PI shielding

[Tinkerer]

Electrostatic shielding of the coils
In the old times, with low sensitivity detectors, a quick and easy "hand test", was good enough for verifying the shielding of the coils.
As the detectors become more and more sensitive, they become capable of sensing much weaker conductors like the conductivity of the human body or the blood it contains.
Sometimes we intend to reduce this sensitivity, "dumb down" the detector. But, most of the time we want to keep all the sensitivity we can get.
So, what do we want to shield against? Electrostatic. Electrostatic is of capacitive nature.
Probably there are quite a few different ways to implement the shielding and many valid, different ways to implement a test to verify the effectiveness of the shielding.
Let's hear about it.
Ah, yes, we need to separate the coil shielding into 2 categories: VLF and PI, because different rules apply.
For myself, I have been using the following method for PI coils:
Apply a spacer over the coil, so that the shielding does not add too much capacitance to the coil and thus change the SRF self resonant frequency of the coil.
Trial and error has shown that about 1/8 of an inch or 3mm spacer works OK, more is better.
Over the spacer I apply a coat of graphite paint. Now here it gets tricky. How conductive should this paint be? It appears that we can compensate for lesser conductivity by applying a thicker coat.
Gap or no gap?
PI detectors transmit a very powerful magnetic field pulse that induces eddy currents in any and every conductor within it's reach. If the shielding makes a loop, the eddy currents will race around in this loop until all the energy has been dissipated.
A big eddy current loop generates a big magnetic field around it. This is not what we want from the shield.
What we want from the shield, is for the eddy currents to form many very small loops, very small magnetic fields that run in every direction, so that they null each other by coupling.
Therefor we want to leave a gap in the loop of shielding that we apply on the coil loop.
Testing the shielding effectiveness. We will talk about that next time.

[Altair]

Electrostatic shielding of the coils
....
As the detectors become more and more sensitive, they become capable of sensing much weaker conductors like the conductivity of the human body or the blood it contains.
....

Interesting and condivisible statements.
The above point is in my interest as I am study how to detect human hands for a particular application. Coil shield for sure has primary relevance in that case. If anybody has esperience on this tema it would be appreciated.

[Monolith]

Interesting and condivisible statements.
The above point is in my interest as I am study how to detect human hands for a particular application. Coil shield for sure has primary relevance in that case. If anybody has esperience on this tema it would be appreciated.

Interesting to look at it from your perspective. What is the hand? A conductive region? An antenna for AC frequencies? A capacitor for electrostatic? I believe it can be all of these.
So how can we separate the one characteristic from the others? Definitely an interesting challenge.

Please tell me more of what you want to achieve.

[Altair]

Hi Monolith,
yes, a couple of phisicals principle could be useful to detect the hand, also others more than you mentioned but in my research I am concentrated to the conductivity sensing derived to a magnetic field. So the answer is "A conductive region" with a caracteristic TC.

[Monolith]

Hi Monolith,
yes, a couple of phisicals principle could be useful to detect the hand, also others more than you mentioned but in my research I am concentrated to the conductivity sensing derived to a magnetic field. So the answer is "A conductive region" with a caracteristic TC.

Never tried to look at a TC of a hand.
It would seem that a TC could be quite variable due to many circumstances.
For example:
Is the hand big or small?
Is the hand sweathing?
Is the hand "pumped up" like after heavy exercise when the blood is flowing strongly?
Is the hand hot or cold?
What is the health condition of the person attached to the hand?
I wonder if a male hand would be different from a female hand?
Would a nervous, twitching hand be different from a relaxed, at rest hand?
So many questions.
Is anybody interested in the answers?

[bbsailor]

Electrostatic shielding of the coils
In the old times, with low sensitivity detectors, a quick and easy "hand test", was good enough for verifying the shielding of the coils.
As the detectors become more and more sensitive, they become capable of sensing much weaker conductors like the conductivity of the human body or the blood it contains.
Sometimes we intend to reduce this sensitivity, "dumb down" the detector. But, most of the time we want to keep all the sensitivity we can get.
So, what do we want to shield against? Electrostatic. Electrostatic is of capacitive nature.
Probably there are quite a few different ways to implement the shielding and many valid, different ways to implement a test to verify the effectiveness of the shielding.
Let's hear about it.
Ah, yes, we need to separate the coil shielding into 2 categories: VLF and PI, because different rules apply.
For myself, I have been using the following method for PI coils:
Apply a spacer over the coil, so that the shielding does not add too much capacitance to the coil and thus change the SRF self resonant frequency of the coil.
Trial and error has shown that about 1/8 of an inch or 3mm spacer works OK, more is better.
Over the spacer I apply a coat of graphite paint. Now here it gets tricky. How conductive should this paint be? It appears that we can compensate for lesser conductivity by applying a thicker coat.
Gap or no gap?
PI detectors transmit a very powerful magnetic field pulse that induces eddy currents in any and every conductor within it's reach. If the shielding makes a loop, the eddy currents will race around in this loop until all the energy has been dissipated.
A big eddy current loop generates a big magnetic field around it. This is not what we want from the shield.
What we want from the shield, is for the eddy currents to form many very small loops, very small magnetic fields that run in every direction, so that they null each other by coupling.
Therefor we want to leave a gap in the loop of shielding that we apply on the coil loop.
Testing the shielding effectiveness. We will talk about that next time.

Tinkerer, and all interested

When discussing shielding you need to consider the following things.
1. Coil self resonance by itself which represents the coil capacitance with a known coil inductance. The wire insulation thickness and dielectric constant will affect the self resonance frequency and the number of coil turns to get the desired inductance.
2. Changes in coil self resonance when one or more spacer layers are added.
3. The dielectric constant of your chosen spacer materiel. Lower is better.
4. Detectability of your chosen shield materiel. It should not be very detectable.
5. Area of the shield materiel and how much it lowers the self resonant frequency.
6. Coil to shield capacitance between the shield and either of the coil lead wires. Distributed capacitance is a very complex formula but from my experience it is about one quarter of the totally measured coil to shield capacitance based on the coil self resonance measurements.

If you record the data of the above points, you will start seeing repeatable patterns that will help you understand how all these variables interact. Reducing coil capacitance allows faster sampling.

Tinkerer introduced a very good topic that in not normally covered in enough detail.

Thanks,
Joseph J. Rogowski

[green]

Attempt at measuring hand TC. Used the coil I was playing with. Recorded amplifier out, persistence on to get both Tx polarities.

Some observations.
One Tx polarity effected more than other, a ferrite core(positioned on coil) was used to adjust what no target signal looked like.
It appears X signal dominates R signal out to maybe 4us.
With the nickel, signal reversed end A to end B. With hand as target, signal changed from no target but appears to be the same end A and end B(maybe hand isn't acting as a target?).
Don't know it means? Looking at each end of Tx coil with the scope, similar but not the same.
Hand doesn't seem to be detected with scope disconnected from circuit. Sometimes with scope connected seems it does detect the hand, where the scope is connected can make it worse.
Wish I had some answers not questions.
Anything else I might try?

CH2 not important, should have turned off.

[Monolith]

Tinkerer, and all interested

When discussing shielding you need to consider the following things.
1. Coil self resonance by itself which represents the coil capacitance with a known coil inductance. The wire insulation thickness and dielectric constant will affect the self resonance frequency and the number of coil turns to get the desired inductance.
2. Changes in coil self resonance when one or more spacer layers are added.
3. The dielectric constant of your chosen spacer materiel. Lower is better.
4. Detectability of your chosen shield materiel. It should not be very detectable.
5. Area of the shield materiel and how much it lowers the self resonant frequency.
6. Coil to shield capacitance between the shield and either of the coil lead wires. Distributed capacitance is a very complex formula but from my experience it is about one quarter of the totally measured coil to shield capacitance based on the coil self resonance measurements.

If you record the data of the above points, you will start seeing repeatable patterns that will help you understand how all these variables interact. Reducing coil capacitance allows faster sampling.

Tinkerer introduced a very good topic that in not normally covered in enough detail.

Thanks,
Joseph J. Rogowski

Thank you for your input.
Coil building is very complex.
One could call it an art, if there were not the need to design the coils so that they can also be manufactured with a high degree of repetability.
I definitively noted considerable diferences in coil capacitance, using stranded, 1x 19, tinned insulated wire for the TX coils. Same AWG, different type of insulation, like PVC, Silicone, PTFT and many other variations.

[green]

Hand doesn't seem to be detected with scope disconnected from circuit. Sometimes with scope connected seems it does detect the hand

Some scope pictures with ferrite core located in three different positions on the coil to balance. Had some hand detection with scope connected to circuit with all positions, position #1 only with scope disconnected.

[green]

Looking at each end of Tx coil with the scope, similar but not the same.

Scope pictures, waveform similar but Tx off decayed to a different value. Maybe NFET Coss not the same? Some other reason? Should it matter?

[bbsailor]

Thank you for your input.
Coil building is very complex.
One could call it an art, if there were not the need to design the coils so that they can also be manufactured with a high degree of repetability.
I definitively noted considerable diferences in coil capacitance, using stranded, 1x 19, tinned insulated wire for the TX coils. Same AWG, different type of insulation, like PVC, Silicone, PTFT and many other variations.

Monolith,

Teflon insulation has the lowest dielectric constant near 2. PVC has a variable dielectric constant between 4 and 6. When you get down to low delays below 10 uS, and the lower the more critical the coil wire type chosen to make the coil becomes more critical so you are not detecting the eddy currents inside the coil wire at very low delays. Stranded wire with a slight resistance between strands will quench potential eddy currents within the wire itself. One of my favorite wires is a single strand of AWG30 with a silver coating and Teflon insulation because it is small enough to not retain eddy currents much beyond some very low delays.

My experiments with eddy currents in wires led me to making guitar pickups with a very thick, AWG 4, 6, and 8 solid wire going through a current transformer forming a single hairpin loop around 6 magnets or a single magnet placed under guitar strings to make a low noise guitar pickup to match XLR low Z microphone inputs. When I looked up the skin effect, I discovered that thicker diameter wires sounded different than thinner solid wires as higher frequencies do not full penetrate to the thicker wire core and that reduces the current in the string hairpin loop that could be transformed to the output. When I looked at the high frequencies that PI detectors use for small targets, I realized that wire size and type has consequences.

Look up the skin depth for the wire size and type you choose for your coil.

Joseph J. Rogowski

[Tinkerer]

Testing the shielding effectiveness.
How can we measure the TC of the hand?
The first thing that comes to mind, is to eliminate the other possible responses that we might see when we approach the hand to the coil.
I have heard the expression "hand capacitance". This is an unwanted effect. How can we differentiate the "hand capacitance" from the effect of induction on the conductivity of the hand?
In the case of "hand capacitance" we have no current flowing. Therefore it is an electrostatic effect.


Let's start with the electrostatic shield.
My very crude test to confirm if the shielding is efficient is as follows.
I take a 3 foot length of 1inch diameter PVC pipe and rub it on my jacket until I hear crackling sounds. Then I test for electrostatic by passing the end of the pipe near my bare arm to see how the hairs stand up. If the static charge is strong, then I approach the end of the pipe to the coil of the running detector.
If the detector shows no reaction, I consider the shielding good enough.
This is a very simple and empiric test. It only works when the air is dry enough. Up in the mountains where I live, the relative humidity needs to be below 50%. With about 30% relative humidity it works real good. I manage to induce enough triboelectric charge to get nice sparks.
I would like to hear about a better way to test if the coil is well protected against electrostatic.

[Altair]

Looking at each end of Tx coil with the scope, similar but not the same.

Scope pictures, waveform similar but Tx off decayed to a different value. Maybe NFET Coss not the same? Some other reason? Should it matter?

Hi green,
seems me the TX_bipolar is not balanced respect to 0V.
Do you tried to swap the scope probes?

[Altair]

...
This is a very simple and empiric test. It only works when the air is dry enough. Up in the mountains where I live, the relative humidity needs to be below 50%. With about 30% relative humidity it works real good. I manage to induce enough triboelectric charge to get nice sparks.
I would like to hear about a better way to test if the coil is well protected against electrostatic.

Hi Tinkerer,
it could try with a dehumidifier in a close room in order to have a repeatable conditions of temperature and relative humidity; it's just an idea.

[green]

Testing the shielding effectiveness.
How can we measure the TC of the hand?
The first thing that comes to mind, is to eliminate the other possible responses that we might see when we approach the hand to the coil.
I have heard the expression "hand capacitance". This is an unwanted effect. How can we differentiate the "hand capacitance" from the effect of induction on the conductivity of the hand?
In the case of "hand capacitance" we have no current flowing. Therefore it is an electrostatic effect.


Let's start with the electrostatic shield.
My very crude test to confirm if the shielding is efficient is as follows.
I take a 3 foot length of 1inch diameter PVC pipe and rub it on my jacket until I hear crackling sounds. Then I test for electrostatic by passing the end of the pipe near my bare arm to see how the hairs stand up. If the static charge is strong, then I approach the end of the pipe to the coil of the running detector.
If the detector shows no reaction, I consider the shielding good enough.
This is a very simple and empiric test. It only works when the air is dry enough. Up in the mountains where I live, the relative humidity needs to be below 50%. With about 30% relative humidity it works real good. I manage to induce enough triboelectric charge to get nice sparks.
I would like to hear about a better way to test if the coil is well protected against electrostatic.

Hi Tinkerer
Good thread. I have wondered why the hand can be detected. Have no idea why. Tried some things this morning. Could get the PVC pipe to pickup a small piece of paper but couldn't get it to light the led with scope ground connected to circuit common or not connected(led lights when signal exceeds threshold). With a 6us sample delay, led lights if scope ground lead is connected to circuit common when I move my hand near the coil. Doesn't if scope ground lead isn't connected. With a 10us sample delay, doesn't light with scope ground lead connected or not. At times have thought hand is acting as a target but when Rx signal is the same with hand over either figure 8 coil(should be opposite if it's a target)I'm thinking it's not a target.