Hello everyone!

Recently we had an issue with a electronic board production that some capacitor exploted and destroy the PCB. There were PCB working with no problem and others may failed in 2 weeks, 1 month or even several months. Well, after some months of confusion and trying to solve the problem searching in the design, we found that our electronic assembler used a 2.2uF/25V instead of 2.2uF/100V. Thus, this capacitor is connected to a 65V and there are some other capacitors in parallel with the right voltage rate.
So, my question could be:
1.How is the effect of using a undervoltage rated capacitor with right rated voltage capacitor in parallel??
2.Is there an effect of having the Cs1 and Cin2 and Cin3 very closed??
I'm attaching some details about schematic and capacitor's position in the PCB with effect of its explosion.


Csn1: 33nF/100V - Cin2: 2.2uF/25V - Cin3: 2.2uF/25V
There 3 branches of an ACIM inverter with the same battery package.
I'm curious understanding it because I never could not reproduce the issue in my lab and I would like to know why this happened in some other conditions.

Thank you.
Best regards,
daniel

 

Welcome to AAC!

How is the effect of using a undervoltage rated capacitor with right rated voltage capacitor in parallel?

The three capacitance values add together to give a combined capacitance of 4.433 uF, but the 25V value limits the overall rating to 25V. Hence the magic smoke if the combination sees 65V.

Is there an effect of having the Cs1 and Cin2 and Cin3 very closed?

If you mean 'close together', then the explosion (or overheating) of either 25V rated cap could damage the other cap.

 

I would imagine that the correctly rated voltage capacitor wouldn't have any issue being mixed with the underrated ones. The underrated ones would be under severe pressure, rated for 25V and getting 65V instead. Depending on your circuit, load and current, some caps may fail sooner than later especially if there are large voltage switching surges going through them. Your bit of board looks like a relatively high current section.
I don't see any issues with the caps being close to each other.

 

I guessed that the 25V voltage rated capacitor should have explose when 65V is applied, but this is my confussion, they don't explose inmediately... there are some boards working in installation from months.
Well, the assembler told me that the rated voltage was 25V instead of 100V... but they can be wrong again and use another rated voltage closed to 65V... let me said 50V. So the effect of explosion should be not inmediately but soon infulenced by temperature and 1,5 times margin of capacitor rated voltage capacitor: 75V.
The only way is try to apply voltage to the isolated capacitor and see what voltage is applied it explose.
I guess the capacitor rated voltage is not 25V... but not 100V either.

 

Also remember, they don't need to necessarily explode, they can just crack internally and therefore become open circuit. And if the crack is internal or not visible it would look like they are still working.

 

Hola Daniel

Out of curiosity, could you show some pictures of the damaged PCBs?

 

I guessed that the 25V voltage rated capacitor should have explose when 65V is applied, but this is my confussion, they don't explose inmediately

The 25V rating is not a hard value.
Due to manufacturing variation, there likely is a large range of voltages at which the leakage becomes large enough to overheat the capacitor and cause it to explode.

 

Also remember, they don't need to necessarily explode, they can just crack internally and therefore become open circuit. And if the crack is internal or not visible it would look like they are still working.

Ok, ok... this can give me new clue to follow, because if this capacitor breaks down, its snubber function is disabled and it can breaks other components like the MOSFET.

 

Hola Daniel

Out of curiosity, could you show some pictures of the damaged PCBs?

Picture above, left to Cin3.

 

If the circuit needs all of the total capacitance to function correctly, and then some of that capacitance is lost because of over-voltage failure, then the remaining capacitors will be stressed quite a bit and fail eventually. So the damage to the correctly rated devices is from a different cause, and not as likely to be instant.
Good engineering would avoid having parts with different specifications like that all looking the same. But even the best of engineering can be undone by purchasing fools.