Balanced VLF Tx

[Aziz]

Doing The Magic Again

Hi all,

Oh boys!, I have made either a super magic to the TX circuit from Davor or the spice simulation isn't working properly.

Result: More bang, simplified circuit, super high efficiency, low noise, perfect sine (less distortion)...

In a test simulation, I have:
Battery input: 12V
Battery drain: 29 mA (note, including amplifier too -> complete laptop sound card controller )
Coil: 300µH/0.5 Ohm
Coil current: 1.6 A peak-to-peak
Frequency: ~12 kHz

There is a very very tricky magic one can do to the TX circuit. I have to build it to check, whether it works or my spice simulation sucks. No doubt, you will get it. Just be patient.

Cheers,
Aziz

[Davor]

I checked again, the "N" scheme I thought about uses PLLs to establish GEB and Disc phases, and not Tx oscillation. My bad. There is a nice symmetrical Tx oscillator with complementary layout, yet not balanced, so it is only half way there

[Davor]

Maybe the 'total resonance' type machines beginning with N, have Tx and Rx coils with similar construction giving same amount of detune with proximity effects - and importantly in the same direction.

That is precisely why they work. I've seen one of the "N" scheme with a Tx realised kind of as a third walk of the balanced Tx solution as listed in my initial post of this topic:

- a digital source with phase sense at the tank to enable frequency steering of the digital source

where a PLL is used as a signal source and a phase sense.

IMHO for Rx it is better to optimise noise directly at a Rx coil using a low noise preamp optimised for ~100ohm source impedance and a non resonant design. It is incredibly easier to just plug a coil - any coil - and continue detecting than adjusting this or that, losing time and patience. Resonant design is low noise, but again it is limited by Rx coil resistance, so your net result is the same. Only non resonant yields an incredibly less complicated design. Low noise non resonant Rx front end is easy to obtain using only cheap garden variety components.

[kt315]

< L - 300uH >in
Q = wL/r = w/Δw => Δw = w/Q =r/L
Δw <=> Δф (-45°…45° = 90°)
Δф/Δw = 90°*L/r = 90°*Q/w
(Δф/Δw)min => Qmin, Lmin (one turn)

use http://translate.google.ru/#ru/en/ it is VERY SIMPE TO USE - only Copy your text and will get translating one in the English!!!
give this people big texts and your comments! HUGE TEXTS! because they so like them!
and may be this forum will see new detector. which will be patented by Minelab.

[golfnut]

I have no idea about X-Terra, but both Verator and Quasar have crystal clock controlled PA and not oscillator for their Tx-es. So actually it is a technical solution in lieu of proper tuning which is complicated at VLF, not a big advantage. I can't imagine a high Q tank being successfully supplied with external oscillation.
Oscillators are always dead on with phases, but very few work in balanced mode.

BTW, check the Tejon Rx

It may be that xtal plus PA type drive non resonant coil

But the Rx coil if resonant will drift in phase with proximity effects.

Maybe the 'total resonance' type machines beginning with N, have Tx and Rx coils with similar construction giving same amount of detune with proximity effects - and importantly in the same direction.

S

[Davor]

You lost me here. Phase exactness is a function of a phase shifter. In case it is supplied with a signal rich in 2nd harmonic, yes, it is exactly what happens in ground proximity, at diferrent phases, e.g. Geb and Disc, you have a completely different offset as a result of 2nd harmonic. See the picture, and just imagine what happens with a bit different 2nd harmonic phase. In such case a constant frequency oscillator is undoubtedly a better choice, but only because of the missing 2nd harmonic. Being voltage source driven, a Tx coil will always have u and i in almost perfect 90° relationship, and it is largely irrelevant for MD operation, yet marginally better if it compensates a Rx phase as an oscillator does. This advantage is completely lost in case of a Tejon-like Rx front end - it has negligible phase change because Rx Q is <0.5.

My reason for rising awareness of balanced Tx is for two reasons
- no 2nd harmonic in a symmetric design
- no common mode component against ground and a Rx coil, so all capacitive couplings effectively cancel each other -> deeper coil balance

My reason for oscillator as a Tx balanced source is simple, it works at a bare minimum power consumption for a given coil, and perfectly provides the above mentioned requirements.

[Sergey_P]

at resonance - phase=0
oscillator supports phase balance and distorts the received signal -> noise from the ground

[Davor]

No Aziz, this is OK. Sergey has troubles with English, and the way he puts his ideas here is quite OK considering the google translate alternative, which is much worse and less concise.

My interpretation...

< L - 300uH >

for a given coil with inductance of 300uH

Q = wL/r = w/Δw => Δw = w/Q =r/L

we have a Q that is a function of coil resistance, and is observed as a Δw bandwidth

Δw <=> Δф (-45°…45° = 90°)

within a Δw bandwidth we have a phase shifted -45°…45° against a perfect 90° at resonance

Δф/Δw = 90°*L/r = 90°*Q/w

Δф/Δw can be expressed as a function of Q (and a resonant frequency w)

(Δф/Δw)min => Qmin, Lmin (one turn)

and the minimum Δф/Δw is achieved with minimum Q

It almost makes sense in a system that is driven by a constant frequency. When such a coil that is within a tank is exposed to ground proximity effects, e.g. Δμ it's resonant frequency will go somewhere, say down. Because it is driven from a constant frequency source, it may observe some phase change if it is constructed as a loosely coupled filter, and it happens in a Rx front end with Frx > Ftx and high Z op amp input. With Rx we can not do anything about this phase change in a system driven by a constant frequency source, except lowering the Q. In Tx of such a system we actually can not observe such a phase behaviour because a parallel Tx tank is driven by a constant frequency voltage source, and a Tx coil, being clamped directly to the voltage source, has current constantly at ~90°, spoiled only by a coil resistance. In this configuration, a capacitor is placed in a tank simply to reduce power consumption, otherwise it has no influence on Tx operation. Provided you are happy throwing 0.5A with 300uH coil and without tank capacitor.

In conclusion, It is beneficial to have low Q in a Rx frontend, especially if Tx is a constant frequency driven coil. With Tx Q is irrelevant, unless it is ridiculously low so that power consumption goes through the roof.

[Aziz]

Soil has a minimal impact on the phase-frequency characteristic Tx if Q (quality factor) - the minimum. Minimum Q - at the minimum L (inductance) ... 1 turn
In the bandwidth (Δw) - phase changes by 90 °( Δф)
(simple arithmetic ...)

Thanks for your effort this time.

Aziz

[Sergey_P]

Soil has a minimal impact on the phase-frequency characteristic Tx if Q (quality factor) - the minimum. Minimum Q - at the minimum L (inductance) ... 1 turn
In the bandwidth (Δw) - phase changes by 90 °( Δф)
(simple arithmetic ...)

[Aziz]

coil for optimal Tx, with a minimum slope of the phase-frequency response
(PP – pinpointer)

Pinpointer (PP)
then you can use PP in your context.

<=>, =>, Qmin, Lmin.....
I'm missing some important sentences, what you are intending to say.
That's not enough. Please try to be more precise.
Aziz

[Sergey_P]

coil for optimal Tx, with a minimum slope of the phase-frequency response
(PP – pinpointer)

[Aziz]

< L - 300uH >
Q = wL/r = w/Δw => Δw = w/Q =r/L
Δw <=> Δф (-45°…45° = 90°)
Δф/Δw = 90°*L/r = 90°*Q/w
(Δф/Δw)min => Qmin, Lmin (one turn)

Hey Sergey,

that's a bad style. You have to make some comments on it as there are people here, who don't understand it what you mean. Same applies to the short cuts (like PP??). At least, you have to declare what you are meaning with PP.

Aziz

[Sergey_P]

< L - 300uH >
Q = wL/r = w/Δw => Δw = w/Q =r/L
Δw <=> Δф (-45°…45° = 90°)
Δф/Δw = 90°*L/r = 90°*Q/w
(Δф/Δw)min => Qmin, Lmin (one turn)

[Davor]

I have no idea about X-Terra, but both Verator and Quasar have crystal clock controlled PA and not oscillator for their Tx-es. So actually it is a technical solution in lieu of proper tuning which is complicated at VLF, not a big advantage. I can't imagine a high Q tank being successfully supplied with external oscillation.
Oscillators are always dead on with phases, but very few work in balanced mode.

BTW, check the Tejon Rx