Capacitor failure

TANGHONGXI

Capacitor failure [Copy link]

 

Recently, I have a problem to analyze. It is about the failure of the capacitor and the short circuit that caused the fuse to burn out. The circuit is DC-DC 24V to 5V. The failed capacitor is at the input end, 2 22uf/35V 0805 package ceramic capacitors. According to the user, there is no phenomenon when the air switch is turned on. We finally judged that the capacitor failed, but according to their statement, the problem cannot be reproduced. The same situation has occurred several times. I would like to ask if you have encountered such a situation and what is the general cause.

maychang

When a ceramic capacitor shorts, it is often a quality problem of the capacitor itself, rather than a circuit design error. The most annoying thing is that the capacitor is fine when tested separately before welding, and it is also fine when tested after welding, but it becomes short-circuited after working for a period of time (maybe as long as several months).

TANGHONGXI

But when I told the boss that it was the quality of the capacitor that was problematic, he was not convinced.

TANGHONGXI

maychang posted on 2018-10-25 10:36 When a ceramic capacitor short-circuits, it is often a quality problem of the capacitor itself, rather than a mistake in the circuit design. The most annoying thing is to test it separately before welding...

What you said is the same as what I am experiencing now. I have tested the capacitor separately and the voltage has reached 40V. There is no problem. I also tested it on the board and there is no problem.

爷乐子

chunyang

To make a specific judgment, you need to get the faulty device and test it. It is impossible to make a judgment based on the limited information provided by the original poster.

ienglgge

TANGHONGXI posted on 2018-10-25 10:39 But when I told the boss that the capacitors had quality issues, he was not convinced

You can do a comparative experiment. Buy a batch of high-quality capacitors. Build a good test environment. Get more test boards. Work for a long time. Let time prove everything.

PowerAnts

22uF, 35V, 0805?? It's so fake, either the capacity is twice as much or the withstand voltage is twice as much

qwqwqw2088

In this case, it is possible that the ceramic capacitor is fake, but it needs to be tested and analyzed carefully. Replace it with a better capacitor to make a judgment.

topwon

There is no phenomenon when turning on the circuit breaker board------What does the circuit breaker control here? 220V AC input? Does it mean there is an AC to 24V DC power supply at the front end? Is its output normal? Does the 24V have large ripples or high voltage at the moment of power-on? What is the withstand voltage of the chip? Is the chip also damaged?

topwon

TANGHONGXI posted on 2018-10-25 10:40 What you said is the same as what I am experiencing now. I have tested the capacitor voltage up to 40V alone and there is no problem. There is no problem when testing it on the board

Then you may need to pay attention to the test method and environment. For example, is the 24V input the same as that on site (the DC power supply in the laboratory and the on-site power supply with large ripple may have different impacts on the capacitor)? Are the ambient temperature and humidity consistent? Is the site harsh? Can the design, manufacturing process, and installation method of the PCBA prevent the capacitor from being easily affected by stress and causing deformation and rupture?

topwon

PowerAnts posted on 2018-10-26 08:09 22uF,35V,0805?? It's so fake. Either the capacity is 2 times fake or the withstand voltage is 2 times fake

For example, TDK has capacitors with such parameters. I just don't know if the ones the OP bought have false labels. C2012X5R1V226M125AC Capacitor 22 uF Voltage rating 35 V Tolerance 20 % Temperature coefficient/code X5R Case code - in 0805 Case code - mm 2012 Operating temperature range - 55 C to + 85 C Strange, I don't see this material on TDK's official website, is it a fake or the website is not updated? This needs to be confirmed with the seller.

TANGHONGXI

topwon posted on 2018-10-26 09:35 For example, TDK has capacitors with this kind of parameters, but I don’t know if the one you bought is falsely labeled. C2012X5R1V226M125AC capacitor 22 u ...

When we purchase from SMT, we directly ask them to provide us with materials. Only they know whether they are real or not.

TANGHONGXI

TANGHONGXI posted on 2018-10-25 10:40 What you said is the same as what I am experiencing now. I am testing the capacitor voltage alone and it reaches 40V without any problem. There is no problem when testing on the board either

The test method is the same, but the environment is not good. I let the PCB work for two days and there is no phenomenon

TANGHONGXI

topwon posted on 2018-10-26 09:19 There is no phenomenon when I turn on the circuit breaker board------What does the circuit breaker control here? 220V AC input? That means there is an AC to 24V DC converter at the front end...

The battery circuit breaker here controls the input voltage of the battery. The ripple is very small and can be basically ignored. The voltage at the moment of power-on has been captured by the oscilloscope and does not exceed 25V. The chip is not broken, only the capacitor is broken.

topwon

TANGHONGXI posted on 2018-10-26 10:36 The test method is the same, but the environment is different. I let the PCB work for two days and there was no phenomenon

You can find out what the environment is like on site. How is it installed, open or closed. You can also ask for a sample of the 24V power supply to be sent over. The 35V 0805 22uF capacitor may really be falsely marked. Even if it is available at this price, it is not something that ordinary products can use. The 35V 10UF used in the lighting power supply we designed are all in 1210 package. When designing, we must leave a margin and try to use conventional materials. Otherwise, mass production will cause trouble for procurement and production line planning.

zhuyebb

What a rare and powerful post.

TANGHONGXI

topwon posted on 2018-10-26 10:42 You can find out what the on-site environment is like. How is it installed? Is it open or closed? You can also ask for a sample of the 24V power supply. 35V 0805 22 ...

There is a machine in the company's R&D. The ordinary power board used also had this problem. There is no experience in mass production. The boards are basically 0603 packages. The capacitors involved are a little larger.