HP 6200B power supply blowing fuses

[GreyWoolfe]

Lowimpedence, I did check that the secondaries were isolated from the primary.  I did not check to see if the secondaries were isolated from each other.  I will check that.  SeanB, I did replace all 4.  Actually I replaced all the rectifier diodes when I found 3 of them on 2 secondaries in pieces.  However, what I bought were "NOS".  I am not sure of the quality of them.

[sync]

Have you installed the diodes with the right orientation/polarity?

[pomonabill221]

AC on caps means shorted diodes. Did you replace all 4?

True... but in the case of C16, that is across the transformer secondary BEFORE the bridge, so it will have AC on it.

[pomonabill221]

You could try..
remove tap 14, and using an ohmmeter (not on diode check), measure across the AC input of the bridge.
That would be on CR26 K/ CR29 A to CR28 A/CR27 K.  Basically you are measuring to see if the diode(s) in the bridge are shorted... probably not, but that would be what I would check next, although acording to your scope pic, you are getting full wave AC on C13/14.
There appears to be a VERY low resistance or a short, in the power section and that is what is dragging the whole supply down.

Now we need to go through that section.

Just a quick glance at the circuit and I would question CR34.  That is one of the few components that is directly across the output and seems to be there to protect against reverse voltage being applied from the outside world.  Also, C19 (paralleled to CR34)
This is just a guess (hunch) but you could just measure across the output jacks with you ohmmeter as well.  I don't think you should see very much of anything as far as continuity, but this test might also confirm whether or not there is something going on on the output side of the supply.
This test should be easy (for starters).

Another thought.... see if the terminal strip on the back is strapped correctly.  Someone could have missed a terminal and placed a short across the output.
Also, C20 might be shorted... you can check that by removing either the jumper on A10/S+ or A9/S- and probe A9 to A10.

I am trying to find the easiest and "most possible" that I can think of before really diving into the supply.

[lowimpedance]

Since the thread is getting rather long some tests and checks have already been done, and I too cannot remember all .
The output strapping was checked as well as C20 etc , cannot remember if CR34 was checked (It should be of course being right across the output!).
 As Sync asked .. you did put the diode bridge back with the correct polarity?. What diodes did you use (part number) ?.
When you had the transistors removed what was the DC volts measured on the filter caps (correct polarity) ?.

While the taps 13 and 14 are isolated can you load those taps with a resistor (say 100 ohms /10W) for a short time to measure if the secondary volts drop significantly. Just to satisfy the winding is able to hold up under a reasonable load.

Do you have another or can borrow a power supply with current limit control?.
I think it had been suggested already!. Connect an external supply at around 25V across the filter caps (transformer taps 13/14 disconnected!).
Set the current limit low and then slowly increase to see if the hp6200 supply comes up or continues to draw too much current.
If it looks good ie no excessive current then do the same procedure again with the bias supplies operating (with bulb tester in place and only the main taps 13/14 disconnected). Does the supply start to work or does it still drag too much current?.

This repair is becoming a real mission! , When you look at the cct. there is not that much in it.
 lots of learning at the least.

Maybe a nice photo of the main bridge and filter caps area of the PCB.

Just a last thought, there is no chance a previous owner had done something silly like put a screw or some such in the wrong place causing a mechanical short, did the supply look 'fiddled with' when you first received it?.

[GreyWoolfe]

I want to thank everyone for their help.  Unfortunately, the saga is over.  I have been working on the psu without case which is basically the pcb and front and rear bezels.  I had it standing up on it's side-don't remember exactly why.  I didn't secure my office door properly, and my stepdaughter's (damn) cats got in and while playing around, managed to knock the psu off my workbench.  PCB cracked almost in half.   The lessons I learned is to make sure all DUTs are sitting flat and make sure the office door is secure.  I'm thinking dinner tonight is linguini con gatto al fresco (linguini with fresh cat)   That is why I prefer dogs as pets.  Time to move on to the next project.

[pomonabill221]

OOOHHHHHHH NNNOOOOO!!!!!!!!!!!!!!!
SO sorry to hear this!!!!!!
This was really getting interesting!!!!  and I TRULY hoping that you(we (all of us here)) with our inputs and suggestions, could find the problem!!!
This REALLY put a damper on my day!  (strange that an inanimate object, and a person that is trying to repair it, could have such an effect! guess I need to get a partener!) hehe

Well you can't say "we" didn't try!  hhmmmm...cat on a stick!

[SeanB]

Put the kit back together and post it to me.....

[robrenz]

Just pretend that you bought it that way and it's just going to be a little bit bigger challenge. If it is just a double layer board all the traces can be jumped across the crack on both sides. Don't let those felines get the best of you.    Come on, you can fix this thing.

[lowimpedance]

Bit of a shame  , Now you have spares for another 6200B (got cheap of course!) .

(Looking at the bright side BTW)

[GreyWoolfe]

Bit of a shame  , Now you have spares for another 6200B (got cheap of course!) .

(Looking at the bright side BTW)

Actually, I am considering getting another, but one that works 

[pomonabill221]

Bit of a shame  , Now you have spares for another 6200B (got cheap of course!) .

(Looking at the bright side BTW)

Actually, I am considering getting another, but one that works 

Now THAT'S the spirit!!!  Just think, you have spare parts as well!

[orbiter]

Damn that's unlucky.  I was following this thread with interest and was sooo disappointed when I read that the cats had killed off your chances of a repair :-(

Sent from my HTC One using Tapatalk Pro.

[charles]

So, I may be able to keep the broken power supply love alive. I recently bought a 6200b off ebay. It worked until I turned the voltage to maximum, around 40V. I heard a click and the gauge on the front went from reading 40v to being off the scale below zero, I.e. all the way to the left. It still worked, i.e. it was outputting 40v and could still be adjusted from 0 to 40v but the display was broken. Also, the thing will no longer go into the 5v range, i.e. the 5v range goes from 0 to 40v. So, I measured the voltage across the terminals of the gauge/meter and found that it was reading approximately -150 mV and would go up to about -100 mV when the desired voltage was increased. Any thoughts?

Also, as a heads up I am a noob but a determined noob. To prove this statement, I have an embarrassingly stupid follow up question. In my haste to l look around inside, I decided to take out the meter knob, I.e. cut all its 4 million leads and set it aside. I took pictures before doing this but unfortunately not very good ones. Is there any chance someone still has this supply and could tell me how this knob is wired up?

[Rigby]

 

[free_electron]

The cat may know...

[Andy Watson]

The cat may know...

+1

[charles]

Re-read original "joke" that I wrote in this post and decided it was too macabre and needed to die... So I removed it. Anyway, I really would appreciate any (preferably non-feline) help that anyone is willing to give.

[Andy Watson]

If you have not already done it, read the whole of this thread and find/download the links to the service manual(s). Good luck ... you're certainly going to need it!

[charles]

 

I fixed the 'major' fault! This is the first time I have ever fixed a piece of electronics before! Hooray!



I started by following the 'though shalt measure voltages' and found that the voltages were a bit out, at least according to the datasheet. Namely, my +6.2 volt rail was +6.4 and my +12.4 rail was +12.51. So, I looked at the regulator circuit, wasn't sure how it worked exactly, so I read up on zener based regulators and then simulated the circuit in LTSpice. Interestingly, the simulated regulator agreed with the voltages I measured and not the voltages on the datasheet. So, I left that line of inquiry for the time.

Then, I went and looked at what I should have in the first place, namely the meter circuit. I read up on differential amplifiers then simulated the circuit. I compared my simulated voltages to the voltages in the meter and the base of one of the transistors was way off (Q12). So, I replaced it with a new transistor and bingo, the meter works again!

Thanks to everyone who provided advice and links to manuals.

Now, the remaining thing that is slightly weird is that the 20v and 40v range selection seems to be way off. Specifically, when unloaded, the 20v range seems to be able to go to 40 volts and the 40v range seems to go above 50 volts (I don't really want to see how high in case I blow something up). I don't really understand the mechanism which limits the voltage in the first place, so I will need to read further into the schematic. However, it appears to be at least usable in its current state.

Anyway, I am happy with my mini-victory.

[megajocke]

The voltage range of the voltage control doesn't change with the range selected. It's just that you are not guaranteed the power supply will be able to supply more than 20 V over the full load range in the 20 V range and likewise for the 40 V range. It's not just that you have more ripple and voltage drop at the smoothing capacitor but one also has to account for line voltage variations. The rating is 105-125 V (or 210-250 V) so the worst case would be full load current at 105 V line voltage.

The voltage range of the control is determined by the +6.2V reference voltage, the voltage control pot resistance and the R12//R13 resistance.

There are some good description in the service manual:
http://cp.literature.agilent.com/litweb/pdf/06200-90001.pdf

From the table of specifications you can find that the ouptut voltage is 1 V per each 200 ohms of programming resistance. Therefore I'd expect the voltage control to have a 50 V full scale range. This also matches calculations using the R12 value well. I can't find any information on what the R13 value is, but it is likely it is (or could be) selected for a precision 200 ohms/volt programming gain.

That your 6.2 V voltage is 3 % off and 12.4 V is 1 % off is perfectly normal. The reference diodes (1N821) used to provide the reference voltage do not have an initial tolerance better than 5%. Their main feature is that they are temperature stable, which is the requirement here.

[charles]

Hi megajocke,

Thanks so much for your very detailed reply!

The point about the line voltage ratings clarifies the situation greatly. Thank you! That being said, do the guaranteed power supply voltage ratings also indicate the maximum output range that should be used? In other words, if I put the range on 40 volts and crank the voltage to maximum, say 55 volts, is it likely that I will see the magic blue smoke?

As to the voltage control, I will need to look at this some more. If I understand your comment, I should look at the +6.2v reference that then passes through the voltage divider composed of R12 and R10, then to -S. Now, if I let the middle of the voltage divider be S, then I have S=vin*(R10/(R12+R10)), where vin = S + 6.2V. So, I can solve for S and get a value. This value matches the figure of 1 volt/200 ohms across R10.

However, this feels a bit like magic because it's not clear where the voltage S comes from. In other words, if I were to draw a schematic of this system, it would contain a voltage source of unknown value +S connected to the middle of the voltage divider and the minus terminal of a voltage source of value +6.4. Then, the desired voltage +S is sort of 'wished' into existence so that the equations are satisfied. To put it another way, this circuit requires a seemingly magical battery that knows the exact voltage +S that it needs to be at that moment and then sets its voltage to +S accordingly.

Thanks again!