Concentric coil for TGSL.

[dbanner]

Yes, I read somewhere in article by George Payne, where he mentioned ratio of tx/ buck of 3.3:1 for coils around 8 inches.
Yet on article it recommends 4 to 1.
Then again I seem to remember another recommendation of 3.4, so there is some flexibility depending on size of coil and exact method of construction with regards winding buck coil on top of Rx coil. But we understand basic premise of cancelling coil by anti-phase (feedback coil). Also, it is better to wind tx coil with thicker wire(0.4mm), as it directly related to battery drain.
Rx coil wound with relatively thin wire. At resonance, the Rx coil Xc and Xl cancel, leaving only R. Higher Rx inductance means higher Q.
I also read that motion detectors with fixed internal GB(as in the tgsl) are fixed at 0.5 degrees to the right of X.
So all the off resonance coils are making more sense to me for better ground cancelling. Though I am yet to do in depth study of the subject.
The difference between tx frequency and Rx resonance creats a phase shift. If you change the tx frequency, you change the residual phase of the Rx coil.
Now exactly how this is calculated, I couldn't say.
Take for example the golden saberl and the silver sabre. Both can use same coil, but tgs is at 15khz(14.5-16.1) and silver sabre at 12khz(11.6-13.2) tx-rx. Notice same frequency spread, alas same residual phase shift
I am not sure if this phase shift is linear over some range. Nevertheless understanding residual phase shift gives tremendous insight into what engineers were thinking when designing these circuits.
It is also for this very reason why coils are balanced slightly to one side of null. A coil which is balanced at null will have greater depth on the bench test but will be useless in the real ground. Because they are designed for motion circuit where ground is averaged out.(internal preset to reject ferrite). This is also why Nautilus DMC 2 coil gives such superior depth, because it can be manually nulled and ground balance adjustment made in all METAL channel.
Something to think about.

[waltr]

Read it again:
"The bucking or feedback coil is where most people get lost
when trying to make a coplanar coil. In reality, the bucking
coil is very simple to calculate. The turns ratio between the
TX coil and the Bucking coil is simply the ratio between the
area enclosed by the transmit coil and the area enclosed by the
receive coil.
OK, so for a transmit coil which is twice the diameter of the
receive coil, the turns ratio is simply 4:1."

The 4:1 ratio is ONLY when Tx coil is Twice the diameter (Area) of the RX coil.
If the TX/RX diameters are not 2:1 then you will get a different ratio for the buck coil turns.

Thicker TX coil wire would lower the total TX coil series resistance. Good if wanting higher current in TX coil.

[dbanner]

Read it again:
"The bucking or feedback coil is where most people get lost
when trying to make a coplanar coil. In reality, the bucking
coil is very simple to calculate. The turns ratio between the
TX coil and the Bucking coil is simply the ratio between the
area enclosed by the transmit coil and the area enclosed by the
receive coil.
OK, so for a transmit coil which is twice the diameter of the
receive coil, the turns ratio is simply 4:1."

The 4:1 ratio is ONLY when Tx coil is Twice the diameter (Area) of the RX coil.
If the TX/RX diameters are not 2:1 then you will get a different ratio for the buck coil turns.

Thicker TX coil wire would lower the total TX coil series resistance. Good if wanting higher current in TX coil.

Yes I'm aware of the 4 to 1 ratio. However here is how George Payne describes concentric coil construction, I didn't pluck the 3.3 to 1 ratio out of thin air. He states: ".....the reciever is about half the width of the transmit". Yet he recommends 3.3 ratio ONLY for coils 7 to 8 inches in diameter. Not 4 to 1. Curious indeed. Emery states that he uses twice the number of turns for the rx coil as compared to the tx coil(looks like he uses same gauge wire as well)so 4 to 1 works in that scenario. This may not be the case when replicating a brand name coil even if the rx coil is half the diameter of the tx coil. Maybe this is why the ratios diverge a bit. It not only depends on the ratio of the tx/ rx diameters, but the ratio of the tx/rx number of turns as well.

[dbanner]

Thinking about it some more:
Strictly speaking, it is the magnetic coupling between the tx and rx coils that determine the strength of the feeback coil required to null the rx coil, and NOT just the ratio between the tx and rx diameters. The 4 to 1 ratio works when the rx contains twice the number of turns of the tx and is half the diameter of the tx( using same gauge wire as the tx). This will not hold true if the rx coil uses different wire gauge and more than twice the number of turns, even if it is half the diameter of the tx coil.

[Hyena]

Yes, its not given by diameter, but area. After simplification its Feedback coil turns = Tx turns * Tx diameter squared / Feedback diameter squared
For Rx with half diameter of Tx its Tx turns * 1/4.

From unknown reason - for my 250mm tx with 38 turns and 80mm rx it gives only 4 turns, but I needed 8 turns in reality... I am still missing something.

[dbanner]

This is precisely what I am saying, not strictly area of Rx with respect to tx( or diameter, by which ever geometric ratio), but by the inductive coupling and the relative strength of the transformer action. The only reason why Emery can use 4 to 1 is because he maintains same gauge wire throughout and keeps the turns ratio at 2 to 1 between tx and Rx, so that the geometry of half the area results in 4 to 1 feedback turns. This will not be the case for a Rx which uses smaller gauge wire and more than twice the number of turns. The 4 to one ratio does not only rely on geometric ratio alone. This is the point I am making.

[dbanner]

Now when the Rx number of turns are far in excess of double the tx number of turns, even though the Rx coil is half the diameter of tx, it now requires more feedback cancelling hence smaller than the 4 to 1 ratio(this means more turns of feedback coil). It now requires 3.3 or 3.4 to one just as an example.
This is my observations, now more experienced men can better explain.

[waltr]

Thanks for the info from Payne.

Ahh... I am now agreeing that the ratio of areas I linked to is over simplified. Also a PI detector can get away with a wide range of coil parameters whereas a VLF needs coil parameters to very closely match the detector's circuit requirements.

[Ozoz_xba]

This is my concentric rx is not 1/4tx but i should admit that this is the best coil that i have made for tgsl.tx 0.45 and rx 0.30 wire

[dbanner]

Splendid, very good results. How many turns did you require for the buck(feedback)coil to get the null? I'll bet it was more than the 1/4 number of turns of tx coil. That's partly because the difference in the wire gauges between Tx and Rx, the thinner the Rx wire, the stronger the transformer action, resulting in more feedback turns in order to achieve null.
This is just how I see it at the moment.

[Ozoz_xba]

It's good that you're doing this, I just started a 10" using some foam cardboard composite stuff.
It is 5.25mm thick with 4.5mm foam using .25mm wire so this will hopefully be close to the same.
[ATTACH]23045[/ATTACH][ATTACH]23046[/ATTACH]

Splendid, very good results. How many turns did you require for the buck(feedback)coil to get the null? I'll bet it was more than the 1/4 number of turns of tx coil. That's partly because the difference in the wire gauges between Tx and Rx, the thinner the Rx wire, the stronger the transformer action, resulting in more feedback turns in order to achieve null.
This is just how I see it at the moment.

Oh nulling was very hard to get it in fact at first,but bucking coil was not mor than 30turns ,something with 20-30turn 0.3 wire.depth and stability work on very good.tx is 14.9 khz and puting the rx at tx is 13.9khz.when nulling take the 15nf on rx at the same at from the voltmetre get 2.6mV which is the best less than 4mV should be...if you tryying to null in without 15nf with voltmetre than you make mistake.you null it less than 4mv in voltmetre withouy 15nf but in fact it is 30mV with 15nf so the circuit will not work good.null it with voltmetre with 15nf if u dont have scope