Tx and Rx coil inductance ratio

[Davor]

Easy: just see what is on offer at fleabay, and adopt it for your project.
30 years ago you couldn't afford a low noise front end to use low impedance coils, so the coils were not low impedance. Now you can.

[Polymer]

Thanks KT

See attached pdf in english

Thanks KT and Altra for posting the translation. I really enjoyed reading it.

[Kev]

Thanks Guys for posting this.
It is a bit hard to follow the translation, which is a shame since Kostin has done lots of experimentation and thinking.
I apreciate what it's like for others here whose mother tongue isn't English.

Anyway, is the gist of what he's saying something like - a front-end fully resonant to the fundamental highly sensitive but impractical in the field, but an rx coil resonant to the first upper or lower tx harmonic optimal?

Kev.

[Altra]

Kev, here's some more bits and pieces. George Payne discussing some thoughts.

http://cleotusrus.tripod.com/Baron/gb.htm

[Nandor]

interesting article!

[Carl-NC]

Good article, but a lot gets lost in translation.

In thinking of the coil as part of a transformer (actually, a double transformer), maximum transformer "gain" seems to be the obvious answer. Some of the old White's coils were as high as 8:1, and many other coils are 3:1 or higher. Yet a lot of Tesoro coils are approx 1:1, and Minelab BBS/FBS is less than 1:1, and no one is complaining they lack depth.

The reality of transformer ratio is that the "gain" you get applies equally to targets and ground. That is, a 3:1 coil gives you a 3x signal for targets over a 1:1 coil, but also a 3x signal for ground. Although external EMI isn't affected by the transmit side, a higher RX turns also couples in proportionally higher EMI levels. So as far as incoming SNR, transformer gain is a wash.

When I was at White's I did an experiment. The V3 coil has TX=540uH and RX=14.6mH (relying on memory here!) which is ~ 5:1 ratio. I replaced it with a Minelab Explorer coil which has TX=580uH and RX=470uH; the TX-L was close enough to the White's TX-L that no changes were needed on the TX side; on the RX side I bumped up the preamp gain by 5x. And because the V3 has all-electronic calibration, re-calibrating to the Minelab coil was simply the push of a button. The results: exactly the same performance, I could not (1st order) tell a difference.

High transformer gain does help in one respect: in the V3 example I had to increase the preamp gain to compensate for the lower transformer gain. This will increase thermal noise. However, the lower RX winding also has less resistance, which helps thermal noise. But, in general, higher transformer gain does slightly benefit thermal noise.

My own personal preference is toward a low transformer gain, between 1:1 and 2:1, especially for multifrequency designs. MF requires a wideband RX coil and that gets really difficult in a hurry for a high-turn coil because for self-capacitance. For single frequency, I would still stick with a low transformer gain because, for a production detector, repeatability is much easier. Also, low transformer gain coils have less copper and are lighter.

[Nandor]

What about the Quality factor of the Rx coil, should I make Rx coils with ticker wire than 0.25mm to reduce resistance? That Explorer coil with only 470uH must have low resistance so I think the Q factor must be high.
I was thinking about a DD coil for my Minelab oscillator TGS, with Tx 1mH 0.45mm and Rx 1mH with 0.30mm wire.

[WM6]

Good article, but a lot gets lost in translation.

Translation?

Wasn't originally written in English as above?

[Qiaozhi]

Translation?

Wasn't originally written in English as above?

The original Kostin article is written in Russian, and Altra supplied a translation in English.
Carl meant that a lot of the technical meaning has become lost during the translation from Russian to English.

[kt315]

first pic in the article

[Nandor]

So the goal to reach the best sensitivity is to find a frequency other than Tx were the link is strong, at it can have a pretty high phase shift ralitive to Tx signal right?

[ifrenide]

Hello, actually I want to mention something important. every on talks about transformer,witch has two bibines . here if we want anderstand this subjectwe have to bring all elecments wich is the window of any vlf metal detector.elements are:
1-(TX bobinel inductance ,frequency and amplitude ,wire gauge)
2-(RX bobine inductance, frequency and amlitude to be amplified,wire gauge)
3-( op amp gain and its characterestics)
4-RX tuning capacitor.can we make for example RX bobine with 456 mh and a 470 pf to match the frequency desired?we will take care about RX tuning capacitor characterestics to avoid amplification quality problems.
5-one more thing is this transformer you talking about, has (four) 4 bobines if we consider target and ground as bobines too .So this transformer is complicated ,since there is two dynamic bobines witch is ground and target, and two other static bobines in terms of inductance. so we have to introduce all these elements in how we choose characterestics of TX and RX bobines, to get the best result.
we can't take only one element and we say: this is the rule of thumb to make a good coil (transformer).
all these elements affect our coil including temperature,and elecromagnetic noise in metal detector's environment and we can consider it also as a micro bobine.
so we have a transformer with 5 bobines interacting with each other and affect RX signal witch is our litle boy ti take care about, and make it a strong man wich tell as the truth about what is under the ground .
best regards

[ifrenide]

Hello,I forget ,bobines shape and size.we can't make a vlf coil with 4 meter of diameter, or a coil with unregular shape. good bye

[Davor]

Oh, but we can. And we do.

[ifrenide]

I don't think so Mr Davor . I never seen a VLF coil bigger than this one .