Need help interpreting an old circuit diagram: Transistor and RF Transformer

I tried connecting various coax cables, up to about 1.5m long, to my circuit.
Putting a short circuit across the far end of the cable resulted in the oscillator stopping, as I would have expected.

kdausman said:

This means that I cannot measure dielectric constants by this method on liquids that are conductive,

Click to expand...

I have previously done a few experiments relating to oil/water mixtures for other people here on ETO.
Look here:
In those two cases the waveform applied to the test sample was anything but a sinewave, more like a triangle wave.

Why your version of my oscillator will not work with a conductive fluid is still eluding me.
I feel another experiment coming on with my circuit and a jug of water.

kdausman said:

And my offer of a co-authorship is still open!

Click to expand...

I am pleased to accept. Thank you.

JimB

JimB said:

I am pleased to accept. Thank you.

JimB

Click to expand...

Well deserved.

JimB said:

I am pleased to accept. Thank you.

Click to expand...

Wonderful! I will be in touch as I get the manuscript together. I have a student working on the measurements we need.

JimB said:

I feel another experiment coming on with my circuit and a jug of water.

Click to expand...

Well, I dunked the capacitor into a tub of water. Just plain old tap water as supplied by Scottish Water, and guess what?
It stopped the oscillation dead.

I tried to make a rough and ready measurement of the resistance across the capacitor while it was in the water, this was impossible using a multimeter as there was about 50mV of electrolytic emf being generated which made a complete nonsense of the readings on a multimeter (-4.7 Meg Ohm !?).
I tried again using an improvised set-up to measure the resistance at 1000Hz and disregarding any capacitive phase shift effects I measured a resistance of less than 10 Ohms, which will damp the tuned circuit sufficiently to kill any oscillation.

JimB

JimB said:

this was impossible using a multimeter as there was about 50mV of electrolytic emf being generated which made a complete nonsense of the readings on a multimeter (-4.7 Meg Ohm !?).

Click to expand...

the DMM uses a current source for measuring resistance, and the meter reads the voltage across the probes, so that kinda makes sense.

putting a layer of clear enamel on the plates might cure the problem.

Test Post.

Hola atferrari, are you here?

JimB

Hola JimB

Thanks for granting my access. Couldn't tag you nor the OP. (?)

I have no idea how I landed in this thread but I believed it was quite recent; now I realize it is older than one year. Sorry for complaining.

The content, I found it interesting to the point that after lot of reading I can distinguish between, Colpitts, Hartley and Clapp oscillators and even managed to build a Colpitts one by following your circuit.

My questions:

The OP explains that the procedure demands min / max values measured in air and min / max immersed in the sample under test. That seems to imply a linear relationship frequency vis a vis capacitance. Is it actually linear? I plotted your values from one of the last posts and they did not look so.

What if I simply use a fixed capacitor? Prone to error?

Why do we use an RF frequency which could compromise the readings? Isn't better going with a lower one thinking that it could be even easier to mesure?

Since I intend to reproduce your experience, could you give some practical suggestions?

The goal is to have it working just for the pleasure of it.

Gracias Jim

atferrari said:

The OP explains that the procedure demands min / max values measured in air and min / max immersed in the sample under test. That seems to imply a linear relationship frequency vis a vis capacitance. Is it actually linear? I plotted your values from one of the last posts and they did not look so.

Click to expand...

The OP is trying to measure the permittivity of various liquids.

Capacitance has a linear relationship to the permittivity of the dielectric between the capacitor plates.

By measuring the oscillator frequency at two values of capacitance (plates open and plates closed) with two different dielectrics, air (permittivity =1) and the fluid under test (permittivity unknown), by manipulating the maths it is possible to calculate the unknown permittivity.

The relationship between capacitance and frequency is not linear.
Using my favourite simplified formula F = 159.2/(sqrt(L.C)) frequency varies as the reciprocal of the square root of the capacitance.

It is possible to make an oscillator where the frequency is a very close approximation to linear with change in capacitance (OK linear to the reciprocal of change in capacitance), the change in capacitance must be very small compared with the total capacitance in the tuned circuit.

atferrari said:

What if I simply use a fixed capacitor? Prone to error?

Click to expand...

Do you mean to use a fixed capacitance for the sampling cell?
Yes in that case you will nor be able to get the best accuracy for the permittivity measurement.

atferrari said:

Why do we use an RF frequency which could compromise the readings? Isn't better going with a lower one thinking that it could be even easier to mesure?

Click to expand...

Good question, I think in the chemistry world they like to measure permittivity at high frequencies.

atferrari said:

The goal is to have it working just for the pleasure of it.

Click to expand...

A very good reason.

JimB

Many materials (plastics) are sold with a datasheet (Like electronic components) and permittivity is a common Line item. Permittivity at 50/60Hz is common for plastics intended for electrical housings on generic electric enclosures (from project boxes to LED light bulbs).

More recently, higher frequencies are getting published with frequencies at 2.4 and 5GHz permittivity. More recently, the various 5G frequencies and even for 76-81GHz for the Automotive object detection radar frequencies that are now the standard worldwide (originally were 24GHz). At 80GHz, the thickness of plastic parts are close to the half-wavelength in the material so they can cause high reflection losses or be almost invisible to the signal if properly designed.

Why the interest in permittivity?

Interesting tidbit....
BMW uses permittivity to measure the quality of the engine oil:

Mainly, lack of hydrogen-bonding alcohols and water (very high in permittivity vs organics).

Mickster said:

BMW uses permittivity to measure the quality of the engine oil:

Click to expand...

Interesting, thank you for that insight.

In the 1980s I had a car (Ford XR3) with an electronic dipstick for the oil...
what a pain in the sump that was!

gophert said:

Why the interest in permittivity?

Click to expand...

Well the OP is a professor of chemistry, so I guess he is interested in chemical stuff.
My own interest was piqued by "hydrocarbon leak detectors" which were fitted to oil/gas production trees installed subsea, in order to monitor the seal between the top of the well casing and the tree.
Since then I have assisted a couple of people here on ETO with oil in water and water in oil detection problems.

JimB

Found this: https://www.surplussales.com/INDUCTORS/IND-SMADJRF/IND-SMADJRF-1.HTML

One has a +-10 uH range.

JimB

Being about to build the oscillator, I still have the concept not clear enough yet.

JimB said:

The relationship between capacitance and frequency is not linear.
Using my favourite simplified formula F = 159.2/(sqrt(L.C)) frequency varies as the reciprocal of the square root of the capacitance.

Click to expand...

I recall that factor being used somewhere in physics (vibration of plates or something related to it)

JimB said:

By measuring the oscillator frequency at two values of capacitance (plates open and plates closed) with two different dielectrics, air (permittivity =1) and the fluid under test (permittivity unknown), by manipulating the maths it is possible to calculate the unknown permittivity.

Click to expand...

Would you mind showing the maths and explain the measurement sequence?

What if once built the oscillator I calibrate the variable capacitor in air with 10 or 15 well defined positions against frequency? I dare to say that I could go away without caring about linearity at all (?!) Kind of reading a look up table in a micro so to speak. The key concept would be repeatability instead. Am I too much off base here?

ATF, what is it that you want to do? ie what is the final objective of this exercise?

Build a variable frequency oscillator
or
Build a device for measuring permittivity

JimB

When I asked, above,

gophert said:

Why the interest in permittivity?

Click to expand...

it was intended for atferrari, not the chemistry teacher OP.

now JimB asks,

JimB said:

ATF, what is it that you want to do? ie what is the final objective of this exercise?

Build a variable frequency oscillator
or
Build a device for measuring permittivity

Click to expand...

Hopefully, a clarification comes....

gophert said:

it was intended for @atferrari, not the chemistry teacher OP.

Click to expand...

Thats OK, I was not sure whether you were asking me or ATF, so in the interests of completeness I answered from my own point of view.

JimB

JimB said:

ATF, what is it that you want to do? ie what is the final objective of this exercise?

Build a variable frequency oscillator
or
Build a device for measuring permittivity

JimB

Click to expand...

The intention is to build the oscillator that would allow implementing the device.

I've built a Colpitts already (rather unstable) but this morning I completed the simulation of a Clapp one which looks that it could work.

gophert said:

Many materials (plastics) are sold with a datasheet (Like electronic components) and permittivity is a common Line item. Permittivity at 50/60Hz is common for plastics intended for electrical housings on generic electric enclosures (from project boxes to LED light bulbs).

More recently, higher frequencies are getting published with frequencies at 2.4 and 5GHz permittivity. More recently, the various 5G frequencies and even for 76-81GHz for the Automotive object detection radar frequencies that are now the standard worldwide (originally were 24GHz). At 80GHz, the thickness of plastic parts are close to the half-wavelength in the material so they can cause high reflection losses or be almost invisible to the signal if properly designed.

Why the interest in permittivity?

Click to expand...

Found the subject very appealing. It moved me to read about Hartley, Colpitts and Clapp oscillators. I see this as a challenge to do something that works.

atferrari said:

Found the subject very appealing. It moved me to read about Hartley, Colpitts and Clapp oscillators. I see this as a challenge to do something that works.

Click to expand...

The colpitts seem to be much more stable.