COG/ (NPO) ceramic caps, I must be the last to know.

Don't know about layers. But the 330 pF is a 1206, and the 22n is an

John

Interesting. With DC bias (eg. coupling capacitors)?

Could you measure *any* microphonics definitely related to the capacitor or was it below your noise floor?

I should probably do some testing soon. Thinking of a calibrated piezo accelerometer and a vibratory actuator.

I've whacked various ceramic caps, surfmount on PCBs, and never observed microphonics. Maybe I didn't try hard enough.

Has anybody had actual microphonic problems?

John

Dielectric absorption. From RaP (RIP):

As Bob said, the model of a real film cap looks like an ideal cap in parallel with a bunch of series RCs.

Modern Polyphenylene Sulfide dielectric film caps are pretty good for DF (~0.05% typically) compared to polyester, but NPO is better, and PP film better again.

They were basically low pass filter circuits, either filtering a voltage reference or a signal to an ADC. The X7R's would generate measurable spikes when the board was tapped with a screwdriver. I thought it might be the PCB bending causing a strain in the capacitor body.

I *think* it was below the noise floor for the NPO/COG ones. I ended up using NPO for one ~1.5nF, and replacing some larger values with tantalums (resistor fed).

I will actually be doing just this soon for a new board revision, see if anything falls off. I will try some different capacitors if I get a chance.

Never "in the field" but I have observed it while testing. Even absent a piezo effect, it is reasonable that a strain would change the capacitance slightly (due to the dimensions changing). Hence the voltage when DC biased via an impedance.

Sorry, George, I have no idea about C0G's soakage.

John

mic=92s.

they

Is there

orption?)

n why

but

rature,

ted text -

I see polypropylene is nearly as good as teflon.

ere

Ceramic cap microphonic problems have shown up on ATE setups. Chip handlers can cause a bit of vibration.

Yes. Before I learned that ceramics cannot directly replace other types in audio circuits without microphonics. I could hear the swish/bumpetty/crackle of picking up the amplifier PC board with my hand. Changed to tants/electros.

John

John

C0G's have low DI/soakage. I don't think the cap will limit the Q of your circuit.

Building a mag field meter? The water resonance must be pretty weak at

John

I have. Not yet with surface mount, but I have experienced microphonic ceramic capacitors.

My worst memory of those is when I helped a friend fix a guitar amp. He took on some "dogs" (difficult-to-repair units) from some repair shop or music instruments/supplies store, for fixed price each if successfully repaired.

That line of making money did not last long - that looks to me like an OK job for someone who can do this and is living in someone else's basement.

The guitar amp that I helped my friend fix had strange feedback problems. My guess was that ceramic capacitors had piezoelectric effects, from maybe being made with barium titanate (also used in piezoelectric loudspeakers) - leading to the capacitors being microphonic.

Replace all ceramic capacitors of larger values - maybe .01 or .1 uF or greater. I don't remember what they got replaced with, possibly different ceramics, but the microphonic problem was solved by doing this.

I am expecting that microphonic properties of ceramic capacitors gets worse with dielectric formulations that are more aggressive in achievement of capacitance or stored-energy for the size of the capacitance in question.

And, I expect microphonics to be less of a problem when capacitance and stored energy are not large for the physical size of the capacitor.

Furthermore, dielectric types achieving greater capacitance and stored energy for a given physical size of a ceramic capacitor tend to have greater variation of capacitance with temperature - for a longer-existing example, Z5U. Something milder and common and having been common for a long time is X7R. Going extreme for stability in capacitance at expense of capacitance and stored energy in a given size capacitor among longstanding ceramic dielectrics - NPO/COG.

So, I expect ceramic capacitors with better-temperature-characteristic dielectrics and lack of aggressive ratios of stored energy and capacitance to volume to "generally" be less problematic with microphonic characteristics.

Oh well we already have one... The Earth's Field NMR uses a high Q front end. All the nice polypropylene caps that we used were discontinued... we made a 'life time' buy before they went away. But if COG ceramics would work, then that would be nice to know.

The soakage may not matter in this case.

"The water resonance must be pretty weak at 1 gauss."

The trick we use is to polarize the spins in a strong B field and then turn the polarizing field off and 'listen' to them decay in the Earth's field.

George H.