Field information question using multiple TX and RX coils in Pulse Induction systems

[fixstuff]

Hi folks,

This may be a couple of dumb questions, but I'm curious about something. I know in a "traditional" PI system, the TX pulse has the current going through the coil in one direction. After the TX pulse is turned off, two samples are taken either on the same coil (mono designs), or on a separate RX coil. Now typically, a coil is optimized for a particular frequency which determines the windings, type etc.

If I could run the TX coil with an opposite polarity sequentially after the positive pulse is complete, and the two samples are taken, are there usable differences between the two different sets of signals? This is assuming the +TX pulse has two samples, one for signal, other to eliminate ground effect, and the -TX pulse has two sample, one for signal, the other for ground effect. This scenario is a single frequency for both +TX and -TX.

If I had a second TX coil, and RX Coil optimized for a different frequency, and I did the same process for that different frequency, would there be useful, different, information in the 2nd RX coil?

Could both RX coils be sampled at the same time to see feedback from each of the two TX pulses?

Scenarios:

#1 (Reading RX coils only after TX pulses)
a) TX Coil #1 ON, delay, TX Coil #1 OFF, delay, Read RX Coil #1, Read RX Coil #2, delay, Read RX Coil #1, Read RX Coil #2. <--- Frequency #1
b) TX Coil #2 ON, delay, TX Coil #2 OFF, delay, Read RX Coil #1, Read RX Coil #2, delay, Read RX Coil #1, Read RX Coil #2. <--- Frequency #2

#2 (Reading RX coils during and after TX pulses)
a) TX Coil #1 ON, delay, Read RX Coil #1, Read RX Coil #2, delay Read RX Coil #1, Read RX Coil #2, <--- Frequency #1
b) TX Coil #1 OFF, delay Read RX Coil #1, Read RX Coil #2, delay, Read RX Coil #1, Read RX Coil #2. <--- Frequency #1
c) TX Coil #2 ON, delay, Read RX Coil #1, Read RX Coil #2, delay, Read RX Coil #1, Read RX Coil #2 <--- Frequency #2
d) TX Coil #2 OFF, delay Read RX Coil #1, Read RX Coil #2, delay, Read RX Coil #1, Read RX Coil #2. <--- Frequency #2

[Qiaozhi]

I think your getting a bit confused when you say, "typically, a coil is optimized for a particular frequency". VLF detector coils are tuned to a particular frequency, but PI coils are wideband, and are not tuned. I presume you're referring to the resonant frequency of the coil, which is determined by the self-inductance and interwinding capacitance. This has nothing to do with "tuning" per se, but it does have an effect on the lowest sample delay that can be achieved.

[fixstuff]

I think your getting a bit confused when you say, "typically, a coil is optimized for a particular frequency". VLF detector coils are tuned to a particular frequency, but PI coils are wideband, and are not tuned. I presume you're referring to the resonant frequency of the coil, which is determined by the self-inductance and interwinding capacitance. This has nothing to do with "tuning" per se, but it does have an effect on the lowest sample delay that can be achieved.

Thanks for the reply. I wasn't very clear I guess, but, yes, I did mean the resonant frequency. I guess since the RX coil is more of a wideband, non-tuned application, then I could use the same RX coil for both TX coils, which begs the question of whether 2 TX coils is useful or not. I guess this is why optimally most designs boil down to just a mono coil. Since the RX feature is wideband, its not really critical. I was just thinking that the different coil might have been useful for the sampling on the transmit side as well, since most PI designs are all sampling after the transmit pulse has ended.

Probably the best thing for me would be to actually try out some of that, and look at it with a scope and actual data from the sampling, to see if there is anything there to work with.

Is there anything to the reversing polarity on the TX pulse in terms of the sampling? I was originally thinking of a dual H-bridge coil driver, similar to that driving dc motors, with the one H-bridge driving TX coil #1, and the other H-bridge driving coil #2. Even if there isn't anything gained in terms of extra info, by having a much smaller #2 TX/RX set, you could have perhaps a better signal pin-pointing, with display or audio tone for the different coil information, say a low audio frequency for coil set #1, and high frequency tone for the smaller coil set #2.

[Carl-NC]

If you use a bipolar pulse then you don't need to take the late Earth field samples, Earth field is inherently canceled. Therefore, you can run at a much higher frequency.

[fixstuff]

Thanks Carl.

Which would it make more sense:

1) Pulse +, Pulse -, delay, sample
2) Pulse +, delay, sample +, Pulse -, delay, sample -?

If its the second one, would mixing the two samples to get the difference get you the GB cancellation? Or could I simply average the two samples, since the Ground effect has been eliminated by the two opposing fields?

[Carl-NC]

You have to do #2, and you have to subtract the 2 samples: (Sample+) - (Sample-).

[Waikiki_Sweep]

Probably you can make it simple.

Just make two identical coils and switch it every cycle as TX/RX - RX/TX.
If coils are inverted you will have bipolar pulses.

Schematics I guess will be very simple: two MOSFETs, two dumping resistors, two preamplifiers and some cheap microprocessor like ATTiny84 to drive everything.

Also if coils are balanced you will have opportunity to read additional samples at the beginning of TX pulses (similar to VLF).
Such samples can help to identify is that target ferric or color metal.