Agilent E3632A Power Supply - No Output

[skennedy]

I purchased a broken Agilent E3632A power supply with the intention of repairing the unit but it has really stumped me. The unit seems to be in very good cosmetic condition. The power supply does not output anything other than -0.05V regardless of the settings. I can verify that the display is reading the output voltage correctly by confirming the voltage externally with a multi-meter. The system self test and the complete self-test pass with no errors.

I've done a full visual inspection of the unit and nothing seems out of the ordinary. The supply rails are all at the correct voltage with little ripple.

I have probed around the ADC but don't seem to see any action on the output. There is digital SPI going into the ADC but probably nothing coming out. I suspected that was the issue but though it was strange that the self-test was passing without fault. I relented and replaced this IC but this has had no effect.

Anyone have any suggestions?

Kind Regards,

skennedy

[jackthomson43]

Seems like problems in LCD 16x2 display. Have you checked its wiring ? You have to check all the wirings using multimeter.

Btw buying a faulty electronics is never a good idea, its almost impossible to resolve the issue.

[Old Printer]

Btw buying a faulty electronics is never a good idea, its almost impossible to resolve the issue.

This is simply untrue. This forum if full of stories of members buying non-functioning gear that has been satisfactorily repaired, many times by no more than a loose connector or switch in the wrong position. I got a great deal on a Tektronix Power supply. I spent $15 on a service manual and weeks tracking down a bad 10 cent diode, but as a noob, what I learned in the process was huge. The item needs to be acquired at a good price to allow for the possibility of failure, and the availability of a service manual cannot be understated, but with so many pieces of electronic equipment needing no more than a re-cap of the power supply section it certainly is worth investigating.

[ArthurDent]

If you don't have the service manual, this may help.

https://doc.xdevs.com/doc/HP_Agilent_Keysight/E3632A/pdf/E3632_service_sch.pdf

I have one working and one E3632A with a problem. The problem one works to set up V/I on the display but when I switch the output on, the display reads all zeros. The correct output is at the binding posts and I can change the settings if I view the output on external meters but the display isn't switched on when the output is. The supply is accurate according to the external meters so I'm assuming it is just electronic switching that isn't working. I'll get to it one of these days

I still believe buying high end broken equipment is a good idea if you like a challenge. 

[WattsThat]

Not really sure what you’re telling us. Can you set the desired output voltage to 10.00 volts? Then, regardless of what the desired output voltage is, the actual measured voltage output remains at -0.05 volts?

Have you verified all four of the floating bias supplies are good? They’re listed on page 98 of the service manual.

[cbutlera]

Faulty static RAM perhaps?  My E3642a suffered from this problem, and I seem to recall there are other forum members with E3630 and E3640 series power supplies who have also had RAM failures.  In my case it was heat sensitive, fine when cold, but it would crash and reset after warming up.

[skennedy]

Thanks for your replies everyone. Certainly not my first repair, but normally things are a bit more obvious. I've previously repaired an E3641A which had a very obviously blown up current shunt.  Also jackthomson43 faulty electronics area always a lot more fun than the working kind. I've bought heaps of non-working equipment and repaired almost all of it. Water damage is the only thing to stay well clear of. HP/Agilent even publish full schematics. Learnt a lot more about electronics fixing broken things than at University.

I'm satisfied that the digital circuitry is working as everything is behaving as expected. GPIB is even working and the front panel is behaving as expected. I've done some reading elsewhere that a few other people have had a similar fault with the servo control system. Perhaps an op-amp that has some connection to the front terminals has taken a strike. I'll continue to trace out the circuit and work out where things go strange.

I'll keep you updated.

[Miti]

Have you tried to calibrate it? Maybe it just lost its calibration constants. The calibration procedure is very simple.

[alm]

If I'm understanding the problem correctly, the problem is that the voltage programming is not working, while the readback on the display matches the actual output voltage. I'm confused why you'd consider the ADC to be an issue?

The DAC is multiplexed, so you would need an oscilloscope to check its output. I would start by measuring after the multiplexer, looking at the schematic, the outputs of U31 (CV_REF, CC_REF, etc) look like a convenient place. If those are correct, then I would look at the regulation servo, but also check that the overvoltage crowbar (CR28) is not conducting. If the outputs after the MUX are not correct, then the MUX, DAC and opto-isolators might be suspect. That's my $0.02 based on a brief look at the schematic.

[Mr. Scram]

Seems like problems in LCD display. Have you checked its wiring ? You have to check all the wirings using multimeter.

Btw buying a faulty electronics is never a good idea, its almost impossible to resolve the issue.

Whether it's a good idea depends on your skill set and situation. If you don't have much repair experience or don't have extensive funds and need it to work it may not always be a success. If you accept it just may not work out for you, go for it.

[skennedy]

Yes, I've measured the output voltage and it appears that the LCD is matching the actual output voltage. All of the keypad works also so I'm pretty certain that is all ok. It probably also means that both of the processors are running (front panel and main board).

Thanks alm, I definitely meant DAC not ADC  . Seems I have been working with ADCs so much recently I got a bit finger tied. I will have a look at the output of the multiplexer to see if thing are looking sensible. I hadn't considered the crowbar circuitry. I have disabled the OVP and OCP. This threw a spanner in the works for my last power supply repair.

To be honest, I probably wouldn't have very much gear at all without repairing broken equipment. My lab is built upon equipment that would have cost someone tens of thousands before they either the equipment was surplus or broken. Still to this day my best purchase was a spectrum analyser for $300 that had a shorted tantalum cap on the ADC power supply.

[Ussovak]

Overheating? Check the direction of the fan. Might be turned around blowing into the case. This is incorrect. The fan should be mounted to pull air through the side vents and blown out the back.

See capture from service manual below.

[skennedy]

Sorry for the delay in getting back to everyone. Lockdown has limited my access to my lab here in Australia.

I fixed the supply quite unexpectedly today. I always suspected something odd was going on with the DAC, I finally took the entire board out of the chassis to enable better probing and found the issue pretty much straight away.

Turned out that pin 16 on the DAC was floating rather than connected to the supply. One end of the ferrite feeding that pin was not soldered to it's pad. When I probed on the ferrite it read ok but at the pin and on the pad of the ferrite it was floating. A reflow of the solder and the supply is back and working like new.