is present , but you have to engage the safety switches, if the aluminum cover is removed.
What happens on K1A/K2A?
Check, if the uni-junction circuit Q3 across Q8 produces pulses if the voltage across Q8 is too low..
I think, 36V is required at least, but I don't remember any more correctly.

Frank

[enut11]

Thanks Dr Frank. I checked Q1 on the Preregulator PCB and, as far as I can determine, it is shorted between collector and emitter. Q1 is a very large cylindrical semiconductor the likes of which I have not seen before. I have not been able to determine its specs from the circuit or the manual.
enut11

[Edwin G. Pettis]

That is a nasty transistor to find a replacement for, essentially it is being used as a crowbar to shutdown the circuit so its exact parameters aren't too important.  A 2N2769 I think is pretty close to what this is in the same case, a 2N5883 in a TO-3 case is the same, just a different form and easily found.  The exact transistor can be found from parts brokers (who tend to be quite expensive), probably on eBay (under the Fluke number 193953) for a high price as well.  If the other transistors in the circuit are okay, you could always chance firing it up and seeing if it runs, something triggered this circuit, possibly the load you put on the output may have done it but this transistor shouldn't have failed just doing its job 'usually', who knows it might just have been its time to go, it normally does nothing.  I think I have an old Westinghouse data catalog in my archives by my wife has buried some it behind a wall of stuff, no simple matter to move it.

[DavidAlfa]

I repaired a 335A having this issue (-1.5V output), but happened randomly.
Ultimately it died completely and it was sent to me.

There was a huge damage caused by leaking capacitors, which I replaced with radial ones, keeping them as far as possible from the power resistors and fixing them with some silicone.

The transistors could be saved by brushing them with a glass-fiber pen, discarded the sockets as they were pretty bad too, I only kept the plastic part as spacers.

The diodes where badly damaged too, but needing nothing special, I just replaced them all with 1N5407 (3A 800V) or 1N4007 (1KV 1A).

Finally the random output issue was caused by K1.
Check K1 coil voltage at the pre-regulator board, should read about 60V (DC mode).
The coil was defective, ramdomly going open, in that case the voltage will read about double (126V).
However most of the time it fixed itself when touching with the DMM tips, so it was a real PITA to catch!

Finally I adapted a 48V relay, adding a 3K3 series resistor with R4 to keep the coil voltage around 40V, it's been working non-stop for the entire day.