Burned Capacitor Identification

[johnmx]

Hi,
I'm fixing a power supply of a very expensive lab equipment. One of the burned components is the 33nF capacitor shown in the attached pictures.
I think it is a film capacitor. I don't know what is the voltage rating, but since it is on the +360Vdc rail (after the PFC) it should be rated at least for 400V.
The body has a width of 6.3mm.
Although the capacitor has a hole in it, the multimeter still reads 33nF. The 10 Ohm / 1 W resistor that is connected to this cap is completely fried. Together they form an RC filter on the input of the auxiliar +5V DC/DC converter.
My guess is that the capacitor went short-circuit, the resistor fried with the 360V and later the capacitor healed itself.

The markings on the front side are:
333 -> 33nF
C7K -> C could be the tolerance and 7K the voltage?? 7K does not follow the general marking system.

Back side:
427
(some symbol) CZ

My questions are:
- Is it really a film capacitor?
- What is the rating voltage?
- What is the brand?

Thanks.

[james_s]

It looks like a ceramic capacitor to me, but if you really want to find out you could crack it open since it's already bad.

Doesn't matter what brand or voltage rating it is, you have plenty of information to select a suitable replacement. 400V is probably ok but it wouldn't hurt to go with 600V if you can find something that physically fits.

[coromonadalix]

the 333 value mean 0.033 uf

http://www.robotoid.com/appnotes/electronics-capacitor-markings.html

[Yansi]

C is not tolerance.  C7 is probably voltage or other manufacturer ID, K is very likely the tolerance (K is +-10%, which seems plausible for a ceramic cap).

[johnmx]

I crack opened the capacitor and now I can see the ceramic interior.

[Addicted2AnalogTek]

I can't find anything on C7K manufacturer capacitor code, but like Yansi said:

If the manufacture code is the same as the ones on new capacitors that can be had from supply houses:

C may be manufacture code for case/package style
7 may be reference to the voltage rating
K is likely tolerance, specifying +/- 10%

Just match the lead spacing spec for the replacement MLCC and select a .033uf (33nf)

My guess is a 630V X7R 33nf MLCC

[johnmx]

My guess is a 630V X7R 33nf MLCC

I'm going to order one of that.
Thanks.

[Yansi]

I can't find anything on C7K manufacturer capacitor code, but like Yansi said:

If the manufacture code is the same as the ones on new capacitors that can be had from supply houses:

C may be manufacture code for case/package style
7 may be reference to the voltage rating
K is likely tolerance, specifying +/- 10%

Just match the lead spacing spec for the replacement MLCC and select a .033uf (33nf)

My guess is a 630V X7R 33nf MLCC

no way in hell it'd be 630V.   

63V maybe, 100V maybe, 250V at utmost craze - would make sense for this size of cap. 

[james_s]

He said it's on the 360V rail, so I'd certainly hope it's rated for more than 250V.

[Gyro]

Personally, given that it's a simple RC snubber network, If there is space I think I would ditch the ceramic cap and get a decent 33nF 630V MKP Polypropylene film cap (physically larger) and new 2W 1W resistor. Then I would solder them between the outer resistor and capacitor PCB holes, safely air-wiring the junction between the two to allow it to fit.

If the cap has failed once, it could be a potential future re-offender.


P.S. A photo of the PCB location would help.

[floobydust]

I recognize it as an AVX ceramic cap like SkyCap SL21.
333= 0.033uF
C7K= X7R, 500V, +/-10%
427= date code
ACZ= AVX logo, lot code CZ

[Gyro]

Re-reading the OP, I'm a bit suspicious that the cap still read 33nF after removal. Multilayer ceramics are not really known for their self-healing capabilities. There's just a chance that the snubber has been asked to handle more HF than it was designed for.

It may be nothing, but it just seems a bit strange that the cap would go short for enough time to fry the resistor and then test normally afterwards.

[johnmx]

Re-reading the OP, I'm a bit suspicious that the cap still read 33nF after removal. Multilayer ceramics are not really known for their self-healing capabilities.

That's why I thought it was a film capacitor. The black spot shown on the front image is a hole!! I measured the capacitor with a Fluke multimeter after removing it from the circuit.

The circuit forms an RC filter, not a snubber.
I think there is nothing else burned, except the PCB track between the main supply (360V) and the resistor.

There is another capacitor in parallel with the damaged one, and it is exactly the same value/model.

I recognize it as an AVX ceramic cap like SkyCap SL21.
333= 0.033uF
C7K= X7R, 500V, +/-10%
427= date code
ACZ= AVX logo, lot code CZ

You found exactly the manufacturer and model. Perfect! Even the date code makes sense. The equipment is from 2006 and the capacitor date is 2004.

[johnmx]

If there is space I think I would ditch the ceramic cap and get a decent 33nF 630V MKP Polypropylene film cap (physically larger) and new 2W 1W resistor. Then I would solder them between the outer resistor and capacitor PCB holes, safely air-wiring the junction between the two to allow it to fit.

If the cap has failed once, it could be a potential future re-offender.

P.S. A photo of the PCB location would help.

The PSU is very dense with almost no free space inside. It is a C&D Technologies 5000 Modular series (I could not find the data sheet).

[lowimpedance]

There is another capacitor in parallel with the damaged one, and it is exactly the same value/model.

Make sure you replace it too.