PI shielding

[Tinkerer]

If anyone is interested in how decay chart reply #28 was done. I recorded a no target, target end A, target end B and another no target with a DSO. Copied data to Excel and charted.

The TX pulse induces eddy currents in any and every conductor within its region of influence. The region of influence has a radius of about 50 feet. Actually much more but I have never measured it any further.
A 10" square sheet of graphite coated paper or cardboard over a 5" coil will receive the field lines of within the TX coil area as, say positive field vectors and also the field lines from outside of the TX coil as negative field vectors. Eddy currents will form small eddies due to the relative high resistance. Boundary areas will null fast. Every eddy current eddy will generate its own magnetic field.

[Tinkerer]

The bipolar info sounds interesting, thanks.

Big mistake. I should never have said that. So now I have to open that can of worms to save my face.

OK, here is the first clue: I just measured the current flowing in the RX coil cable shielding by adding a 100m ohm resistor in series. 600mA of current during the positive Flyback and 600mA during the negative Flyback. This is only the RX cable shielding. It is not connected to the coil shielding. This has to be induced current, induced by what? The RX signal flowing in the wire? Or induced by the TX magnetic field pulse? Probably both.
Are the 2 sources of induction aiding? Or are they opposing? The current pulse looks clean, so either they are induced by the sources aiding or mostly by a single source. I would suspect the RX signal current running in the wire.

This seems to be a lot of current. Could somebody please make the same measurement to compare?
How is this relevant for a bi-polar system? Crosstalk?

[green]

If I understood right, The correct way to measure is "per square". For example: 2 each, good contact areas of 1" length, separated by 1", should give the resistance per square (inch).
The difficulty is to have a good contact on the surface of a graphite paint.

When we embed 2 drain wires into the graphite coating we do not necessarily have a perfect contact area, but over a certain length of wire, using stranded wire, we achieve the best we can.

Thanks
I've been using just the probes, get a lot of variation. Made a 1 inch probe per your suggestion. Measured few places on the shield I charted reply #28. 15 to 25 ohms/square. I read it should be a lot higher.

[6666]

Big mistake. I should never have said that. So now I have to open that can of worms to save my face.

You can drop the subject, there is no obligation to continue.

[Tinkerer]

Big mistake. I should never have said that. So now I have to open that can of worms to save my face.

OK, here is the first clue: I just measured the current flowing in the RX coil cable shielding by adding a 100m ohm resistor in series. 600mA of current during the positive Flyback and 600mA during the negative Flyback. This is only the RX cable shielding. It is not connected to the coil shielding. This has to be induced current, induced by what? The RX signal flowing in the wire? Or induced by the TX magnetic field pulse? Probably both.
Are the 2 sources of induction aiding? Or are they opposing? The current pulse looks clean, so either they are induced by the sources aiding or mostly by a single source. I would suspect the RX signal current running in the wire.

This seems to be a lot of current. Could somebody please make the same measurement to compare?
How is this relevant for a bi-polar system? Crosstalk?

Correction:
This really did not seem right, so I checked again. My mistake. It is more like a 6mA current peak at the Flyback. This scope has some great features, like automatically switching scales and adding antialiasing filters. Very useful, but got to check the numbers.

[Tinkerer]

Thanks
I've been using just the probes, get a lot of variation. Made a 1 inch probe per your suggestion. Measured few places on the shield I charted reply #28. 15 to 25 ohms/square. I read it should be a lot higher.

For VLF, they use much more, like 2k Ohm region.
For PI, in general they want much less, like 1-2 Ohms. The idea is less is better.

I think we need to dig deeper into the capacitance - inductance conundrum. There are several ways to look at it.

The shield adds capacitance to the coils. The TX pulse induces eddy currents in the shield. How does it all interact together? This is what we need to spend some time on to figure out.

[Tinkerer]

You can drop the subject, there is no obligation to continue.

Bi-polar TX can solve quite a few problems. However, it is not twice as complicated, it is more like complication squared.
Very few people dare trying bi-polar, but if we manage to get a few people together to discuss and exchange information it is worth doing it.

[green]

Bi-polar TX can solve quite a few problems. However, it is not twice as complicated, it is more like complication squared.
Very few people dare trying bi-polar, but if we manage to get a few people together to discuss and exchange information it is worth doing it.

My attempt at bipolar. D3 and V3 were added to unipolar circuit as a snubber, didn't use a snubber with my real circuits. Lot more parts with bipolar Tx circuit. Control circuits similar. Only advantage I could see with bipolar, can use a higher Tx rate since EF sample isn't needed. Some disadvantages, More Tx parts and I could sample sooner with unipolar Tx. My first attempt so I could be doing something wrong.

[green]

For VLF, they use much more, like 2k Ohm region.
For PI, in general they want much less, like 1-2 Ohms. The idea is less is better.

I think we need to dig deeper into the capacitance - inductance conundrum. There are several ways to look at it.

The shield adds capacitance to the coils. The TX pulse induces eddy currents in the shield. How does it all interact together? This is what we need to spend some time on to figure out.

Agree. Hadn't considered the shield acting as a target using graphite(big mistake). Worse with my figure8 coil since I'm painting the whole side(bigger target, increased TC). Have some thoughts to decrease TC, wondering how to best make shield so not to detect my hand. Wondering if there is a better(more repeatable)test than using the hand?

[Carl-NC]

Correction:
This really did not seem right, so I checked again. My mistake. It is more like a 6mA current peak at the Flyback.

*Phew!* I was hoping that was the case.

[Altair]

Agree. Hadn't considered the shield acting as a target using graphite(big mistake). Worse with my figure8 coil since I'm painting the whole side(bigger target, increased TC). Have some thoughts to decrease TC, wondering how to best make shield so not to detect my hand. Wondering if there is a better(more repeatable)test than using the hand?

About your experiment, I am not believe that hand was detected only due capacitive effect. A part could be by his conductivity. You have measured a TC of 0.7 uS, I am wonder what would be the scope image if shield was not present? I think very different.

[Carl-NC]

Only advantage I could see with bipolar, can use a higher Tx rate since EF sample isn't needed. Some disadvantages, More Tx parts and I could sample sooner with unipolar Tx.

Generally speaking, correct. Bipolar can also help with low frequency EMI such as inductive mains pick-up. The +/- sampling forms a comb filter with a null at DC and another at the TX frequency.

There is another way to do bipolar pulsing using a center-tapped coil, where each half is driven by a unipolar TX driver. Resulting in the same speed as a normal unipolar PI. Downside is a more complicated coil.

[green]

About your experiment, I am not believe that hand was detected only due capacitive effect. A part could be by his conductivity. You have measured a TC of 0.7 uS, I am wonder what would be the scope image if shield was not present? I think very different.

Not clear what you are suggesting. A piece of poster board 10x19 inches painted with graphite paint measured .7us TC(reply #28 and #29). Reply #29 includes target and no target recordings. I have recorded target decays with and without the shield, might change the no target recording but doesn't effect target minus no target chart calculated with Excel.

Couple reasons might not be do to conductivity. I have the unipolar Tx hooked up now. Can see a small change at amplifier out when I bring my hand near the coil but change is the same at either end of the figure8 Rx coil, polarity changes with conductive targets. Can't see change at amplifier out with a piece of #9 lead shot(TC=.7us) but do see a change at integrator out and polarity does change at opposite ends of figure8 coil. Don't see the same with my hand. What I see with my circuit, maybe different with a different circuit?

[green]

Tried charting the poster board graphite shield. A piece 5x10in and 10x10in. Both charted the same TC. Don't think they should, need to find why.

[Tinkerer]

Generally speaking, correct. Bipolar can also help with low frequency EMI such as inductive mains pick-up. The +/- sampling forms a comb filter with a null at DC and another at the TX frequency.

There is another way to do bipolar pulsing using a center-tapped coil, where each half is driven by a unipolar TX driver. Resulting in the same speed as a normal unipolar PI. Downside is a more complicated coil.

The comb filter really does an amazing cleanup of the signal.