Balanced VLF Tx

[Davor]

ideal Tx (Rx) for IB MD - is 90 degrees phase
Resonance - is not optimal. Compromise: Qtx <7 and Frx> Ftx

I think I understand your reasons, but bear with me for a second to see mine and we'll both learn something in the next few posts.
First thing to think about any VLF is what reference you take for a phase 0°. In case we use Tx voltage, and I think we do, then it is all the same what Q is the Tx tank: it provides sine voltage. di/dt is the driving force of any detector, and with constant L gives you a voltage u quite near 90° to the coil i, regardless of the way you apply it. Whatever the phase of the current, your voltage across the coil will still be the phase reference. High Q or low Q, same voltage -> same di/dt with constant L. The current change di/dt, is transformed to the target, and phase shifted according to the target tau, and again picked by the Rx coil. Now, Rx processes a phase shift with phase reference to the Tx coil voltage. Rx coil can work in detuned quasi resonant regime which is phase critical or in an aperiodic regime which is not. In case of a quasi resonant coil operation, it is BENEFICIAL if Tx frequency is adjusted according to the ground influenced Δμ change. This is why I prefer Tx tank being a part of an oscillator - it self-adjusts frequency to the Δμ change. It also does not in any way influence the phase change due to the target tau.

In my understanding, with any kind of voltage applied to a coil, you'll have exactly that voltage across the coil, and a current at 90° ... unless your Q is very bad.

So, where am I wrong in your vision?

[Davor]

its good if you can keep the Tx coil <2ohms.

It is easy if you go for a low L and use a thick wire. A 300uH coil is made of ~20m wire, and it is your choice how thick it will be. In case you use Φ1mm wire you'll get ~0.5Ω.

there is some similarity here in the drive bridge.

There is in fact a great similarity because both are H-bridges, only this one picks it's drive from a resonant tank in a cross-coupled fashion.

[ODM]

The merits of differential oscillator configurations are well discussed in electronics literature, way back.

However - What has been the chief reason of not implementing one earlier? Added number of components and need for differential measurement to get actual phase into inexpensive unipolar circuitry?

There are various highly harmonic-clean oscillators in unipolar configurations as well, though they tend to sport rather large amounts of components. I'd believe complexity to be the case, but does anyone more familiar with MD development history have more on this?

[Davor]

There were some Russian rigs that used a cross coupled oscillator, but it requires a centre tap. People are reluctant to wind their coils bifilarly, and you just can't buy any commercial coils with a centre tap.

Resonance - is not optimal. Compromise: Qtx <7 and Frx> Ftx

There is a way to make a Rx aperiodic, e.g. a series tank with Q<0.5. You'll find this configuration in Tejon and some other rigs as well, but the prerequisite to do so is a low noise preamp at ~4nV/sqrt(Hz) or better. In such case you have no problem with the Rx phase whatsoever, and the C is added to act as a high pass filter and noise optimiser. A quasi resonant circuit requires Frx> Ftx and is very touchy regarding phases, but not the aperiodic circuit. Feel free to consider a quasi resonant circuit to be obsolete. I'll put some of my ideas on how to make a nice and cheap low noise preamps soon.

I played with this oscillator to make it run with a coil at Q~7 as you suggested, and it works fine with some components tweaking, but at sensible currents it gives about 3dB less than the oscillator with good Q. In short - nothing is changed regarding the Δμ drift. See for yourself

[Aziz]

However - What has been the chief reason of not implementing one earlier? Added number of components and need for differential measurement to get actual phase into inexpensive unipolar circuitry?

Patents!!!! The ******* patents!!!!
Aziz

[Davor]

I think not. There is a detector at least 30 years old called Rebeks that employs a cross coupled oscillator, made by Grozdev, see attachment. The patent mania happened after that.

I really have no idea if anyone made a cross-coupled oscillator in this kind of H-bridge configuration yet, and I doubt if this creation of mine is by any means original - it is obvious expansion of a cross-coupled oscillator, yet without a centre tap. As far as I'm concerned you are free to use this oscillator in any kind of device.

In literature you'll find something called "complementary cross-coupled differential LC oscillator", yet it is not properly balanced and it has residuals of a 2nd harmonic because a current source is present only on one side, hence single balanced. It is targeted for VHF to microwave and is always about CMOS integration. You may find references:
Craninckx, J., M. Steyaert, H. Miyakama, "A Fully Integrated Spiral-LC CMOS VCO Set with Prescaler for GSM and DCS-1800 Systems," Proc. CICC, pp.403-406, May, 1997
and
A. Hajimiri and T. Lee, "Design Issues in CMOS Differential LC Oscillators," IEEE Journal of Solid-State Circuits, vol. 34, no. 5, pp. 717–724, May 1999

Mine is for VLF, double balanced, uses BJTs and uses differential amplifiers with resistors instead of a single current source.

Both references describe a basic circuit upon which there are a plethora of patents issued ever since, circuit is as follows:

[Sergey_P]

First thing to think about any VLF is what

Pardon my ignorance, I do not understand – “VLF”
Your oscillator works great in BFO, but not in IB MD.
Magnetic field of the coil is distorted ground, which causes a change of its parameters.
Δμ - it is for BFO (frequency and phase – variable), generator monitors the properties of the environment
For IB MD - X (magnetic susceptibility) and σ (electrical conductivity), rx - receives a response from the environment (need: frequency and phase – constant)

[golfnut]

Your oscillator works great in BFO, but not in IB MD.

Magnetic field of the coil is distorted ground, which causes a change of its parameters.
)

IB coils get affected too with proximity effects.

If you clock a colpitts type osc with the main tank coil in the head and offer it pump it close to ground - the frequency will jump about too - not by a lot admittedly.

S

[Sergey_P]

not by a lot admittedly

target - the same small

[Sean_Goddard]

IB coils get affected too with proximity effects.

If you clock a colpitts type osc with the main tank coil in the head and offer it pump it close to ground - the frequency will jump about too - not by a lot admittedly.

S

Depends Golfnut, if the FTx is in the 100's of KHz then the frequency change will be greater. I suspect it's a log law thing (most things in nature seem to be) rather than an absolute %age of the fundamental (Carl will know this) I also suspect that change will be influenced by ground mineral matrix composition and concentration.

At VLF frequencies the change will, as you state, be minimal but NOT negligible. Try looking at the Tx frequency of any Tesoro machine when you pump the head...looks like FM (albeit with a low deviation). I raised this subject about three years ago but only now do I see circuits looking at the phenomenon.

This might be off topic, but what about “pumped” resonant coils? What I mean is; The timebase of a TV (old CRT type) is free running @ around 15625Hz (UK PAL D) but “synchronised” with one pulse every line. Anyone think there is mileage in doing this with a beeper coil, that is a resonant coil and just pulsing say every 2 or three cycles?

[golfnut]

[QUOTE=Sean_Goddard;159404]
Try looking at the Tx frequency of any Tesoro machine when you pump the head...looks like FM (albeit with a low deviation). I raised this subject about three years ago but only now do I see circuits looking at the phenomenon.
QUOTE]

I guess this is ok as the LO to the Rx gets the same treatment so no real delta.



The only delta potentially is that the Rx coil and the Tx coil will get pulled by different amounts -

S

[ODM]

Patents!!!!

I highly doubt it was the patents. This would fall under obvious and/or widespread things, as there are numerous manufacturers and types of differential oscillators - as many as unipolar ones, undoubtedly. Some clever biasing and level control arrangements have been protected, but mostly under their IC implementations.

So, what would be the actual reason? If it's not patents, then perhaps it is one of those secrets?

[Davor]

I think it is just inertia. All commercial coils are single strand Tx coils with no centre tap. We all learned in our schools some typical oscillator configurations, like Colpitts, Clapp etc. and they all are single ended, and they were all designed back in vacuum tube era, where you did not have something like "complementary tubes". Just inertia.

The only delta potentially is that the Rx coil and the Tx coil will get pulled by different amounts

Exactly, and that can happen only if your oscillator is driven by some external source, e.g. a crystal oscillator, or in any case in a proximity of a large target. Otherwise the resonance change for both Tx and Rx is the very same, and there is no delta to observe. In case of a free running Tx oscillator they are mutually compensated. The thing about targets is that their response is in fact a delayed one against the activation signal, so in that department you'll not see any difference within a wide Tx frequency span. Because of these reasons I favour a free running oscillator for VLF IB rig.

[Davor]

Magnetic field of the coil is distorted ground, which causes a change of its parameters.

Yes, it seem we have some opposite views on this one. My vision is that, since you can't do anything about the resonance change in Rx, your best choice is a Tx that follows the relative frequency of the Rx tank resonance, hence the free running Tx.
There is a phrase, "the proof of the pudding is in the eating", and I suggest you to try some well balanced Tx oscillator in Verator design. I expect it to have more accurate phase response that way. Since I don't have a Verator and you do, it is only up to you to determine if it is so or not.

[Sergey_P]

Discrimination, selection, target detection (GEB, ...) in IB - requires precise magnetic field parameters Tx, Rx ... Resonances - it does not provide.
Quasar, Verator, X-Terra,… have: Qtx<7 (external excitation) and Frx > (or >>) Ftx