[SOLVED] 0805 ceramic capacitor make the noise

[greatmarx]

0805 ceramic capacitor make the audio noise

I don't know the reason why?
Someone told me that it's the Piezoelectric effect.
The SRF is >1MHz
Who can give me some material about it.

 

[vfone]

Generally reactive components (ideally with no losses) doesn't introduce noise in the circuit.
There is a situation, for example at the input match of a high-frequency LNA replacing a 0402 cap with a same value 0805 one (due to different parasitics and SRF) the noise figure gamma optimum match could change, increasing the NF of the LNA.

 

[G0HZU]

I think it depends on the dielectric material used in the ceramic capacitor.
Eg X7R and Y5V and Z5U can produce piezo effects if the cap is subject to stress or vibration. I think NP0 class capacitors are fairly immune to the piezo effect.

For this reason it's best to avoid these caps in high gain audio amplifiers if the electronics will be subject to stress or vibration. eg onstage audio gear would be especially prone.

 

[biff44]

chip parts that are slightly damaged (like microcracks) can give off shot noise. Try replacing the part with a different manufacturer's part and see if problem goes away--you could have a bad lot. Tantalum capacitors can have some "sputtering" shot noise if there is a big voltage across them. And of course any electrolytic cap can have all sorts of noise, especially older ones.

Never found a ceramic cap that was microphonic or piezoelectric, though, and I have shaken them with many G's of force while watching!

 

[G0HZU]

Large, flat ceramic capacitors in X7R or Z5U are not recommended for use in a PLL loop filter if the VCO has a high Kvco. I believe this is due to the piezo effect of the capacitors when the circuit is knocked or vibrated. This microphony can introduce audio noise in the PLL output.

However, I would have thought that the VCO or the reference crystal oscillator would dominate any microphony effects.

 

[greatmarx]

Maybe I didn't expain the situation carefully.
The 22uf capacitor is used for RF PA. I can hear the sound only with the pcb board.There is no MIC or Earphone.Althrough the pcb board is designed for mobile phone.

 

[G0HZU]

You could try reading this IEEE article

“Electrical noise generated from the microphonic effect in capacitors” Nelson &Davidson

IEEE Xplore - Electrical noise generated from the microphonic effect in capacitors

You need IEEE membership to access it. I can access it if I apply for it but I don't have the free time to do this.

Edit: just saw your reply above. It looks like the capacitor is producing audible noise due to the piezo effect.

 

[greatmarx]

It seems like the IEEE article that Gohzu given.
But I can't read the article because I am not the membership of the IEEE.
Can you give any other article which can be help?
Thank everyone .
Yang

 

[biff44]

If that IEEE paper is the only evidence, I am going to cry B.S.!

1st, they had a HUGE 400 hz mechanical resonance on that board, which looked more like a bass drum head than a circuit board at resonance! Surprised the chip did not depart the board after the test! They probably had 100s of Gs of force vibrating the board, with mechanical harmonics and all. All they had to do was to add 20 more screws and posts, the resonance would have shifted up to 20 KHz, and the data might be believeable.

2nd, how do you know it is not noise pickup in their measuring setup. The board is on a HUGE electromagnet, powered by swept sine waves. It would be reasonable to assume that there was going to be a 1/2 nanoamp of induced loop current in their board from the vibration stand itself. Did they elevate their test board off of the vibration stand by an inch and re-run the test to see how much magnetic pickup they had? No!

I am unconvinced. Show me some corroborating data!

 

[keith1200rs]

This might be of interest:


Keith.

 

[G0HZU]

If that IEEE paper is the only evidence, I am going to cry B.S.!

1st, they had a HUGE 400 hz mechanical resonance on that board, which looked more like a bass drum head than a circuit board at resonance! Surprised the chip did not depart the board after the test! They probably had 100s of Gs of force vibrating the board, with mechanical harmonics and all. All they had to do was to add 20 more screws and posts, the resonance would have shifted up to 20 KHz, and the data might be believeable.

2nd, how do you know it is not noise pickup in their measuring setup. The board is on a HUGE electromagnet, powered by swept sine waves. It would be reasonable to assume that there was going to be a 1/2 nanoamp of induced loop current in their board from the vibration stand itself. Did they elevate their test board off of the vibration stand by an inch and re-run the test to see how much magnetic pickup they had? No!

I am unconvinced. Show me some corroborating data!

The piezo effect of high dielectric ceramic capacitors is well known and well documented so I'm slightly surprised why you have such an issue with it.
If you apply stress to a piezo material it generates potential differences within the material and this causes noise. That's the way piezo microphones work.
Also, it you apply a large switching voltage across a (piezo) susceptible ceramic capacitor then the reverse effect takes place. The capacitor starts to 'sing' because it acts like a mini piezo speaker.

---------- Post added at 19:08 ---------- Previous post was at 19:06 ----------

This might be of interest:

http://www.kemet.com/kemet/web/homepage/kfbk3.nsf/vaFeedbackFAQ/118EBDDDFA5D532C852572BF0046B776/$file/2007%20CARTS%20-%20Reduced%20Microphonics%20and%20Sound%20Emissions.pdf

Keith.

I tried a cut and paste but I couldn't get access to the doc. It showed as not found on the Kemet site?

 

[keith1200rs]

I have uploaded it. I am not sure why the link didn't work (although it looked a mess!).

Keith.

 

[greatmarx]

The piezo effect of high dielectric ceramic capacitors is well known and well documented so I'm slightly surprised why you have such an issue with it.
If you apply stress to a piezo material it generates potential differences within the material and this causes noise. That's the way piezo microphones work.
Also, it you apply a large switching voltage across a (piezo) susceptible ceramic capacitor then the reverse effect takes place. The capacitor starts to 'sing' because it acts like a mini piezo speaker.

---------- Post added at 19:08 ---------- Previous post was at 19:06 ----------



I tried a cut and paste but I couldn't get access to the doc. It showed as not found on the Kemet site?

The doc keith given is very good.
That the ceramic capacitor make the noise is not a rare case.
The 22uf capacitor is working on the vbat pin of a GSM PA which is working on the voltage of up to 4.2V.
Yang...

 

[smislahi]

Re: 0805 ceramic capacitor make the audio noise

Check the ESR of that capacitor. High ESR results noise.

 

[FvM]

The question apparently hasn't been clear. Mentioning a GSM PA clarifies, that you are referring to audible noise generated by the pulsed GSM supply current. This is actually a case well covered by the Kemet publication. On the other hand, it's obviously not involving any problems for the GSM operation. And it doesn't change the fact, that high Er ceramic capacitors are superiour for high frequency supply filtering, surely better than tantal caps. So if audible noise is a problem in your case and can't be managed by changing the board geometry, you probably should go for the said COLF MLCC types.

 

[greatmarx]

The photo of the esr of the cap I have uploaded on the website.
It's a cap of Yaiyo.
I have changed the cap of Tantalum,the problem is no longer exist.
Is it any problem?
Yang...

 

[FvM]

Is it any problem?

No, if your satisfied with several 100 mOhm ESR instead of a few mOhm.