HP E3610A Power Supply Help

[glowell]

Hi -

First post to the forums, and I'm a relative beginner to analog electronics.  I'm trying to repair an HP E3610 from around 1992.  When I received the power supply, the AC fuse was blown, and on opening the cover I found several of the electrolytics were leaking.   I replaced all 5 of the large electrolytics and thoroughly degreased the board.  There were no obvious signs of other damage. The power supply is mostly functioning, but isn't stable.

On power-on, with the voltage and current knobs full counter clockwise, the voltage reads around -10 volts, and slowly over 5-10 minutes will creep up to +1-2 volts, then settle at around  0.25 volts.  The CV and CC indicators are both on, but eventually the CC indicator goes off.  Any change of temperature, like a draft, will cause the voltage to drift upwards.  Both the voltage and current adjust knobs will change the settings, but the displayed voltage is higher than measured with a DMM at the output.  I'm waiting for a couple power resistors to arrive before checking out the current settings.

The display voltage at TP9 & TP10 is 4.98 V.

The reference voltage section reads: TP1 12.14 V, TP2 6.92 V, TP8 -11.94 volts  using +output/TP6 as reference.  CREF/TP7 tracks the CC current pot from -0.21 to -6.64.

With the CC current set to 3.0 amps, 1 V on the DMM reads 1.49 V in the display, 2 V on the DMM reads 5.21 V, 8 V reads 11.51 V.

The voltage at the bridge rectifiers is CR13 8.55V, CR9 37.33V, and CR2 16.86V (US 120V power).

I have the 2000 Schematics, but it has some slightly different components than I see on the board.

Any advice before I start messing around with the innards ?

Thanks,
Gary

[singapol]

Reading the troubleshooting flowchart should help.

[glowell]

Thanks Singapol.  The document that I have only has a Performance and Calibration procedure.  I didn't find anything on troubleshooting or fault isolation.  Is there another version of the document I should be looking for ?  I found copies dated 1992, 2000, and 2007 but they are all similar.

Thanks,
Gary

[free_electron]

-replace ALL electrolytics.
- check for open traces, especially around the electrolytics and around the little round to-39 transistor.
- clean off the pcb with alcohol
- take the opamps out of the board and clean under the opamps ...

when revising these supplies i do all of the above by default. the opamps cost like 1$
i spend 5$ in parts ( swap caps and opamps.) and the supply is good as new

[glowell]

I will scrub the board again, and check traces.  There were a few traces showing some darkening under the conformal coating but so far I've not seen any broken or corroded traces.

I have parts on the way for the remaining electrolytics, and the op-amps.  Any issues putting the replacement op-amps in sockets ?

Thanks
Gary

[bingo600]

I have a E3610 that's "acting" , and am a bit lazy 

I'd like to replace the CAPS & Opamps too.

Would you mind posting your components list

/Bingo

[glowell]

Hi bingo600 -

My first go was to replace all the large electrolytics and I did see some improvements but not fully functional.  The original caps were Nichicon and I found pretty much exact matches at Digikey:

       Original Part          DigiKey  Part             1992 Schematic
C2  10,000uF 50V        493-8667-ND
C3   470uF 50V            493-12254-1-ND
C7   330uF 35V            493-12024-1-ND
C13  330uF 35V           493-12024-1-ND        470uF 50V
C17 4700uF 50V          493-11080-ND

I'm still waiting for the second batch of components to arrive.  Those are the small 1uF electrolytics and the op-amps & comparators.  I'm still learning about op-amps, but I've read that op-amps in circuit could be checked by comparing the differences between inverting and non-inverting inputs, and non-inverting input and output.  The differences should be a small value in a working circuit.  That was not true for U1 the LM741 op-amp. Unfortunately when checking the next IC I must have shorted something and blown F1 and F2.  I need to get long pointy probes for my meter.

        Original Part                        Digikey Part                         1992 Schematic
C9     1uF 50V                             493-11596-1-ND                  Same
C15   1uF 50V                              “”                                         “”
C19   1uF 50V                              “”                                         “”
U1    LM741  Op-Amp                  LM741CNNS/NOPB-ND           LF411 Op Amp
U3    LM329  6.9V Ref                  N/A                                     LM336BZ5.0 5V Ref
U4    LM1458  Dual Op-Amp         296-1403-5-ND                    LF442CN Dual Op-Amp
U5    LM393  Dual Comparator     LM393NFS-ND                       Same

The unit I have uses a few different parts than are shown on the 1992 schematic.  The data sheets imply the new parts are functionally equivalent, but I'm sticking to the parts I found on the PCB for the time being.  I think I will use sockets for the ICs.  I've heard there are signal integrity issues using sockets, but from what I can tell that's more relevant for high frequency circuits.

Cheers,
Gary

[Kleinstein]

Frequencies (and GBW) are not that high that sockets would be a problem. Some cheap sockets might cause contact problems after long time and when corroded. Parts that get rather hot could be a problem.

With the set current set all the way to zero, the supply might act a little unpredictable without a load. So for a first test one should have a small load (even if only in the 10 K range) or set the current to a little more than zero. I don't know this supply in detail - but well possible that there is a trimmer to set the minimum current, that might need adjustment.

Changing the OPs from 741 to an LF411 might also change other things and might need adjusted values for a few caps. In the regulator circuit it is well possible that a few of the OPs are not always working in the linear range and thus might have a non zero voltage between the inputs. In many circuits this happens to either the current or voltage regulator most of the time.

The LM329 reference should be OK if it still reads about 7 V - changing to a 5 V one would need extra adjustments. Also the LM329 is better (more stable, lower noise) than the 5 V LM336.

[bingo600]

@glowell & Kleinstein

Thank you both.
I will try to locate the components in EU

/Bingo

[glowell]

The additional parts arrived, but it was a week before I could work on the power supply again.  Added 2 fuses to the list of parts:

F1,F2 1A 125VAC fast blow PCB mount fuse, Digikey Part F2313-ND

Replaced the fuses, the three 1uF electrolytics, and the three op-amps.  Power supply works much better - voltage meter and DMM agree within a couple millivolts.  I will check the current later when I've got hold of some some higher wattage resisters.

So in the end, no real systematic debugging, just replaced the all electrolytics and op-amps.

Thanks,
Gary