I have an accelerometer with an integrated charge amplifier that I have to redesign.
The LF 3dB roll off is said to be 0.2Hz but the charge amplifier roll off should place it about 5 times lower. I unfortunately don't have any mean to measure this. I mean the mechanical rool off.
The given explaination is the sensor time constant, which, for a 1nF sensor, gives about a 1G leakage resistance.
This seems really low to me, even over the -50/+120°C temperature range.
The piezo ceramic? will create a few thousand pf cap which with the input resistance of the charge amplifier will create the LF roll off. Measure the time constant by turning over and watching the time constant on a DC coupled scope. If the time constant is too short pad out with (low leakage) capacitance across the piezo element.
The turning over stunt never works well in practice because even if the TC is long compared with the turn over time then the acceleration caused by the turning rather upsets things.
Accelerometer source capacities range from a few pF to many 10s of nF.
The input resistance of a charge amplifier is a rather dangerous concept - a pure charge amplfier is an inverting amplifier with a feedback capacitor - there is no resistance anywhere. Real charge amplifiers usually have a feedback resistor in // with the feedback capacitor and there will be leakage resistance and capacitance to ground from the input. There will aslo be some series resistance between the charge source and the charge amp input (often some is added on purpose for protection).
If the charge amplifier is working properly (ie the amplier has enough gain so that the voltage on its inverting input is small (in the context of the rest of the ciruit)) then the resistance and capacitance to ground do not affect either the gain of the complete system or the frequency response.
If you find this hard to believe then dig out your simulator and try it.
The way you can measure the frequency response is to inject a signal into the input via a capacitor. If you use one that is less than 10% if the value of the source (accelerometer) then it won't affect the frequency response even if the amplifier is not really working as a charge amplifer at the frequency extremes.
There is a bit about charge amplifiers on my web site.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.