X-class capacitor quality

[japlytic]

When I did a check for capacitance tolerance on about 200 used X-class capacitors, up to half of them were lower than 90% of their rated value (all had a 10% tolerance), with a few being way down in capacitance.
Apparently, half of the yellow X-class units were out of tolerance.
I remember reading the Serviceman's Log in Silicon Chip magazine about X-class capacitors losing their capacitance because too much capacitance was crammed into a small unit, and apparently deteriorated at a much quicker rate.

Could some brands of X-class capacitors be bad?

[b700029]

Re: X-class capacitor quality

No one really notices even if they're bad (unless they fail shorted, which they are specifically designed not to do) because they're basically only for EMI compliance, and really cheap PSUs don't even have them.

[999999999]

Re: X-class capacitor quality

It wouldn't surprise me at all if some brands used lower purity aluminum, lower uniformity in depositing it, or lower quality plastic.

Not sure which mechanism they are referring to by "too much capacitance was cramming into a small unit", that should just result in them having a lower voltage limit or margin but at capacitor tester voltage levels it should not matter. However I'm thinking of a new capacitor, lower voltage tolerance could cause more AC mains voltage spike damage in use. You could tear open one of the worst to see what's inside...

[PeteS in CA]

Re: X-class capacitor quality

Assuming these are film dielectric ...

It would not surprise me if cheap-@$$ brands set their machines to make parts at the low end of their tolerance, and then cheat just a little. Or made caps to the low end of 20% tolerance ("M" instead of "K") but marked them 10%. Like b7... posted, who would notice?

OTOH, the parts you checked were all used. Some one pointed out on another thread that film X and Y caps take voltage spikes, are damaged, but self-heal (the arc that punches through the film clears metallization from the area around the arc site, so the cap is not shorted). Each self-heal does damage, and over time the damage is cumulative and can reduce the capacitance of the part. Also, the self-healing of film capacitors is not utterly perfect. There is carbon residue in and around the arc, through which current can flow. Over time these high resistance current paths accumulate and can cause the part to got hot and even catch fire.

If you live in an area with lots of spikes due to industrial machinery or lightning storms, that can be really hard on X caps. OTOH, if your power is relatively pristine ...

Now, if those X caps were ceramic dielectric, they don't have the vulnerabilities film caps do. BUT getting higher capacitances at the required voltage and surge voltage ratings necessitates using ceramics with horrible temperature coefficients (e.g. Z5U or Y5V). If you connect such a part to your capacitance meter and warm the part with your fingers, you can watch the capacitance reading vary. At the high or low end of such parts' rated temperature range their capacitance could le 25% or 30% (or less!) of the nominal value. Such ceramics also have horrible DC voltage coefficients, but that doesn't apply in this context. Anyway, if those were ceramic X caps, they might have been good, but your room temp (or finger temp) caused to capacitance to drop that much.

When Electronics classes in high school, trade schools or colleges say, "Assume an ideal capacitor," that's fine for calculations, but they really should spend some time talking about and demonstrating the effects of real-world characteristics such as impedance and temperature & voltage coefficients!

[Tobugrynbak]

Re: X-class capacitor quality

I'm going through a similar exercise hand grading new x2 capacitors. Found most to be low but within 10%, just a couple at just above the value and a couple lower than 10%.

A previous batch has caused us some grief, prematurely failing in the field. Too much "self healing" causing them to go way out of spec (from 0.22uF to 0.022uF in one case).

As for dodgy components I'm always on the lookout for counterfeit parts though that usually just concerns semiconductors. Though I've got a bag of capacitors with brand x on the bag but brand Y on the capacitors themselves. With a lot of sub sub sub contracting going on in the manufacture of parts with everyone chasing the lowes possible price, it's gonna happen again and again.

[PeteS in CA]

Re: X-class capacitor quality

There are some parts rated for both "X" and "Y" usage. Though a .01 uF or 4700pF "X" cap seems less that useful to me.