ESR of ceramic capacitor in this snubber

Question

I am replacing a few fried components on this main board of a washing and drying machine that is rated for 127V and was powered on 380V.

It is based around a STR-W6052S.

Afer removing the resin and measuring its input stage components, only the STR-W6052S itself and the R_OCP were damaged:

R_OCP is the destroyed 0.33R non-inductive shunt. C5 (damper snubber, see datasheet image below) is the blue ceramic disk capacitor, that also looks damaged.

It looks like those SMD resistors were damaged, but they are fine.

Here is an example of the circuit from the datasheet:

The real circuit has some minor differences, as shown below. The rest of the circuit is not shown for clarity. Note the 2x 47R after C5 and the choice of using a zener in parallel with the C6 snubber capacitor. I assume the 2x 47R were added to tame oscillations.

The issue: I removed C5 and measured it using a DE-5000 LCR (and ESR) meter. At 100kHz, it measures 96.3pF and an ESR or around 50R. I say around because I don't get a stable reading. It starts at 52R and slowly drifts down, reaching 33R over 10 minutes, and the reading still doesn't stabilize itself. I assume this is caused because this is capacitor probably has a Y5P dielectric (or worse), so the ESR changes when a voltage is applied. As a comparison, no other capacitor I have measured with this instrument shows this wide fluctuation. Other quality Y5P capacitors ESR might drift a bit, but a few ohms at most. I don't know the manufacturer or brand of this part, but there are many components on this board manufactured by the Taiwan based company Walsin. Assuming they manufactured this part, then this datasheet applies:

At 1 MHz, Q = 1000 and C = 100pF gives ESR = X_C / Q = 1.59 ohm. Unfortunately, the DE-5000 is limited at 100 kHz and thus I can't test at 1 MHz.

Does this capacitor need a replacement or is a drift like this considered normal?

The replacement: I am out of the USA and I just don't have access to any major distributor, so I can't order a proper new 100pF quality capacitor. I do have, however, a lot of 220 pF / 2 kV, Y5P, Vishay S221K25Y5PP63K5R. Handpicking 2 of them in series yields 100pF, and a lot of voltage leeway. What is troubling me is that the ESR at 100 kHz of the 2x 220pF in series is 284 ohms. Mind that the soldering is not to blame, since a single part measurement gives 142 ohms also at 100 kHz.

Does that ESR seem right? Can I replace the other capacitor with these in series?

Answer 1

The meter datasheet, https://www.ietlabs.com/pdf/Manuals/DE_5000_im.pdf gives accuracy on the order of 2%; see pp.39-41.

This includes accuracy of measurements and parameters, such as damping factor. The reactance in your measurement is 15.9kΩ, 2% of which is 318Ω. 284Ω is less than this, implying the true value is below measurement error. Note it could just as well be negative, and the reading would still be accurate given the specification.

In short, you need a more precise meter to measure the Q.

Or, some way to cancel out the capacitive reactance so it can measure the remaining impedance more accurately -- trouble is, the only way to do this is an inductor, and high-Q inductors are particularly hard to construct. Or you have to calibrate the inductor's Q, then subtract it from the measurement, which begs the question, what infinite-Q capacitor do you calibrate it with; etc. Suffice it to say, capacitors of this high Q are challenging to measure!

Put another way, the only hypothesis you can justify with this measurement, is the Q is greater than 50; how much greater, you can't really know.