I measure very low capacitance.
And I use the circuit from http://arduino.cc/en/Tutorial/CapacitanceMeter.
But I find that it is not accurate when the capacitor is 4pF.
Why would it happen?
Well, apart from stray capacitances, if you use a 10 Mohm resistor for charging the time constant is 40 uS. Even if you replace "millis" in the code with "micros" you'll not be able to measure that accurately as it's just too fast.
To measure capacitance that low with any accuracy you will need to resonate it with a known inductance in an oscillator circuit and count the frequency.
Russell.
But LC circuit is built in AC circuit.
If I use DC, how can I do it?
Btw, Capacitance measurement with the Arduino Uno | PIC Tutorials it seems it can work well in small capacitance. And it does not use RC circuit.
If you are measuring capacitance with an error of 2 pF, then it will be 2 pF error on 4 pF, or 2 pF in 200 pF. It will not be a proportional percentage error.
Nevertheless, you might find it works better to deliberately add a fixed parallel capacitance - perhaps 15 pF and just subtract that value from the result.
And remember, to measure these small capacitances, you cannot use flexible leads or move any other object near the sensing leads.
The best way to measure tiny capacitances like this is with a tiny measuring
device, otherwise you just tend to measure the capacitance of the meter w.r.t. the
surroundings. 4pF is very small, its the capacitance of a few inches of wire
for instance.
I would guess putting the capacitor in a differential LC oscillator circuit might be
quite a good method since frequency is easy to measure, and symmetric
differential circuits tend to cancel out some of the systematic errors.
The other good way to measure is by comparison with a set of known capacitances,
which will cancel out more of the systematic errors.
I have two capacitors and I want to compare it.
The resolution of measurement should be 0.001pF.
The range is very small and from 0pF to 20pF.
Btw, two capacitors are also in the range of 0pF - 150pF.
I think RC method does not work well in this case.
"you might find it works better to deliberately add a fixed parallel capacitance - perhaps 15 pF and just subtract that value from the result."
I do not quite understand it.
MarkT:
The best way to measure tiny capacitances like this is with a tiny measuring
device, otherwise you just tend to measure the capacitance of the meter w.r.t. the
surroundings. 4pF is very small, its the capacitance of a few inches of wire
for instance.I would guess putting the capacitor in a differential LC oscillator circuit might be
quite a good method since frequency is easy to measure, and symmetric
differential circuits tend to cancel out some of the systematic errors.The other good way to measure is by comparison with a set of known capacitances,
which will cancel out more of the systematic errors.
If I have two capacitors and measure their difference, should I measure them one by one.
And calculate their difference.
Or I should measure a differential capacitance directly.
Btw, I know one of the capacitors' value.
DevilMoo:
The resolution of measurement should be 0.001pF.
The range is very small and from 0pF to 20pF.
0.001 pF in 20 pF is 0.005% resolution. I think that will be well outside your capability (and mine) unless your name is Hewlett or Packard!
Russell
DevilMoo:
But LC circuit is built in AC circuit.
If I use DC, how can I do it?
Btw, Capacitance measurement with the Arduino Uno | PIC Tutorials it seems it can work well in small capacitance. And it does not use RC circuit.
This is express way to damage your Arduino. Don't do it.
Finding 0.001pF reading? You need to place your measurement inside some well-controlled room with temperature and moisture and probably not a lot of metal nearby. How you bend your jumper wires or how far you are to the circuit will change capacitance by way more than that I think. Did you anger your professor so he gave you this task as a punishment?
Lol.
I know there are many ways affecting it because it is really small.
How about 0.01 or 0.1pF reading?
When I use RC circuit, it is not accurate for measuring small capacitance.
Why is it?
Yes, 0.1pF maybe. This meter at $80 claims to do it:
My FLuke multimeter has 1pF as smallest reading and wrapping the test leads around my wrist will increase as much as 10pF. It constantly reads 70-80pF with nothing on the leads (stray capacitance). The meter is a couple hundred USD and is not a specialized capacitance meter.
Using LC circuit you are trying to resonate with the proper inductance value and in principle it is better at detecting small capacitance. RC is looking at time of charging and discharging.
For LC circuit, AC current is applied.
Can Arduino generate ac current?
I have searched giving a freqency to the lC circuit and measure the difference with and without DUT.
I am curious of what is going on.
DevilMoo:
When I use RC circuit, it is not accurate for measuring small capacitance.
Why is it?
Because you have not comprehended all the other advice regarding the size and critical physical construction of your measuring system.
I just have capacitors and uno.
I have bought NE555 and some resistors to try RC circuit or get the frequency from NE555 to calculate capacitance.
But RC circuit cannot measure small capacitance while NE555 is not stable even though the capacitance is fixed.
That's why I am trying LC circuit to obtain more stable and accurate circuit.
For RC circuit, I follow the step in arduino.
For NE555, I use astable mode of NE555. And connect pin3 to uno.
NE555?
You must be joking!
if it is the CMOS version of NE555, that may just be useful.
DevilMoo:
Lol.
I know there are many ways affecting it because it is really small.
How about 0.01 or 0.1pF reading?
When I use RC circuit, it is not accurate for measuring small capacitance.
Why is it?
Because the time you are trying to measure is very short and the Arduino can't do it accurately nor with adequate resolution.
Don't confuse resolution, the smallest change you can detect, with accuracy, the size of the measurement errors.
Russell.
CMOS type?
Do you have any suggestion to do small capacitance measurement?
I have looked AVR LC Meter With Frequency Measurement – Kerry D. Wong
but honestly I do not understand it.
I think resolution means that the smallest difference of two different values.
That is a Collpits oscilator and buffer amplifier to feed the signal to the Arduino for frequency measurement. It should be OK for rough measurements down to a few pF but it will not have 0.001 pF resolution. The accuracy will depend on the accuracy of the inductor, any stray capacitance and inductance in your circuit layout, and of course the accuracy of the frequency measurement.
A more refined version using the same principles would give better performance but I suspect would be beyond your capability so perhaps it's worth trying that one as a start.
Good luck.
Russell.
Would you mind to share some methods which can measure down to 0.01 or 0.001pF?
But maybe I build this one in these days first.
DevilMoo:
Would you mind to share some methods which can measure down to 0.01 or 0.001pF?But maybe I build this one in these days first.
As others said, methods exist but you won't understand. What is the reason you want to measure this capacitance? If there is no meaningful reason, you are better off doing something else. We are wasting time here.