Minimum Capacitance Cx on LM555

[wogoos]

Dear reader,
I want to measure a Capacitance between 10pF and 500pF and feed that information to a microcontroller. I was wondering if an LM555 configured in Astable Operation could do the job. The problem I have is that most LM555 charts show the relation between capacitance connected at the threshold pin and the frequency but they don't go lower then 1nF however the specification don't say anything about the lowest value of this capacitor.

Does anyone have experience with this and can tell me what the minimum capacitance can be on this threshold pin.
Is there an other better and cheap way to measure this capacitance.
hope someone can help.

regards Oscar Goos

 

[danadakk]

Possibility -

DIY Simple Autorange Capacitance Meter (10pF-10000microF)

Inexpensive, precision, autorange capacitance meter with a large range. By Mirko Pavleski.

www.hackster.io

Goals for accuracy ? Do you also want to measure esr ?


Regards, Dana.

 

[Audioguru]

10pF is way too small for a 555. The error will be 150% when the wiring and IC pins capacitance is 5pF and could be a higher wiring capacitance.

 

[mtwieg]

Short answer is yes, you can use a 555 timer for this, but you will actually be measuring stray capacitance on the circuit as well, which will decrease the overall sensitivity of the circuit. If that stray capacitance is known and repeatable, then you can just apply a correction factor to your measurement to extract the actual capacitance of the sensor (or whatever it is). But the SNR of the measurement will still be degraded by the stray capacitance.

Also, you should make sure to stay within the recommended frequency range of the 555, so you'll need to use larger resistance in the charging path.

Hard to give more specific guidance without more info on the application.

 

[danadakk]

This might be of interest -

https://www.qsl.net/iz7ath/web/02_brew/21_LCMeter01/fig/fig08.gif

Regards, Dana.

 

[wogoos]

Short answer is yes, you can use a 555 timer for this, but you will actually be measuring stray capacitance on the circuit as well, which will decrease the overall sensitivity of the circuit. If that stray capacitance is known and repeatable, then you can just apply a correction factor to your measurement to extract the actual capacitance of the sensor (or whatever it is). But the SNR of the measurement will still be degraded by the stray capacitance.

Also, you should make sure to stay within the recommended frequency range of the 555, so you'll need to use larger resistance in the charging path.

Hard to give more specific guidance without more info on the application.

Hi MTwieg,

I did a simulation with an 555 the two R1 and R2 resistors were 1M, and I used a variable capacitance between 50 and 550pF. I did see some nonlinearity for C=10..30pF after that the frequency of the LM555 output followed quite well the formulas. this simulation didn't indicate any weird signals shapes or other problems. The frequency ran between 1.6Khz and 10Khz good enough to read that in to a microcontroller.

Stray capacitance is for me not an issue, this capacitor is a contactless dirty water level meter (see link) of which the capacitance varies with the height of the water between 11 and 236pF. The microcontroller can solve the nonlinearity in case that is an issue, which i don't expect it to be.

This LM555 acts according the simulation quite well as a Capacitance to frequency period length converter and its even linear according to the formulas used for the LM555.

Question:
So according to what you know and what experience is, the C connected to pin 6,2of the LM555 can be as low as 10pF (here 50pF) assuming that the R1 and R2 are big enough to keep the frequency within spec. Have I understood that well?

--- Updated Feb 3, 2022 ---

This might be of interest -

https://www.qsl.net/iz7ath/web/02_brew/21_LCMeter01/fig/fig08.gif

Regards, Dana.

Thanks for your help I wrote a response below.

--- Updated Feb 3, 2022 ---

10pF is way too small for a 555. The error will be 150% when the wiring and IC pins capacitance is 5pF and could be a higher wiring capacitance.

Thanks for your help I wrote a response below. There are others who say it is possible. Have a look at my explanation .

--- Updated Feb 3, 2022 ---

Possibility -

DIY Simple Autorange Capacitance Meter (10pF-10000microF)

Inexpensive, precision, autorange capacitance meter with a large range. By Mirko Pavleski.

www.hackster.io

Goals for accuracy ? Do you also want to measure esr ?


Regards, Dana.

Thanks for your help I wrote a response below. I did look at the software that is used for this meter. It might work well for big capacitors but I have my strong doubts for the pF values. The AnalogRead() instruction takes 104us to complete, that is way too long for low RC times < than 100us

 

[danadakk]

You have done sim, does proto show the same linearity ?

Regards, Dana.

 

[Audioguru]

A simulator knows nothing about stray capacitance and interference pickup.

 

[wogoos]

A simulator knows nothing about stray capacitance and interference pickup.

That is correct, a simulation doesn't know that, that is why I have added a 50pF fix capacitor in parallel with the variable capacitor. That 50pF in the schematic of the simulation symbolizes the sum of stray+ input + zero level sensor capacitance + wires.

 

[dick_freebird]

My cheapo Harbor Freight multimeter (CenTech 61593) will
read down to 1pF (lowest full-scale range 1nF) and appears
pretty stable at the low end (it's not flickering off zero, with
leads but no DUT attached). Set me back about the same as
the parts kit to do what you're imagining.

 

[danadakk]

My cheapo Harbor Freight multimeter (CenTech 61593) will
read down to 1pF (lowest full-scale range 1nF) and appears
pretty stable at the low end (it's not flickering off zero, with
leads but no DUT attached). Set me back about the same as
the parts kit to do what you're imagining.

I have several and they are slowly all failing because function switch assembly made out of , dont say....dont say.....s...it.

 

[wogoos]

My cheapo Harbor Freight multimeter (CenTech 61593) will
read down to 1pF (lowest full-scale range 1nF) and appears
pretty stable at the low end (it's not flickering off zero, with
leads but no DUT attached). Set me back about the same as
the parts kit to do what you're imagining.

Thanks for your response but that doesn't allow me to read it in a microcontroller.