minimum sample delay

[green]

http://www.geotech1.com/forums/attac...4&d=1527807206 attachment from another thread.

Rx(two 1.5inch round coils connected figure eight) Tx(oval surrounding Rx). What is required to reduce sample delay to 2usec maybe 1usec? What measurements from existing circuit are needed if any? Maybe forget about existing circuit and define what is required to have sample delay at 1 or 2usec.

What is required to get the no target signal near zero volts at 1 or 2usec?

[rickb]

Use real litz wire for all coils?

[kt315]

What is required to get the no target signal near zero volts at 1 or 2usec?
--
i think physics rule is one either coil for vlf or pi - the transformator has to have equal number of turns.

TX = RX. ie in IDEAL transformer with coefficient of transformation egual 1. thats is your 'near zero'.
another way you always get dizbalance.

[green]

Use real litz wire for all coils?

Was wondering the same, the reason for flat cable thread as an option for litz. A mono coil sees the coil wire as a target. Wondering if the figure eight Rx coil doesn't cancel. Can position a US nickel near one coil until amplifier saturates. Then adjust position of another US nickel near the other coil until amplifier out looks the same as no nickels. If it cancels nickels why wouldn't it cancel the coil wire? Maybe litz would be better but I think something else has a larger effect on reducing the delay with the figure eight coil.

[green]

Tried adjusting Rd, has a small effect on minimum delay time. I've been holding test coil near an operating PI to determine test coil resonance. Carl mentioned in another thread to disconnect Rd with an operating PI to check resonance. Think my way gives maximum resonance but Carl's way is better if trying to calculate Rd with formula Rd=pi*L*resonant frequency. Wanted to chart coil decay volts with coil critical damped. Coil resonance with leads 1.38MHz, in circuit with Rd disconnected and x10 probe across coil 1.03MHz, in circuit with Rd disconnected and probe near coil not connected 1.17MHz. Coil inductance=416uH. Rd=1.38*pi*416=1803 ohms, Rd=1.03*pi*416=1346 ohms, Rd=1.17*pi*416=1529 ohms. Recorded data with x10 probe across coil with Rd=1332 ohms. The decay TC(1.8usec) is less than calculated L/R(3.12usec). Should it be or am I doing something wrong? Anything else doesn't make sense with the scope pics or chart? Running out of things to try with existing circuit to lower delay. Maybe winding another coil. Any suggestions for the coil or other things to try?

[KingJL]

"... Carl mentioned in another thread to disconnect Rd with an operating PI to check resonance..."

Carl's method of measuring resonance is most accurate, as the result will have included ALL of the capacitance (intended and unintended) of the TX circuit (i.e. coil, MOSFET, cable, internal wiring/routing, etc.). Also the correct damp resistance is not just a function of resonant frequency, but more importantly a function of the ratio of inductive reactance relative to capacitive reactance (which corresponds to the ratio of L to C as well as frequency). In other words if 2 coils have the same resonant frequency but different ratios of L to C, they require different damp R values.

[green]

Carl's method of measuring resonance is most accurate, as the result will have included ALL of the capacitance (intended and unintended) of the TX circuit (i.e. coil, MOSFET, cable, internal wiring/routing, etc.). Also the correct damp resistance is not just a function of resonant frequency, but more importantly a function of the ratio of inductive reactance relative to capacitive reactance (which corresponds to the ratio of L to C as well as frequency). In other words if 2 coils have the same resonant frequency but different ratios of L to C, they require different damp R values.

Agree. Rd=pi*L*resonant frequency(circuit not coil resonance). My point was, it can matter where the measurement probe is connected. I wanted to record coil decay so I measured resonance with the probe across the coil. With the probe removed Rd increases from 1345 to 1530 ohms. Maybe the best place for the probe to measure resonance is amplifier out?

[rickb]

Using litz wire where individual strand radius is smaller than skin depth at the highest frequency of interest, means that the detector will no longer detect its own coil, i.e. no eddy currents. Then you pick an overall strand count that will support the current you need to put through the coil. Then pick the next largest standard strand count that is available. You cannot achieve that with uninsulated twisted strands. Then and only then do you work out the coil topology for the lowest SRF. That will be the plan for my PI.

[KingJL]

... My point was, it can matter where the measurement probe is connected. I wanted to record coil decay so I measured resonance with the probe across the coil.Agree With the probe removed Rd increases from 1345 to 1530 ohms...

I am of the opinion that you never really KNOW that what you are observing is what is really happening when you are not observing it. In other words, when you observe anything, the result you see is an artifact caused by simply observing it in the first place! Think about it...!

[waltr]

I am of the opinion that you never really KNOW that what you are observing is what is really happening when you are not observing it. In other words, when you observe anything, the result you see is an artifact caused by simply observing it in the first place! Think about it...!

Yep, been there many, many times. Scope and meter probes have capacitance which alters the circuit operation. I've have circuits not working until a scope probe is connected, then the circuit works. Makes for very frustrating troubleshooting.

We like to call this the "Heisenberg Effect" (Although technically this is actually the "observer effect").

[KingJL]

"We like to call this the "Heisenberg Effect..."

I know... I apply it to life in general... not just electronics (and physics in general). That way, I eliminate a lot of conflict in my thought processes that would affect understanding vs. knowing.

[green]

Some of the reasons I want to sample at or near 1usec. I have charted TC vs AWG wire size. http://www.geotech1.com/forums/attac...4&d=1526683069 have tried AWG24, charts about .8usec(projected on chart), would like to chart AWG28 that I have been using to make my coils to see if it charts just over .3usec projected on the chart. At 5TC's, 1.5usec AWG28 would lose 99% of it's peak signal. I'm hoping if I understand what is required to sample at 1usec I can make a better coil to sample at 5 or 6usec. Eric posted in another thread he has sampled at 1usec, challenge to see if I can do it. I'm thinking coil inductance, wire size,Tx circuit resonance, peak current and amplifier response have an effect, maybe something else?

[KingJL]

... maybe something else?

TX current waveform (triangular, half sine, trapezoidal, etc.).

What you are doing is commendable and worthy of continued efforts. However, I would also think of your frame of reference... "minimum sample delay". In reference to what? In reference to removal of TX input pulse, actual MOSFET current shut-off (can be in excess of 2 usec after removal of gate signal), after coil current achieves some threshold level, etc. What is the objective? To achieve a fast coil or to detect a specific minimum TC target goal? To achieve the later does not necessitate the former... it is one means to achieve the result, but is not the only one. In fact I know of one detector that has excellent performance against a 3-5 usec TC target and samples at 50 usec after TX turn-off with a sample width of 50 usec... it really does not care about "how fast" the coil is (it has a TON of capacitance in the TX). What really needs to be thought about during observations of target responses is what is actually happening with target eddy currents within the context of the exciting TX current waveform (like Carl keeps saying "coil current waveform, not voltage"). Then you can implement a suitable solution to obtain the desired result. But that is probably suitable for a separate thread and discussion.

As food for thought about the above discussion, I give you a link to an article to tickle your brain cells.
http://iieta.org/sites/default/files..._A/54.2_02.pdf

[Ferric Toes]

Have a read of this article. It changed my thinking on using expensive Litz wire.

https://engineering.dartmouth.edu/in...s/stranded.pdf

Eric.

[KingJL]

Have a read of this article. It changed my thinking on using expensive Litz wire.

https://engineering.dartmouth.edu/in...s/stranded.pdf

Eric.

I can't justify purchasing 5000', otherwise this would be GONE!!!
https://www.ebay.com/itm/5000-FEET-S...gAAOSwnHZYj5by