HP 6200B power supply blowing fuses

[Shock]

With replacing the caps repair first.  A minimal replacement job would be if you had a few of the same series or brand caps fail showing a trend.
If a lot of them were contributors, or on their way out and economical to replace, and the device was going to be a keeper then replace what you can.  Same goes for transistors and also checking resistors for drift etc. It's a reliability, accuracy, cost, time balance.

[sync]

CR8, CR9 are not zener diodes. They are basically a string of three diodes to get a forward voltage drop of 2.4V. If you need to replace one then use three silicon diodes (1N4148) in series.

A zener diode has it's specified voltage on reverse bias and conducts (0.7V) on forward bias. Big difference.

[mos6502]

When you're checking caps, ESR by itself is actually pretty meaningless, because an ESR value that would be fine for one type of cap can mean that it's defective for another type of cap. A much better parameter is the dissipation factor, D, which the DE 5000 displays by default in the top display in auto mode. So, just set the DE 5000 to 120 Hz (for electrolytics), and check that the D is less than 0.25 (0.15 for low ESR caps, but you don't have those in your PSU). If the capacitance is also OK, the cap is good.

[oldway]

Are you using a slow blow fuse ?

[GreyWoolfe]

mos6502, thanks for the tip.  I will go back and check.  Sync, the fuses are fast blow.  I found a bunch of 1N4148s so I am going to change out CR8, 9 and 10.  I will be hitting up home depot tonight to get the parts for a dim bulb tester and I will update over the weekend what I find.

You guys are great, pointing a second time around newbie in the right direction.  If I get it working, I will be sure to post pics of the guts, the crappy bezel and what it looks like after cleaning.  This will go great with my Agilent 3610A, Elenco XP-720K and Astron VS-35M power supplies on my bench.

[Andy Watson]

... the fuses are fast blow.

Unless you've gone way over the top with the rating, the initial inrush current will probably be sufficient to take-out a fast blow fuse, fault or no fault.

[Shock]

... the fuses are fast blow.

Unless you've gone way over the top with the rating, the initial inrush current will probably be sufficient to take-out a fast blow fuse, fault or no fault.

On the schematic it's not specifically marked slow blow only 2A how did you determine it will fail regardless?

[Andy Watson]

... the fuses are fast blow.

Unless you've gone way over the top with the rating, the initial inrush current will probably be sufficient to take-out a fast blow fuse, fault or no fault.

On the schematic it's not specifically marked slow blow only 2A how did you determine it will fail regardless?

Experience.
And note, I only said probably fail. But, the chances are that Mr HP made the fuse specification as low as possible, consistent with it not blowing every time the unit is powered-up. Using a fast-blow fuse in place of a normal/slow blow fuse allows less energy into the circuit; not what you need when there's a transformer to magnetise and smoothing capacitors to charge up.

[Wytnucls]

Agree with Andy. I would replace that fuse with a medium or slow blow one, 'tout de suite'.

[Po6ept]

The HP fuse number cross references to Littelfuse, Newark, and Boeing numbers that are specified as fast-acting.  This isn't to say that a slo-blo wouldn't be useful to see if inrush current is the culprit...

[tautech]

Not much point if you use a Bulb tester.
Bulb will indicate the inrush and Cap charging, then dim if there is no load and all is well.

[Shock]

Not much point if you use a Bulb tester.
Bulb will indicate the inrush and Cap charging, then dim if there is no load and all is well.

You mean not much point using a slow blow or a dim bulb tester? Unless I'm not mistaken the issue present is a short causing somewhere above 2A blowing the fuse.

As far as I'm aware the best approach is either to get the DUT to power up in a current or voltage limited state.
Or depowered and cap drained state, test and replace caps then move onto checking the transistors with a possible sacrificial fuse to check in between.
Or depowered and cap drained state, isolating parts of the circuit testing resistances.

Is there any better way to approach this?

[tautech]

Not much point if you use a Bulb tester.
Bulb will indicate the inrush and Cap charging, then dim if there is no load and all is well.

You mean not much point using a slow blow or a dim bulb tester? Unless I'm not mistaken the issue present is a short causing somewhere above 2A blowing the fuse.

As far as I'm aware the best approach is either to get the DUT to power up in a current or voltage limited state.
Or depowered and cap drained state, test and replace caps then move onto checking the transistors with a possible sacrificial fuse to check in between.
Or depowered and cap drained state, isolating parts of the circuit testing resistances.

Is there any better way to approach this?

Bulb tester will power DUT in a Voltage and Current limited state and keep the magic smoke in!
Then use your knowledge to fault find the areas of the DUT where low/no voltage is present. Powered up or otherwise.
Poor mans Variac !

[megajocke]

OK, brain cramp on CR8 & 9.  They are not 1N485B diodes but GD8-001 silicon diodes rated at 2.4v @ 100mA.  Googling finds nothing.  In looking for for something similar, I found 1N4617-2.4v @ 200mA.  Could this be a reasonable substitution?  Assuming that CR8 and 9 are bad.  With readings over 1.7, I would think that they would warrant a replacement.  Someone correct me if I am wrong.  Don.r, maybe I really am a dim bulb tester!

Your earlier measurements suggest CR8 and CR9 are perfectly fine. They wouldn't cause the fuse to blow if broken in any case.

My prime suspects would be rectifier diodes, filter capacitors, CR11, CR34, Q4, Q6 and Q7.

Removing Q6,Q7 and Q4 and testing with a dim bulb tester sounds like a way to go, but it might be easier to start by disconnecting the CR26-29 rectifier and use the dim bulb tester to identify whether the problem lies in the regulator circuit or the power circuit.

There was a mention of CR10 earlier, but I can't find it on the schematic.

[lowimpedance]

Your earlier measurements suggest CR8 and CR9 are perfectly fine. They wouldn't cause the fuse to blow if broken in any case.

My prime suspects would be rectifier diodes, filter capacitors, CR11, CR34, Q4, Q6 and Q7.

Removing Q6,Q7 and Q4 and testing with a dim bulb tester sounds like a way to go, but it might be easier to start by disconnecting the CR26-29 rectifier and use the dim bulb tester to identify whether the problem lies in the regulator circuit or the power circuit.

There was a mention of CR10 earlier, but I can't find it on the schematic.

As I said earlier with CR8,9 not to assume they were suspect !! so move on and check CR11 (((((NOT 10, NOT 10 !! ))))) , and having looked again at the circuit CR34 would be good to check too.
 Q4 6, and 7 were supposed to be checked as good! so testing with the bulb tester and Q6 and 7 out of circuit there is not much left, but would be important to eliminate the rectifier and filter section before moving on to other possible areas.

[GreyWoolfe]

Lowimpedence, I replaced CR 8,9 and 11.  I powered the PSU without Q6 and & with the dim bulb tester and the PSU came on.  I checked CR34 out of circuit and it appears fine.  My Fluke 27/FM in diode test shows .585 and open in reverse.  With the tester, I am showing 862 mV.  I put Q6 and 7 back in and the PSU came on with the tester.  I will probe the circuit around CR34 and come back with findings.

[pomonabill221]

I am watching your progress with GREAT anticipation!  Looks like you are making progress!
Keep us updated (as I am sure you will)!

[Po6ept]

You're making good progress.  Assuming the supply is jumpered as shown in the schematic, what happens if you remove the sense jumpers -R to -S and +R to +S?

As lowimpedence suggested, if you look at this as three separate supplies, what happens if you open taps 7 and 10 on T1 one at a time to isolate the problem to one of the three?

[TerraHertz]

First off, put something like a 40W light bulb in series with the mains input. This will prevent you from blowing any more fuses (or semiconductors), and also serves as a visual indicator. If it glows brightly, the PSU is still broken.

 
All the years I've been servicing stuff, and that never occurred to me. Doh. Thanks.
I generally just used a variac, and an AC ammeter, and wound the output up slowly while watching the current.

Too bad lightbulbs are soon to be completely unobtainable. Better lay in some more stock.

[GreyWoolfe]

Here  is another update.  MOS6502, I went back and checked the dissipation factor on C12, 13, 14 and 20.  The caps are still in circuit and the D was well over .25-- .7 and up.  Do I need to check the D with the caps out of circuit?  They are a bear to try to unsolder as the mounting tabs are also soldered.  I just got some wider solder wick so I can try if necessary.  PO6EPT, I opened taps 7 and 10, the PSU powered up with the tester and blew the fuse plugged into the mains.  I am now officially out of fuses.  I refuse to pay Radio Shack ripoff prices so I ordered in 20 fuses of eBay, they will be here by Tuesday.  I will go and get a pack of 2A slow blow to see if the inrush current is a problem.  More to follow.  Once again, thanks for all the assistance.  If nothing else, I will have some time to try to understand the schematic better.

[pomonabill221]

Would be nice if you could find a circuit breaker to use rather than eat fuses?  A fast trip 2 amp Heinman from a scrap piece of electronic gear?  Even a 3 amp you could use but that is pushing it a little.  NOT a thermal breaker but a mag trip.

I agree with buying ratshack stuff... they are $$$$!
Aren't you using the light bulb for current limiting?  If you are using a 40 watt lamp, the fuse should NEVER blow.

[Po6ept]

A guy I used to work with was of the opinion that the best way to fix shorts was to jumper over the fuse and deal with whatever smoked.      (Don't do this, of course.)  After a few burned circuit traces and transformers, he was invited to go work elsewhere...

I realize this is a bit silly to ask, but have you verified that all the jumpers are where they should be?  I ask because I have spent hours working on equipment only to find that the person before me did something that deserves a face-palm. 

With the T1 taps open at 7 and 10, there really isn't much there that should blow the fuse.  CR26-CR29, C16A, C13, C14, Q6, Q7, C20 are the things that immediately jump onto the must-check list. 

With the supply off and the caps discharged, what is the DC resistance between TP24 and the negative rail? How about at TP23?   Same thing at +R with the jumper to +S on and off.

You can sometimes unsolder one lead of a filter cap well enough to center it in the via to take readings.  That gets around having to unsolder the tabs. 

[lowimpedance]

 So with tap7 and 10 disconnected and thus only leaving the main supply windings  the fuse still gives up the ghost!. Is Q6 and 7 still out of circuit?.
If so there is very little left and if you have already replaced the rectifiers that just leaves the caps (or maybe the use of a slow blow fuse is required).
 AS Po6ept said try and desolder the cap pins and give them a circular wiggle in there holes to make sure there is no 'little' whiskers of solder still connected from pad to pin!. Check with you DMM that the pin is indeed now isolated from the pad. Then try powering up with the 40W bulb tester and again test for an appropriate secondary voltage - if the new fuse remains intact. If all looks good try again without the bulb tester, does it look like previous measurement or does the fuse pop. (lets hope at this point it does not!) If the fast blow pops try the slow blow to see if it survives.
 If the fuses survive try bodging in another modern cap of approximately the same value and try with bulb tester that you now get an appropriate DC across the bodged in cap. If it looks good then try putting Q6 and 7 back in circuit and so on with taps 7 and 10. The upshot being hopefully the original filter caps have to go.
 Just a thought perhaps the Radio shack fuses were dodgy/mislabeled etc. With the ebay fuses it would be interesting if they pop the same way.

On a side note I own a 6236B triple supply and it uses a fast blow 1A (240V mains!), just checked!.

[pomonabill221]

To keep you inventory of fuses, why aren't you using the 40 watt lightblulb?
IF the fuses are blowing WITH the 40 watt lightbulb, something is wrong with your wiring because a 40 watt lightblub will NOT blow a 2 amp fuse!!!

[pomonabill221]

Just another thought.... I know you are in the US and line voltage is 110 (I assume that is what you are using to power the supply).
Is the supply strapped for 110 and is it done correctly?
May seem like a stupid question... but stranger things have been done and overlooked!