Bench power supply failure 4

[ScottyUK]

Yesterday my beloved HP 6675A power supply died. It hadn't been used for a while - it's a great big lump for most jobs - and as far as I can determine so far at least one of the reservoir caps on the front end has failed through aging and lack of use. I should have thought about this possibility before plugging it in, but it's easy being wise afterwards. It hasn't failed dead short, but seems to be an intermittent short which is severe enough to clear the 13A line fuse and then blow clear. Is this a normal failure mode for an electrolytic? I'm aware that this happens with self-healing film types but hadn't seen it on big can types.

On to the second problem which is that the caps are an old Philips line rated at 400V / 1500uF and are about 46mm diameter. Nowhere can I find caps matching the both physical and electrical parameters. The diameter is important. I've tried Epcos, Vishay (Sprague), Evox-Rifa (Kemet), BHC (KEMET), Siemens. Any more ideas? I think this is likely a US part, and I'm largely looking at European sizes. I'd really like to avoid having to do a hatchet job on a supply worth over £2000 so I'm hoping for direct replacements rather than a homebrewed repair.

Alternatively does anyone know of an equipment breakers who sell parts for high end test gear?


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Skogsgurra]

These guys impressed me:

http://www.stealthcomponents.com

I have been talking to Chris Bryant. They are in North Carolina.

Good luck



Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Comcokid]

OK, took me only about 10 minutes.

The service manual with parts list and schematics can be found on the internet at agilent.com in their service area as a PDF.

The only 1500uf/400V cap I found listed in the manual is part number 0180-4369.

Searching this part from Agilent indicated the part is orderable, they have 18 on hand, and is about $41 (US dollars).

So, there you go.

 

[ScottyUK]

Thanks ComcoKid - I have the service manual and couldn't see the 1500uF cap. What page is it on (just so I can embarrass myself!)

$41 isn't too bad considering it's from the OEM.

Much appreciated both of you.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[OperaHouse]

That new cap may be as old as the one you take out. This is general info for others in this situation. I have a graveyard of old caps and take a little time to reform them before installing. Nothing overboard, just charging for a couple hours through a high ohm resistor so natural leakage slowly ramps up the voltage. There is a real change. The VFD guys talk about caps blowing up when a drive is stored for a couple years and then put into immediate use. Many of the manuals have a warning to power the drive up for 24 hours before motor use after long storage. High switching currents of a power supply would be similar to a VFD. A little caution may pay off in extra component life.

 

[ScottyUK]

How high a voltage is required for reforming? Does it have to be full rated voltage (through a limiting resistor as you say) or does reforming take place to an adequate level at a lower voltage? I've usually chucked suspect caps out rather than tinker with them. Would appreciate some practical experience.



----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Skogsgurra]

This is what Siemens says.

I would use a series resistor also. Or an incandescent lamp.



Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[OperaHouse]

After posting I became a little curious about the procedure. I have a Sencore LC101, a little hokie as a capacitor tester though love the leakage feature for semiconductors and capacitors. It has this spring plunger that Keeps the leakage button pressed in for forming capacitors. I remember downloading a manual for it off the internet a while ago but couldn't find it on the computer or web again or their procedure. Found a good quality new 40uF 250V that was at least 12 years old. These are the results using the LC101 to reform.

Starting off at 50V the leakage was 1200uA fifteen minutes later it was 20uA.

Switching to 100V the leakage was 2200uA fifteen minutes later it was 0uA.

Switching to 200V the leakage was 2900uA thirty minutes later it was 0uA.

Initial current readings were made after the capacitor had reached full voltage for the setting. During forming the current decay was gradual. There were numerous wild swings of current.

The old radio restorers put a variac on the line power and raise the voltage over an 8 hour period. A line cord with a diode and 47K to 100K resistor in series would probably be good enough. Voltage doubler if more voltage is needed. From the data, half the full rated voltage did not fully form the capacitor. It would likely form it enough for it not to be damaged by application of full voltage.

 

[OperaHouse]

Lots of info out there. One site

http://www.vmars.org.uk/capacitor_reforming.htm

Basically charges form a 400V source through a 68K ohm resistor. When voltage drops below 75V across resistor it if formed.

Othr sites say to continue forming till this voltage drops to almost zero. Leave overnight

ABB says reform an hour for every year it was stored

They all say the same thing. Limit current to less than a couple ma and wait for voltage to approach source value.

 

[Comcokid]

ScottyUK-

In the Agilent service manual dated 2000-09-01, The capacitor appears on page 87 (the thrid page of Section 5 "Replacable Parts"), in Table 5-3 "Main Chassis, Replacable Parts" as C416-C419.

Whatever Agilent has in stock is probably not old - this particular power supply model is still available brand new from Agilent and their distributors.