HP 6038A DOA, troubleshooting, FIXED

[valley001]

Hello all,

I found myself in need of a smallish power supply and learned of these system power supplies made by HP.  I really like the quality of the older HP gear so I searched one out. 

I will not have my example until I pick it up next week.  It "does not power up" so I welcome informed guesses on potential avenues of investigation.  I have the service manual and plan to follow the flow chart there, as a start.  Anyone have experience with these?  I see limited info out there aside from some general commentary.   As I improve my troubleshooting skills I have become less inhibited re the purchase of dead units, hopefully that does not bite me with this one 

[Jay_Diddy_B]

Hi,

First of all you can get the service manual directly from the Keysight website. Search for 6033A, the 6033A and the 6038A share the same Service Manual.

These supplies share a lot of the circuitry found in the 6024A. You can read about the 6024A here:

https://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1981-08.pdf

Troubleshooting

Before I get into the details, these power supplies are off-line switching supplies. A large portion of the circuit is connected directly to the line. Read and understand all the warnings in the HP Service Manual before attempting repair of these units.

The nature of this circuit, is that if you don't repair everything, it will fail catastrophically again. It is very important to find all the damaged components.

First I would do a visual inspection. Check the electrolytic capacitors, especially in the bias supply have not leaked and corroded all the tracks on the board.

You can also check the main power MOSFETs.

If the MOSFETs are damaged you will probably have to rebuilt all the circuitry on the gates.

Have a look at this thread:

https://www.eevblog.com/forum/testgear/hp6012a-1kw-power-supply-repair/msg1019892/#msg1019892

Hard to give more advice until the unit arrives.

Jay_Diddy_B

[coromonadalix]

yep   As Jay wrote    do a visual check for fuse(s)  power switch ....

This psu is one tough cookie,   if everything seems visually fine,  i would isolate the main board from the power mesh section, hoping to test all the bias power supply

Careful  you will have two ground levels in this psu, the high voltage ground  on the input rectifier   4a 4c,  and the "low power section" bias power supply   triangle 1 and 3  after the r50 0.7 ohms

It must be like a floating supply section

On the input filter section 12a, 12c   with  4a, 4c   you will have a high voltage 240v and up  be careful

Follow the service manual trouble shooting  section

[valley001]

Thank you both for the insight.  I am studying the service manual/schematics and learning as much as I can about this unit.  I will plan to do a visual inspection and perhaps some testing of caps. Look for discoloration etc.

I assume if a fuse is blown I am in for some fun.  I see two fuses on the A8 board along with the primary fuse on the back panel. 

[valley001]

I now have the unit in hand.  It is in fairly good shape cosmetically, nothing obviously toasted and I believe I am the first one to open it up.  F2 on the main board blows immediately when the power is applied. 

I checked the 120v winding on T3 and it seems ok, 8-9 ohms or so.  Further testing of the main board may require additional disassembly.  My plan next is to test the filter caps, rectifier diodes etc.  Anything else I should look at?

[valley001]

I removed the A4 power mesh board and attempted to apply power with a fresh F2, it blows immediately. 

F1 and 2 on A4 are intact, the FETs seem to be OK.  No obviously blown components there.

There is an audible buzzing coming from the IEC socket when the power cord is installed.  R20 and 21 are getting warm (not much but noticeably more than the surrounding components, up to maybe 90 deg F.  This is all with the power switch in the off position and F2 empty.  I have checked C2-4 filters in situ and they are at least not a dead short, they test 400-900uf with my DMM. 

I am thinking my problem is on the A1 main board, wondering if you all agree before I disassemble further (removing A1 for complete access requires total disassembly of the unit, more or less).

[George Edmonds]

Hi

Buzzing IEC's are not uncommon, at lot of Chinese ones are not made to the correct dimensions and make a poor contact.

Your best move is to put a 100W incandescent bulb is series with the incoming phase, this prevents instant fuse blowing.  BE AWARE THAT THERE IS SOME 360VDC AT LOW IMPEDANCE THROUGHOUT THE HIGH VOLTAGE SECTION, THIS WILL KILL YOU.

George G6HIG   

[bdunham7]

I'd think you'd want to pull out or disconnect the control board to see if the short is in the rectifiers or transformer section on the main board or further down the line on the control board.

[valley001]

Yes, disconnecting the A4 and A2 boards still results in a blown fuse.  I do need to fab up a dim bulb tester so I stop blowing them. 

I have been testing through the components on the A1 board.  All of the diodes seem OK.  I have checked most of the windings on T3 and the readings seem reasonable, 9 ohms for the primary for instance.  The caps I suppose will require removal, none measure a dead short they all register capacitance on my DMM.  Is there a way to narrow down a bad cap short of disordering them?  Hate to de solder when I don't have to, especially with these really nice HP PCBs. 

EDIT- I also checked the fan motor winding and it measures around 215 ohms. 

I do appreciate the input.

[George Edmonds]

Hi

Where are you in the world and what is your mains voltage?

Personally I would pull the A1 board and do a DC resistance check across what looks like R9 1M, the schematic that I have is of poor quality.  If not short circuit then apply mains with a 100W or lower wattage incandescent bulb in series across R9 DO NOT DO THIS WITHOUT THE BULB IN SERIES as there will be no fuse present.  It is not clear from the schematic as to how line module connects to the A1 board.

If the bulb flashes on power up and then goes dim of goes out all is well with the A1 board, if it remains at full brilliance you  have a fault on the A1 board.

Leave A1 board for some time before handling it to allow main capacitors to discharge via R20/R21 and then check with DMM that there is no voltage present across R20/R21

George G6HIG

[George Edmonds]

Hi

Having looked further at the manual the A1 board is the mother board, just pull all other boards and cary out my suggested tests.

George G6HIG

[bdunham7]

This is where the iso-variac earns its keep.

With A2 and A4 removed, there doesn't seem to be much left.  Measure the resistance across C26 with F2 out, that should be your primary resistance.  If that is low, check C26 and the voltage selection switches. Also test between both sides and ground to make sure the transformer hasn't shorted internally.

I doubt the caps have shorted, and if the diodes all test good in circuit, the next step IMO is to desolder the transformer secondary windings one by one, starting with 9 & 11, then 6 & 8.  Or, if you have a DBT or Variac, and a current clamp, you can test for secondary current--there shouldn't be any with everything unplugged. Also, can you tell me where 14/15/16 at the bottom go to?  If that unplugs, perhaps try unplugging that first.

[valley001]

I am in the USA, 126ish volts this evening. 

Interestingly R9 is a jumper on my example.  This is odd because the schematic shows R9 as 1M, but the parts list for the A1 board indicates a "wire, tinned copper 22awg."  Measuring across the mains input yields 1M which made sense to me when I saw R9 in the schematic but now I am not sure what is going on there. 

[George Edmonds]

Hi

May I suggest that you contact Artek and get a better copy of the manual, they have it for $10.  The copy that I have from the keysight web site  is almost unreadable,  My advice to use a series bulb has not changed as this will prevent damage whist putting full voltage on the components. 

George G6HIG

[valley001]

Strange, I have a copy from Keysight as well and it is very clear.  Perhaps there is more than one available through them?

Here are the secondary winding resistance measurements, t3 in situ:

6-junction of CR6/CR9- 3.2 ohms
7-junction of CR6/CR9- 1.6 ohms
6-7- 2 ohms

9-11- 8.2 ohms
10-9- 4.1 ohms
10-11- 4.1 ohms

14-16- 1.2 ohms
14-15- .5 ohms
15-16- .7 ohms

These seem reasonable no?

I am suspect of the voltage selector switch, still puzzling on how to test this and verify it is not shorted.  Just need to study it more to understand its connections. . 

[valley001]

T3 primary:

5-4- 9 ohms or so.

I read open between what I interpret as the core/ground of T3 and any of the primary or secondary leads. 

George-  I would like to carry out your test but I am not clear on it.  I should insert a 100W or less incandescent across the R9 1M position (jumper in my example), am I to jumper the F2 position or should that be open?  I apologize if this should have been obvious. 

[bdunham7]

14-16- 1.2 ohms
14-15- .5 ohms
15-16- .7 ohms

I am suspect of the voltage selector switch, still puzzling on how to test this and verify it is not shorted.  Just need to study it more to understand its connections. .

Where do 14/15/16 go to?  I can't seem to find where they go anywhere on the diagram.  1.2 ohms seems very low unless it is a very low voltage winding like for a VFD.  And this is not a VFD display, is it?

As for the voltage selector, test it the way I said to start.  Pull F2 and measure across C26.

[valley001]

I actually cant find that either (on the diagram).  Found it:  14/15/16 go to a three wire cable that connects from A1J2 to A8J2 J10 on the GPIB board, where it is met by diodes for rectification.  (top left of figure 6-9A)

Across C26 with F2 removed I get 8.7 ohms or so. 

[bdunham7]

I actually cant find that either (on the diagram).  14/15/16 go to a three wire cable that connects from A1J2 to A8J2, where it is met by diodes for rectification. 

Across C26 with F2 removed I get 8.7 ohms or so.

OK, so it is power for the GPIB board and the schematic for A8 seems off--on the diagram I think it would be J10.  Just disconnect it for now to eliminate that as an issue.

If you have reasonable resistance across C26, then your voltage selector and primaries are likely fine.  You could try setting the voltage selectors for 240V, verifying the resistance at C26 goes to about 16 ohms.  Then you can try one more fuse at half-voltage and if it doesn't blow, maybe you can find your short with some quick measurements of transformer output.

[valley001]

Hmm..with the selector set for 240v I get 35 ohms across C26.  I'm guessing this is not a good sign for the 120v primary winding.

EDIT- But then, primary winding 1-3 also measures 8.6ish ohms.  I am not sure how the 240v position yields 35 ohms.   

[George Edmonds]

Hi

I have now found a better copy of the service manual/schematic, there are multiple versions on the Net. most of which are incomplete.

When I get back home later in the day I will respond to you.  HP have thought about how to fix this psu and have built in test facilities

George G6HIG Dover UK

[HighPrecision]

Hmm..with the selector set for 240v I get 35 ohms across C26.  I'm guessing this is not a good sign for the 120v primary winding.

EDIT- But then, primary winding 1-3 also measures 8.6ish ohms.  I am not sure how the 240v position yields 35 ohms.   

Hi

looks correct to me, 8.6 Ohm in 120V setting is the result of TWO windings in parallel, each of 17.2 Ohm.

Setting to 240V the resistance is double: 17.2 Ohm, windings in series 34.4 Ohm.

[valley001]

That makes sense, I did not think about 120v windings being in parallel. 

Am I correct that F2 blowing is related to T3 and then the associated rectifier U1 and filters C20/21?  Can CR1-4 and C2-4 be related to this issue?

[bdunham7]

That makes sense, I did not think about 120v windings being in parallel. 

Am I correct that F2 blowing is related to T3 and then the associated rectifier U1 and filters C20/21?  Can CR1-4 and C2-4 be related to this issue?

I didn't follow the primary switching either, although it is pretty standard.  How they get 100V is sort of interesting, but also common.  Anyway, that looks fine.

As far as F2, if the schematic is correct and you have removed the A2 and A4 boards and disconnected the cable to A8-J10, then your problem is confined to the area to the left of connector XA2P1 on the main board--T3, the rectifiers, capacitors, resistors and wiring.  Unless I'm wrong or something weird is going on.  I would be pulling out the transformer leads in the order I mentioned before, unless that is very difficult.  I'm not sure how you tested the rectifiers, but my prime suspect would be U1.

[valley001]

The trouble is I do not get any dead shorts in U1 measuring it in place. Is this a flawed methodology?

T3 is soldered directly to the board, so it is not trivial removing leads from the circuit.  Would be easier to remove U1 I think.