[SOLVED] ultra low current sensing circuit

[nanock]

hi,

Is there any off the shelve circuit for sensing or monitoring currents of femto to atto ampere?
I have four capacitors that have initial fixed and equal value of 5pF. Their value changes with resolution of tens of zF, 10^-18F. One way is to sense the attoampere current. Even if there were a Capacitance to Voltage Converter, I will be appreciated.

Regards,

 

[mtwieg]

That's in the domain of laboratory grade equipment. The packaging alone of most devices will have leakage currents on the order of femtoamps.

Capacitance meters with resolution in the zF range exist, but work by using high frequency AC perturbation to bridge circuits.

 

[FvM]

Consider that an atto ampere is only 6 electrons per second. This restricts achievable bandwidth and signal-to-noise ratio seriously. Capacitance sensors like ADI AD7745/AD7746 are already operating near the physical noise limit.

 

[KlausST]

Hi,

Look at "AD7745/AD7746" it is a so called capacitance-to-digital converter.
They say 4aF resolution.

Klaus

 

[nanock]

Hi,
Thank you very much for AD7745/AD7746. But the datasheet mentions that the accuracy is fF limit. What is the difference between accuracy and resolution in AD7745/AD7746?
Regards

 

[KlausST]

Hi,

What is the difference between accuracy and resolution

Those are two different things.
And "precision" is the third, that often causes confusion.


Let´s say you have a DVM.
* Resolution is the step size you can detect. Lets say you have a 10.000V reference voltage source: your DVM show "9.97" Volts. Then it has 0.01V resolution. (you can not detect 0.005V for example)
* Accuracy is how close you are to the "real" value. Lets say you have a 10.000V reference voltage source: your DVM show "9.97" Volts. It is 0.03V OFF. This 0.03V is an accuracy issue.
* Precision: Means repeatability. Lets say you have a 10.000V reference voltage source: you measure the value on five days. Once each day.. The results are: 9.97, 9.97, 9.98, 9.96, 9.97, then the repeatability = precision = +/-0.1V.

Klaus

 

[mtwieg]

hi,
Their value changes with resolution of tens of zF, 10^-18F.

zF is 10^-21F, which are you actually talking about?

 

[nanock]

Hi,


Those are two different things.
And "precision" is the third, that often causes confusion.


Let´s say you have a DVM.
* Resolution is the step size you can detect. Lets say you have a 10.000V reference voltage source: your DVM show "9.97" Volts. Then it has 0.01V resolution. (you can not detect 0.005V for example)
* Accuracy is how close you are to the "real" value. Lets say you have a 10.000V reference voltage source: your DVM show "9.97" Volts. It is 0.03V OFF. This 0.03V is an accuracy issue.
* Precision: Means repeatability. Lets say you have a 10.000V reference voltage source: you measure the value on five days. Once each day.. The results are: 9.97, 9.97, 9.98, 9.96, 9.97, then the repeatability = precision = +/-0.1V.

Klaus

Thanks a lot Klaus. That is perfect. So when the datasheet says the resolution is 4aF, AD7745 can detect for example 4.004fF or 4.016fF and the accuracy is 4fF so when the output is 4.004fF maybe it is 0.004fF or 8.004fF. Am I right?

On the other hand, what is effect of PCB soldering? I mean, it will add unwanted capacitance but haw can I model this effect?

Regards

 

[c_mitra]

The results are: 9.97, 9.97, 9.98, 9.96, 9.97, then the repeatability = precision = +/-0.1V.

I am afraid you dropped one zero somewhere- the precision should be 0.01V.

Often we represent that in relative and percent terms, but that is a different story.