some tips on a twisted DD for PI?

[rolling rock]

Lately I've gotten interested in building a PI with bi-polar drive.
And so I'm thinking that a differential wound DD coil would be the way to go, with each channel's receive input tapped from the opposite channel's inactive coil. To eliminate (adjacent channel interference (more like next-door-neighbors), two separate receive channels will be used which will get combined post integrators.

Seems great, but then, does anybody know of any off-the shelf differential DD coils besides the GPX-7000 D0D monsters? Somebody out there making those?

I'm sure results would probably not be so great if I were to just swap wires at the plug because the shield would be unbalanced.

Which brings up another question: What are some things I can do to perhaps minimize the need for coil screening?

Thanks everybody.

[Carl-NC]

I don't recall that anyone was ever licensed to make aftermarket GPZ coils. Probably you'll have to build one. It doesn't have to be DD or balanced, you can wind a single coil with bifilar (or twisted pair) wire. You can also run both coils to a single differential preamp and get both flybacks at the same time.

[AK48]

look here if it may helm :
https://www.nexusdetectors.com/after...archcoils.html

[rolling rock]

I don't recall that anyone was ever licensed to make aftermarket GPZ coils. Probably you'll have to build one. It doesn't have to be DD or balanced, you can wind a single coil with bifilar (or twisted pair) wire. You can also run both coils to a single differential preamp and get both flybacks at the same time.

I'm not looking for a GPZ style coil. I need a DD coil but with wires reversed on one side. I thought there some other folks experimenting with bi-polar drive but I'm just not

ready to accept a bi-filar coil's extra capacitance, so bi-filar wound, no, but maybe a dual-field or other concentric IB would work, given appropriate attention to shield connections (!), but D-twisted-D seems simplest.

Using a differential amplifier might work but that seems counter-productive. Shortly after a transmit pulse, one half of the system is in overload, so why not take the target sample only from the OTHER half, just to avoid

potential conflict.


I suppose ML's GPX patent was written to try and cover a wide range of coil coil configurations besides just their DOD, but really all I need is something that could be called a twisted double DD. I hope their patent didn't

cover that , too?

[Carl-NC]

When you said "GPX-7000 DOD" I assumed you meant GPZ-7000 because there is no GPX-7000 and the DOD is used on the GPZ, not GPX. In any case, there are multiple ways to do bipolar. What I understand you want to do is have 2 TX circuits drive 2 TX coils in an alternating manner. A DOD coil won't help, it only has 1 TX coil (and 2 RX coils). What you want is a DD coil where both windings are identical. But it doesn't have to be DD, the coils can be co-located; they don't have to be balanced. However, an IB coil may let you sample earlier.

Offhand, I don't know of any commercial coils with matched DD windings. I also don't recall that any Minelab patents covered any particular coil configurations. Finally, I don't know what you mean by "D-twisted-D"... do you mean a butterfly coil?

[rolling rock]

Sorry if I sometimes don't express myself plainly... hey I got the ML models mixed up. I'm not very familiar with their offerings.

Here's what I mean. Two D coils side by side, mirror images except that one is wound the opposite direction from the other.

Each channel's receiver taps the rx signal from the other channel's transmit coil. It will be a negative excursion and very weak in relation to what's on the other channel, so extra amplification is on the menu. I considered using a bi-filar coil and/or differential amplification but ruled those out. I think a differential signal path would be detrimental because the two opposite ends of the coil see temporally disparate signals that are in no way even approximately equal. That is to say, I would rather treat each channel as more or less independent.

Does that sound workable? Sometimes I go of in the wrong direction. When I first started this project I was expecting to be cancelling the ground signal with bi-polar pulsing. Then I started reading again and realized my mistake, so it's a good thing I hadn't expended too much effort on PCB art because what was there got ripped apart, but circuitry is simplified, so its a net plus.

[Carl-NC]

Yes, what you want to do will work, except that your diagram actually shows both coil wound in the same direction. That is, starting with TX and ending with Gnd they are both counter-clockwise. One needs to be clockwise.

[rolling rock]

one tiny detail...

OK I see. Talk about a false start. I think you have just stopped me from erroneously trying to use a DD coil and for that I thank you.


Er, so what happens when pretend I know what I'm doing, and I reverse the leads on one side. Is there a name for THAT? The inclination is to call it a DD but I think that would be wrong. It seems like such a picayune little detail except that that it's really not.

[Carl-NC]

Still a DD. DD describes the windings' geometry, regardless of how they are connected.

[Daves]

Still a DD. DD describes the windings' geometry, regardless of how they are connected.

Carl, if we are going to use a normal DD coil for single frequency VLF, should TX and RX be wound in reverse direction?

[Carl-NC]

It doesn't matter, it only affects the polarity of the RX signal. Personally, I always wind TX & RX with the same polarity.