Is there any way for commutation of TX coil between PI and IB schematics electronically?

[Waikiki_Sweep]

I have a theoretical question. I am thinking to make a PI-IB hybrid detector that will feed sinusoidal wave into TX coil between PI impulses.
Is there any way for commutation of TX coil between PI and IB schematics electronically? Looks like high voltage fly-back may burn any switch.

----> ????

[Tepco]

Actually no need to switch anything. There is one patent, using linear current ramp superimposed with sinewave signal (will post here if I find it, file is on another machine). Another approach (I tried something similar) is to use half bridge driver for coil, 2 mosfets and something like IR21xx, then in PI mode both mosfets turned off using shutdown command to generate flyback, in VLF mode normal square wave drive is applied to bridge. If you want sinusoidal, use series resonance and another mosfet pair to connect\disconnect tank capacitor, but square have more advantage.

[Waikiki_Sweep]

Thank you for good ideas. I need more time to digest. If you find some images thanks twice.
Also I have suspected that square wave is better for VLF. For square wave 2 mosfets is OK!
I saw many great multifrequency machines using square wave: Most Minelab machines like Excalibur, Sovereign, Explorer, ........ and Fisher CZ series.
But there is not much attention to multifrequency design on the forum. Such machines works great in salt water.
Many hunters preferring multifrequency detectors for salt water instead of PI machines because of descrimination and target ID you can't find in PI detectors.
Many times I have read on the forum that without sine wave there is no good target identification. I think it is outdated believes. Go square wave go !!!

[Tepco]

Here is patent I mentioned. (I don't think anyone ever actually built detector this way, outside lab) With square wave drive, this is technically PI, time domain data extraction. Yes, better target ID (using balanced coil) Also, in half or (better full bridge) configuration asymmetric drive, not 50% duty cycle, can be used to simultaneously charge coil with DC component and sample target response.

[Waikiki_Sweep]

US7078906.pdf Interesting idea. They using VLF at the time of transmit impulse. It is really easy solution. Just modulate current to charge coil with some frequency.
Only as I remember in classic PI transmit is only 50 microsecond and 1600 microsecond pause. So we have more time to use VLF in pause between impulses.

OK. We have square wave VLF signal sent. How to process it in receiver?
For square wave there is a well known schematic of two frequencies 5,15 kHz machine Fisher CZ5.
Machine is great, dip and sensitive to get a dime up to 12 inches in salt water beach. Only failure is to report beer caps as gold rings.
Also it is too easy to make a coil because it is nonresonant. I have made a 17" SEF coil for it from the first shot.



Schematic contains everything that needed to work with square wave but it is too big so simplification absolutely necessary.
First it is reasonable to replace timing block with microcontroller, lets say ATTiny84. Then transistors preamplifier asking to be made with some OPA350, OPA2350 op-amp.
Then I see tons of op-amps and switches I can barely understand. Should be some simple solution. I hope.

But anyway I can see my Surf PI is more dip for some targets and for basic search it is better to use PI then target identification with VLF.