Trying to make a parallel plate capacitor

[vk6kro]

That circuit is a "T" network and, as you found, it will produce less output if the center capacitor is increased in value.

However, your test frequency was very low for those capacitance values. 100 pF has a reactance of about 22 Megohms at that frequency and 11 pF has a reactance of about 200 Megohms.

This is going to produce an analog output which may be OK if you are going to drive an analog meter.

 

[mathew086]

I don't have a 555 timer or schmitt trigger. How can i modify the astable oscillator circuit to get a logical output when capacitance changes?
I am not using an anlog meter.
By test frequency do u mean the frequency produced bythe astable oscillator??

Another idea that came to mind is this one.

Make an astable oscilator producing around 70KHz frequency. But the capacitor that we use in the astable oscilatoor is our test capacitor. So when ever there is a change in that C value, number of sqaure waves produced for a particular time period changes. i.e for 10pF ( which is the initial capacitnace, there will be too many number of sqaure wave and as C increases number decreases. )
So my question is, is there a way to count the number of waves produced by this astable oscilator and create a checker??

This is just an idea that came to mind. Dont know whether it is logical . anyway i tried a simulation and the results showed differences in number of square waves.

 

[vk6kro]

Yes, exactly right.

If the capacitor in the astable oscillator was the test capacitor, then the frequency of the output would depend on the value of this capacitor.

Having produced this output, you could use a frequency counter to read off the frequency:

Or, a better way would be to use the frequency counter to measure period. This is the time between rising edges of the square wave being tested.
The output from your pipe capacitor was fairly linear in period.

There are fairly cheap frequency counters available, especially as kits.
http://www.sparkfun.com/commerce/product_info.php?products_id=9003

There is also a chip called a LM2917 which converts from frequency to voltage.

However, the main thing, just now, is to test the actual pipe capacitor. You can get a reasonable idea of frequency by measuring the frequency with an oscilloscope.
The plastic bag in the pipe trick would let you do a test without setting up the whole plumbing setup. Just hold the pipe vertical, put the plastic bag in the pipe and gradually fill it with water with an electrode in the bottom of the bag.

You haven't said what this is for. Is the water in the pipe moving?

 

[mathew086]

Yep.. I ll try with a Frq to voltage convetr. I ll soon test this with pipe.

Wate in the pipe will be flowing. Actually this pipe is a T shaped pipe and the sensor is situated on the outlet side. Water from outside ( like rain water) flows to the pipe . SOmetimes objects can also come along the pipe. so the sensor has to detect the presence and give signal to the user.

 

[vk6kro]

I think you just said the magic words. "rain water"

Rain water is a poor conductor, so we may have a problem.

Maybe you could just get a transparent pipe and point a cheap video camera at it?

 

[mathew086]

Oops... does that mean this system won't function at all?? rain water is not directly flowing to the pipe. It flows over roofs and then comes to the pipe. so there are some kind of impurities along with it ( impurities on the roof surface)

Other options i evaluated where as follows. Each of them had some problems and that is why i came to capacitive sensor.
Pressure- How to detect pressure difference? Also only water presence could be detected.
Temperature - using a RTD, but its too expensive and also not directly applicable in water. I am looking for a cheap solution.
Optical - Using a IR emitter and IR receiver. But then the receiver and emitter heads project out in the path where objects come along. so they could damage the receiver and emmiter head. I actually designed a circuit for this one not yet implemented as the heads may get destroyed. Is there any option to modify this idea. The IR emitter and Receiver are placed on the bottom side of the pipe with them inside and circuit part outside. See figure. The emitter and receiver is too small 5 X 5 mm ( L X B) .

Is there any other idea or is it that not possible to get a result via capacitive sensor? Or is there a way to increase the conductivity by some means?

 

[mathew086]

It seems i will have to adopt the othe roption i.e to place the 2 plates parallel to each other on the top and bottom of the outlet pipe. When i calculate the capacitances produced by them it ranges from 0.2 pF to 15pF ( 0mm water level to 100mm water level).
How is it possible to detect such a small change of capacitance.?

 

[vk6kro]

It would be very difficult to measure such low capacitances.

I had a sort of an idea that might work.

What if you had a parallel chamber that would fill up with water when the main pipe filled up but had a wire grill on it so that all these foreign objects could not enter?

You could then mount a capacitor in there that changed its capacitance depending on how much of its plates were immersed in water. It could have parallel plates so it might not be very large.

I suppose the big question is, do you have room for something like that?

Otherwise, I think you would have to use some sort of a float arrangement on a pivoting arm, but that is hard to do if you have these objects passing down the pipe.

What are they, and what are you really doing?

 

[mathew086]

Hmm. actually there is no space for that. But how is it then possible to throw away the water collected in the parallel chamber.?

Is there any other way to detect the presence of rain water ( just the rain water only).

I thought of using the principle of conductance between two probes so that when water comes in between them, a path if formed and a transistor is fwd biased and an LED lights ON. But the problem is I am using a 9VDC. So when water comes in between the probes, electroysis or somethin like that happens and the resistance is increased. That means conductivity is there only for a very short time and then nothing could be seen.

What effect does it have when two plates ( one positively charged and other negatively charged) are placed horizontally with a small gap in between them and when somthin enters its field?? See attachment

Objects coming through the pipe could be small leaves, stones, sand, etc...They flow along the pipe and goes out..

 

[vk6kro]

See this diagram. I drew it with square pipe to show the idea. It could be circular.

There would be a flywire or similar side to the main pipe with a box on the side of the main pipe at this point.
So, water could flow freely beteen the two areas, but leaves, sand and rocks could not.

If the water level in the main pipe dropped, the water would run out of the side chamber.

In the side box could be a capacitor which would get its dielectric changed by being immersed in water.

If the water was rain water and didn't conduct, you would not be able to use conductance measurements.

 

[vk6kro]

If you just wanted to know if there was water in the pipe there was a thread about this, last week.
https://www.physicsforums.com/showthread.php?t=355740
Read right through the posts to get the idea.

You would just need to get one of these switches and a piece of plastic hose (3 mm) and mount it in your pipe. You could glue it to the side of the pipe.

When the water was 3 or 4 mm above the bottom of the pipe, the switch would turn off due to increased pressure in the pipe.

 

[mathew086]

I have already seen such pressure sensors. OK still iam evaluating the possible chances...I will let u know if there are any problems.

Meanwhile i looked out for something else.

I have seen in an article
http://www.plantstress.com/Articles/salinity_i/salinity_i.htm
that the electrical conductance of rain water ranges is around 0.01micro Siemens/m

So for 1 mm it is 0.000001 Siemens

So Resistance = 1/conductance = 1MOhm /mm

I made a circuit and incorporated a 10M Ohm reistor in place of the sneing probe and checked the reuslt. .See the attachment. WIll it work with the rain water??

 

[mathew086]

Hey Vk6kro

i would like to ask u something. I connected a copper plate to positive terminal and steel plate to negative terminal of a battery. Now when i connect those plates to a LCR meter and check whether the C changes, there is no change. But when i remove the battery and connect the terminals to LCR meter and chek for C changes, there ocurs the changes. Why is it that ??Is it not possible to detect the C change when the plates are being charged??

Also about the copper sheet u mentioned in your prevoius posts, will such a sticking copper sheet work??
https://www.distrelec.com/distrelec...7/$File/3M_Kupferband_glatt_Datasheet_EN.pdf"

 

[vk6kro]

It depends on the design of the LCR meter. You would have to read the book that came with it.
Normally, you would remove a component from any voltages before you test it, but I have seen "in circuit" testers. Yours doesn't like external DC voltages.

I meant to comment on your post about conductances. The article you mentioned was talking about conductivity. To get conductance from this you have to put it in a formula with the dimensions of the conductor.
It is a bit like getting resistance from resistivity.
However, rainwater is very variable in conductivity, so it would be best to just collect a sample and test it.

Yes, stick-on copper sheeting would be ideal. If you had to use multiple turns of it, you could solder it at the joins.

 

[mathew086]

ok..I will have tocheck with the LCR meter
I have also noticed another thing. At the moment, during the testing some water came in contact with the outer copper plate and also in the gap between it and the pipe.( I tested with a copper plate. I had a small gap in between them. When i get the stick on copper sheet i hope there will be no gap) Also in the real sitiuation the outlet of the pipe is connected to something else and so there is less chance of water coming in contact with the outer copper plate.
My question is, is it possible to isolate the outer copper plate from the surrounding. I mean like a covering material? what type of material could be used to cover the outer copper plate to isolate it from surrounding? ( FOr example if a hand is coming in contact with the outer copper plate, there is still a variation in capacitance and the sensor gives output in the wrong situation. so inorder to avoid such a thing...))

 

[vk6kro]

I did a calculation as per post 21 above, but with a spacing of 2.54 cm (1 inch) and an air dielectric to an outer cylinder that was grounded. I got about 12 pF.

This is imperfect because the outer cylinder has a greater surface area than the inner one, but it indicated that a 1 inch spacing would not introduce much extra capacitance. Compared with the 500 pF maximum capacitance, 12 pF doesn't matter much and may be an advantage. It would limit the frequency range of the oscillator.

Possibly a sleeve of polystyrene foam would be OK too.

Or, you could just have a 1 inch gap between the pipe with adhesive copper on it and any grounded objects, including the ground.

 

[mathew086]

Is polystyrene a good insulator for electric field? I gues its just a thermal insulator...Any way i going to try it.
Is there a way to detect a change in peak to peak voltage and obtain this change as high or low output?? For example, to detect such a change as in attachment ( 80mV change)

 

[vk6kro]

Could you draw a picture of the waveform you have and what you want to get?

I find MS Paint good for this.
If you are using Windows XP, it is available under Start/Programs/Accessories.

Show the relative DC voltages as well.

I think polystyrene foam is an excellent insulator, although it can not cope with heat or some solvents. In this case, it does not have to deal with high voltages, just provide a non-conducting space filler.

 

[mathew086]

I am using MS paint all the time. :p

When i use a astable oscillator that produces 4Khz output and connect it to a T bridge and an Op AMplifier. I tried to use an online circuit simulator http://www.falstad.com/circuit/"
[Click Circuits on menu bar>555 timer chip > squarewave generator, chnage voltage to 9 V , add C and op Amp by right clicking the circuit"passive components and active components"]

I tried to use PSpice for simulation , but I am new to it and the simulation did not work for me.( although i created a schematic)

When i give the parameters as in figure 1 ( R1 = 10K Ohm, R2 = 160 KOHm, C = 1 nF) F is supposed to be = 4.36KHz. The waveform that is obtained at the output is shown in figure 1. For both 10pF and 11pF values. Basically they are sqaure waves.since the frequency is too high they are too crowded. If we redce the frequency one can see it.
When the C ( red marked one)= changes, the output voltage also changes accordingly.

In figure1, upper one is the circuit and the lower 2 diagrams shows the output for 10pF and 11pF. The values shown in them are the upper and lower peak values. The difference between them is the peak to peak volatge and that is waht changes when C varies.

 

[mathew086]

Hey all..

i tried another method for designing the circuit. Used an astable oscilator with the C as my test capacitor. ( producing around 6 KHz to 130 Hz )( from C = 10 pF to 500pF)
then used a frequency to voltage converter and get voltage values from 5.6 V to 0.1 V.
I am now in designing and testing the circuit.
Hope it will work.

I have another question.. I saw this in another site http://www.electronicspoint.com/capacity-voltage-converter-two-555s-t33728.html"
a method by using 2 555 timers ( atable and monostable) . Can anyone explain the working of that one? I mean my C range is from 10pF to 500pF. So what frequecy output should have the astable and R& C for monostable??


Also Vk6kro: I need a bit theoretical explanation for this type of capacitance measurement. means this is not actually a parallel plate capacitor. But depends on the conductivity for change in C. am I right?? So the basic principle is "Capacitance due to conductance" or what??

 

[vk6kro]

If you have a string of very short pulses and send them to a low pass filter, the output of the filter will depend on the width of the pulses or on the number of the pulses.

In this case, the number of pulses is kept constant, but the width of the pulses is varied with the capacitance.

So, this circuit works like that. The circuit should be fairly linear and it will give a greater reading for a greater capacitance.
This would be a good time to graduate to CMOS 555's though. They have many advantages and don't cost much more than ordinary 555s.

If we assume the water is a conductor, then the surface of the water touching the inside of the pipe is one plate of a capacitor and it connects through the water to the stainless steel or copper electrode at the bottom of the pipe. The other plate of the capacitor is the copper stuck to the outside of the pipe.

 

[mathew086]

Heyy all... I connected the test capacitor in place of C1 in an astable oscillator. There was very speculative and perfect chnages in frequency. Now I am trying to connect the output of the astable oscilator to monostable and then use a level detector (somethin like a schmitt trigger)

and vk6kro, i experimented both with tap water and rain water. there is no much difference, the results werre same...:D. So it only the dielectric that makes the change, i guess...

thanks guys...

 

[vk6kro]

Heyy all... I connected the test capacitor in place of C1 in an astable oscillator. There was very speculative and perfect chnages in frequency. Now I am trying to connect the output of the astable oscilator to monostable and then use a level detector (somethin like a schmitt trigger)

and vk6kro, i experimented both with tap water and rain water. there is no much difference, the results werre same...:D. So it only the dielectric that makes the change, i guess...

thanks guys...

That sounds promising.

If you have a moving coil meter, you could set it up as a 0 to 5 v voltmeter (assuming a 5 volt supply for your 555s). Then if you put this on the output of the monostable, the reading on the meter will be something between 0 and 5 v depending on the pulse width of the pulses from the monostable.

You probably want the readout to increase with depth of water. This would correspond with increased capacitance.
So, the pipe capacitance should be in the monostable, not in the astable.
In the astable, increased capacitance would result in a lower frequency and hence a lower reading.

 

[mathew086]

I connected an astable oscillator ( F = 5.6KHz i.e t = 0.18ms) . The output was connected to a monistable oscilator, whose F = 10KHZ i.e t = 0.1ms). It was thenconnected to an RC low pass filter.

I need a lilttel help with next part. WHen the test Capacitor changes from 10F to 500 pF the peak voltage changes from 5.48 V to 800 mV. So how can I implement a schmitt trigger to set a HIGH when voltage drops down below 5.48 V and LOW when voltage = or greater than 5.48V.? See attahcment

 

[vk6kro]

Do you just want to know if there is water in the pipe at any level?

That can be achieved a lot easier than with this elaborate capacitor arrangement. That was to give you a continuous readout of how deep the water was in the pipe.

All you would need is a pair of contacts in the water to detect conductance through the water.