IB Rx coil at Tx off

[green]

A couple of thoughts.

1) Regarding the high capacitance you measured between coil to shield. You mentioned that your tester works on the principle that it measures the time it takes to charge the capacitance. From my experience with graphite shielding, it has a high resistance compared to the metallised fabric that I use. Could it be that this fact increases the time to charge and which results in an erroneous reading?

2) I suffer from pickup noise in my workshop and have not yet identified all of the sources. One major source is the LED lighting which I have to switch off when taking measurements. The other source is the security alarm. Both systems run off voltage converters which are a source of radiated and conducted noise. I go round with a hand held AM band radio which pinpoints these noise sources. Even if switch off everything in sight that has a voltage converter, there is still noise which appears to be radiated by the ring main power wiring. Maybe what is needed is a heavy duty filter where the power comes into the building.

Like you, I can align the search coil for minimum pickup and that is what I do when doing range and sensitivity tests, even when known sources are switched off.

Eric.

Some testing with my capacitance meter. Measured a 100p and 1000p capacitor and with a 500 ohm resistor or 234uh flat spiral coil(not shielded) in series. Added resistance or inductance had little effect.
Made some more capacitors, glued aluminum foil to each side of pieces cut from 3/16 foam board. The (2x4)one side, (4x4)opposite side measuring the same as (4x4) both sides surprised me. Foam boards were 4x4inches for both.

Going to do more noise testing.

Forgot the p after capacitance readings for the foil capacitors.

[green]

Some testing with my capacitance meter. Measured a 100p and 1000p capacitor and with a 500 ohm resistor or 234uh flat spiral coil(not shielded) in series. Added resistance or inductance had little effect.
Made some more capacitors, glued aluminum foil to each side of pieces cut from 3/16 foam board. The (2x4)one side, (4x4)opposite side measuring the same as (4x4) both sides surprised me. Foam boards were 4x4inches for both.

Going to do more noise testing.

Forgot the p after capacitance readings for the foil capacitors.

More tests with the foam board.

[green]

Was playing with 24in fig8 coil couple days ago. Tried to detect a quarter at 24inches. To much noise, still haven't found source of noise that I can cancel by rotating the coil. Measured signal at AC coupled stage output absolute valued after the integrator with scope. Held a quarter at 16inches from coil and removed quickly, got about 400mV peak. The 24in coil looses about 1/2 the signal for every 4inch increase in distance at distances above 16inches. Should have about 100mV signal at 24inches. Get about 20mV peak noise with coil input jumpered. Signal should be greater than twice the noise so it looks like I could detect a quarter at 24inches if noise is low enough outside.

Moved outside today. Fastened a quarter to a drink bottle swinging from a tree limb(pendulum). Laid coil on the ground, to much noise. Did some testing with coil touching ground, sheet of plastic between coil and ground and coil spaced 10inches above ground on a foam container(_26). Don't know why coil is noisier on the ground. Plastic probably helps since I didn't coat the coil after applying the graphite shield? Rotating the coil didn't seem to make much difference. Noise less than 40mV with coil spaced 10inches from ground, maybe could detect the quarter at 24inches. Tested the 30in DD also for noise, lot noisier.

Pendulum test(_27)11 feet from limb to coil. Tx_324us, 1A peak constant rate. Delay_25us. Sample_103us. Tx rate_481pps. Got little over half the signal swinging from a pendulum that I got inside removing the quarter, integrator and differentiator TC not what they should be? Can see the quarter at 24in with target swinging across center of one coil but not good. Lot better when swinging from end to end. Shouldn't be a problem detecting a quarter with 4A peak Tx current.

Next is learning how to minimize noise.

[Ferric Toes]

I remember, a long time ago when testing a large coil outdoors, I got a big reduction in noise by actually grounding the coil. Drive a non-ferrous rod into the ground with a wire going to the electronics ground of the coil. Worth a try.

On the topic of graphite shielding, I have some Surfmaster coil shells that I believe have graphite shielding on the inside. I get about 1K with a simple test meter measurement across the diameter of the shell. For better accuracy it should be a measurement of the sheet resistivity which requires a slightly different technique.

Eric.

[green]

I remember, a long time ago when testing a large coil outdoors, I got a big reduction in noise by actually grounding the coil. Drive a non-ferrous rod into the ground with a wire going to the electronics ground of the coil. Worth a try.

On the topic of graphite shielding, I have some Surfmaster coil shells that I believe have graphite shielding on the inside. I get about 1K with a simple test meter measurement across the diameter of the shell. For better accuracy it should be a measurement of the sheet resistivity which requires a slightly different technique.

Eric.

Thanks for the suggestion. Didn't have a problem with the 8in fig8 coil touching ground with my first detector. Didn't try to look at signal with a scope, threshold didn't seem to change raising or lowering the coil. Guessing scope and coil having different grounds might be the problem.

I was a long way from detecting a quarter at 24inches with the 8inch coil and was wondering what it would take to detect a quarter at 24inches. Wasn't expecting to detect a quarter at 24inches with 1A peak, thinking maybe 4A peak. Been wondering what I would change to make it better. Looks like circuit noise(Rx jumpered)is about the same as figure8 coil connected with the coil off the ground. Figure8 canceling noise outside in the yard, not as much in the house. Reducing circuit noise would help, been using 1MHz amplifier response. Needing 25us delay, using lower frequency response might lower the noise? With target signal about 1/2 second swinging 1meter/second with a 24inch coil my integrator and differential time constants are probably to high, don't know how to calculate what they should be? Lower integrator and differential frequency might lower circuit noise and increase signal?

Like the figure8 coil, not happy with my shielding procedure. Need to understand coil parameters better. Was thinking fewer turns since inductances were higher than calculated. Probably wouldn't help since circuit noise seems to control minimum signal.

Any thoughts?

Was going to make a detector circuit just for detecting a quarter at 24inches but decided to finish a detector circuit I hadn't finished to see how close I could get. The reason for higher frequency response.

[green]

Been reading some on shielding. https://www.geotech1.com/forums/show...061#post149061 (Then come up with a test method and we all try different shield materials?)from reply #38
(If not intended for water, just check any capacitive effect, laying hand over coil etc must not produce response or change in waveform.)from reply #3

A test using the hand to test for shield effectiveness. Coil(Rx 8in figure8, oval Tx)
With no shield, big effect when coil decays. Longer delay before sampling reduces change, wasn't expecting a change with 17us delay.
Is there a better way to test for shielding than using the hand?
Want to try some other shielding than graphite.
Any comments appreciated.

US quarter at 8inches causes 50mV change at integrator out with 5.3us delay.

[Tinkerer]

Been reading some on shielding. https://www.geotech1.com/forums/show...061#post149061 (Then come up with a test method and we all try different shield materials?)from reply #38
(If not intended for water, just check any capacitive effect, laying hand over coil etc must not produce response or change in waveform.)from reply #3

A test using the hand to test for shield effectiveness. Coil(Rx 8in figure8, oval Tx)
With no shield, big effect when coil decays. Longer delay before sampling reduces change, wasn't expecting a change with 17us delay.
Is there a better way to test for shielding than using the hand?
Want to try some other shielding than graphite.
Any comments appreciated.

US quarter at 8inches causes 50mV change at integrator out with 5.3us delay.

We must not forget, that the hand (and body) is also made of conductive material. Thus it is also a target.

How could we differentiate between the target response and the capacitive effect?

On your figure 8 coils, do you see a target if you put a quarter on the same position on each coil? How about putting the hand so it is over both coils? Left hand on one coil, right hand on other coil might make a loop and falsify the result.

[green]

We must not forget, that the hand (and body) is also made of conductive material. Thus it is also a target.

How could we differentiate between the target response and the capacitive effect?

On your figure 8 coils, do you see a target if you put a quarter on the same position on each coil? How about putting the hand so it is over both coils? Left hand on one coil, right hand on other coil might make a loop and falsify the result.

Signal is direct coupled thru integrator out. Signal is positive over one coil and negative over the other. Hands or targets over both coils can cancel. Last stage is capacitor coupled, then absolute valued so a target over either coil is detected.

Tried another test. Cut a piece of space blanket 22.5cmx30cm and taped it to a piece of 2mm craft foam 22.5cmx30cm to replace the hand for the test. Got a negative going signal during coil decay(picture 2)when I placed the SB over one of the coils, same as placing my hand. As I was bringing my hand close to the SB edge not over the coil the signal went more negative. Think I might be increasing capacitance. Where I was touching the SB wasn't over the coil and wouldn't have caused a signal change if the SB wasn't there. Thought the SB might replace the hand for the shield test until I added the shield and tried it. SB causes a visible scope signal during coil decay where my hand doesn't. SB is a short time constant target, hand might be but if it is it's a lot shorter.

[Tinkerer]

Prepared some SB test material.

1 sheet of cardboard with about 220x300 SB glued on it. will try tomorrow when glue has dried.
1 15mmx15mm SB on a long wooden stick.

for the capacitive effect we need to find a non-conductive material that accepts a static charge. Does anybody have suggestions?

Years ago I used the hand test to judge the shield effectiveness. without shield the hand would be detected at some distance and the hand would also introduce high frequency noise from a distance.
The shielding would reduce the sensitivity a little and would remove most of the high frequency noise.

In the meantime the sensitivity achieved has increased a lot, so that the hand can easily be detected at a greater distance, even with a shield. Adding a strong shield reduces the sensitivity significantly.

At the end we will have to compromise between sensitivity and noise, as always.

[waltr]

for the capacitive effect we need to find a non-conductive material that accepts a static charge. Does anybody have suggestions?

How about a sheet of polystyrene plastic or a sheet of white foam.
Both hold a static charge.

[Tinkerer]

How about a sheet of polystyrene plastic or a sheet of white foam.
Both hold a static charge.

Good idea. Polystyrene really holds a charge well. When cutting it with a saw, the sawdust clings to everything.

During winter, the air gets very dry here, up in the mountains. Static electricity becomes a real problem. When sparks of up to 1/2 inch length fly out of your fingers, it is better not to touch any static sensitive electronic parts. In the lab, I use a humidifier that evaporates about one gallon of water per day to keep the relative humidity above 40% to mitigate the problem.

Now, how could we control and measure a static charge, so that we could test a coil with it?

I think a Theremin is a musical instrument that works with capacitance. Maybe something of that kind?

[Tinkerer]

Graphite shielding test

Testing Graphite shielding.
The Pdf shows the beginning with no target.
Scroll down the pages.
Then about 10 sweeps with a sheet space blanket, of about 220mmx300mm at a distance of about 70mm.
Then about 10 sweeps with a piece of space blanket, of 15x15mm at a distance of 5mm
Then about 10 sweeps with the flat hand at about 70mm
The coil is about 200mm diameter overall, TX maybe 190mm diameter
RX is about 150mm diameter.
Shielded wit 2 coats of graphite paint, fairly thick. Homebrew graphite paint with Chinese graphite powder. The shielding drain wire has a 10 Ohm resistor and a ferrite bead (1206 size) in series at the connection to the board.
Pulse repetition rate is 25,000, bi-polar, that is one positive pulse every 20us and one negative pulse every 20us.
TX voltage is 5V
Signal processing is AC coupling from RX coil, LP filter, Opamp with gain of 10, Antialiasing filter, input to 12 bit ADC.
The scope probes are connected at input to ADC.
The PICO scope is set at 12bits channel A minus Channel B, and LP filter.
I feel that the shielding is acceptable, but still try to improve.

Comments please. What could be improved and how.

[waltr]

Good idea. Polystyrene really holds a charge well. When cutting it with a saw, the sawdust clings to everything.

During winter, the air gets very dry here, up in the mountains. Static electricity becomes a real problem. When sparks of up to 1/2 inch length fly out of your fingers, it is better not to touch any static sensitive electronic parts. In the lab, I use a humidifier that evaporates about one gallon of water per day to keep the relative humidity above 40% to mitigate the problem.

Now, how could we control and measure a static charge, so that we could test a coil with it?

I think a Theremin is a musical instrument that works with capacitance. Maybe something of that kind?

There are Static meters but they are a bit expensive. I have used them at places I work to test/measure work stations for ESD compliance.
Probably could build something that gives good relative measurements.
simple metal plate to a JFET and op-amp meter drive would do it.

[green]

Testing Graphite shielding.
The Pdf shows the beginning with no target.
Scroll down the pages.
Then about 10 sweeps with a sheet space blanket, of about 220mmx300mm at a distance of about 70mm.
Then about 10 sweeps with a piece of space blanket, of 15x15mm at a distance of 5mm
Then about 10 sweeps with the flat hand at about 70mm
The coil is about 200mm diameter overall, TX maybe 190mm diameter
RX is about 150mm diameter.
Shielded wit 2 coats of graphite paint, fairly thick. Homebrew graphite paint with Chinese graphite powder. The shielding drain wire has a 10 Ohm resistor and a ferrite bead (1206 size) in series at the connection to the board.
Pulse repetition rate is 25,000, bi-polar, that is one positive pulse every 20us and one negative pulse every 20us.
TX voltage is 5V
Signal processing is AC coupling from RX coil, LP filter, Opamp with gain of 10, Antialiasing filter, input to 12 bit ADC.
The scope probes are connected at input to ADC.
The PICO scope is set at 12bits channel A minus Channel B, and LP filter.
I feel that the shielding is acceptable, but still try to improve.

Comments please. What could be improved and how.

Is your time scale +-10 seconds? Is hand and space blanket stationary during the 20 second sweep? If yes, what are you measuring with the coil AC coupled?

[Tinkerer]

Is your time scale +-10 seconds? Is hand and space blanket stationary during the 20 second sweep? If yes, what are you measuring with the coil AC coupled?

The time scale is 2 second division.

The coil is stationary, the hand and targets sweep across the coil at a measured distance controlled by a spacer.

Although this setup is definitely a PI, it is not conventional at all. I would call it a very advanced technology PI, with many innovative features.

The TX includes 8 mosfets and the timing is very tricky. It took me years to find somebody who could write the firmware.

The coil is also very unconventional. There is really not more else at the frontend, I mentioned a single opamp stage with a gain of 10.
All the rest is done in digital processing. We use a STM32F446 for the processing.