HOW TO FIND CAPACITANCE VALUE

Use a C meter and measure it at various levels of liquid immersion.Not enough info for a specific calculation.

Hi,

as your gap is 1.5mm how do you make this constant over length or circumference of probe and time or temperature gradients?

How are you sure that your gap is clean?

If this is metallic (except platinum) there may be electrolytically induced corrosion.

If used in air or oil: ok. This is a usable pressure and temperature transducer as the dielectric constant will change with pressure and temperature (and air not constant and dry and without particles).

If used in water: water has a considerable conductance dependent on dissolved minerals and pH and temperature. And it will get some bubbles at the walls formed. So try to measure the capacitance non-disturbed by an unknown parallel resistance. And degas your water and select a frequency that avoids the effects from charging and discharging the absorbed layers of highly oriented water molecules that exist at any surface - very thin but very high capacitance.

Back to your questions: as your gap is not accurate, you can live with a non-accurate capacitance estimate. (As a plane-parallel capacitor). If you search a little bit you will find also the exact formula for the concentric cylindrical layout you try. (Log of d1/d2). But as the errors from eccentricity are likely to be much bigger than the error in thinking the tube-like air-gap as plane - stay with the plane estimate.

Which type of electronics to be used?

We made capacitive sensors of nearly any type: acceleration, distance, vibration, tilt, part detection, also different fluid-sensing applications. Any of these has some very good characteristics but also some unexpected and unwanted characteristics. Most disturbing was capillarity and its dependence on gap and evaporation and condensation - wetting and more.

RHABE

Why not use coax instead of reinventing the wheel? The C values are very accurate, and the attenuation at 100 Mhz is easily referenced.

Also, have you considerd using it as an RF probe?

It will be very difficult to reproduce the same degree of accuracy with a home made unit as is available in the coax.

A link to info on various coax characteristics is listed below:

http://www.antennasystems.com/coaxialcable/af-rgseriescoax.html

Perhaps this will help.

SSB

Hi SSB

100% air-filled coax is really pretty good,

but usually there are spacers inside coax to maintain the proper distance and at these spacers air-bubbles will attach at immersion thus deteriorating the accuracy of measurement.

Thats why I suggested flat ribbon HF cable.

Distances are not really good with HF insulating materials, neither absolute nor stability.

In mechanical applications you often need 10^-3 to 10^-5 resolution of small capacitances.

Highly stable (mechanically) and highly insulating (electrically) material with a really constant and low "dielectric constant"?, quartz-glass yes but anything else only moderately. Mica-Teflon is ok. But pure PE or PP or PTFE is not. Quartz and LR-Invar is good but who will pay for that?

So most users turn to metallic or ceramic structures to be metallised and forget about the stray capacitances and don't care about instability. Temperature gradient (radial) will bend the above mentioned cylindrical probe and change the distances!

`RHABE

I guess I should have gone a little further with my suggestion.Insulate the "wet" end of the coax from the liquid.Attach a weight to the end to keep the coax straight.If there is a lot of wave action or sloshing in the tank, a stilling well surrounding the coax should be used.The stilling well also has the advantage of a fairly constant environment for the probe when the tank is not linear.I have used a similar probe with RF attenuation to control level in a 75000 gallon hot water tank.It was very accurate and reliable.The coax functions as an antenna, and the liquid acts like a wiper on a potentiometer(crude analogy) and changes the tuned freqency of the probe.

The electronics are fairly sophisticated,and I do not recal all of the details,but it used basically a frequency to voltage converter,an A/D converter and digital scaling circuitry to convert to engineering units and display value on LCD display.

The switching point for off/on alarm, etc were easily programmable.

Hope this proves helpful.

--------------SSB----------

Hi,

I do not understand this.

The coax cable has an inner conductor and an outer shield.

The fluid will only affect the outside of the outer shield. This has no signal if the cable is terminated properly .

So what is the effect to be measured?

Do you use the cable without termination: this would cause severe reflections depending on frequency. If cable short or frequency low then the signal imposed onto the inner conductor will couple capacitively to the outer shield and there maybe see some environment of air or fluid. So this would rely ob stray capacitance to an unknown and likely changing environment?

Please clarify.

RHABE

At proper voltage and frequency, the outer insulation acts a a dielectric between the inner conductor and the fluid.The fluid in effect has a variable capacitor effect on the coax depending on how much is immersed, the best I can remember.It has been over 20 years since I saw this application in use.The end result was like tuning an analog radio tuner, with the different "channels"(frequencies) corresponding to variable level in tank.

I will see if I can find a reference to it in my old pre-historic files.

--------------SSB-------------

Here is a link that explains it better than I can.It desribes in principal the system I was referring too.My memory is a little rusty on some of the details.

Perhaps it will give you some design ideas.I do remember that our probe was located inside of a 8inch diametr stainless pipe, that rested on the bottom of the tank, and the probe was anchored by a weight in the bottom of the pipe.The pipe had lV notches cut in the base so it could respond to level changes.

http://www.omega.com/literature/transactions/volume4/T9904-13-RF-CAP.html

__________________SSB_________________

Hi all,

You do not need to know the capacitance. If you are capable of building the circuitry to detect the change in C (usually by having the probe as part of an oscillator circuit) then the output from the detector can be scaled to suit.

As previously advised you will need to insulate at least one pole of your probe unless you can be sure of you fluid being pure and not contaminated with something conductive. Diesel tanks frequently have a problem with water contamination for example.

regards

Chas