Coil Values

[Daves]

What are the exact functions of the values in the coils?

What is the exact purpose of mH, LP, RP, Q, X, what are they used for?

There is something I don't understand, let's say the detector I am currently using has
10 turns 0.40mm tx coil
50 turns 0.20mm rx coil.

But if I take the coil winding number from a different detector and use it on my own detector, unfortunately the performance deteriorates.

Note: If it is a resonant circuit, we change the resonant capacity according to the coil.

Let the coil of the different detector be like this;
40 turns 0.40 tx coil
180 turns 0.20 rx coil.
coil values are random.

The values change in both tx and rx coils, what should I understand from these, which value means what and what should be done for this. Should the RX input filter change according to TX or is there a different ratio, proportion, I have a lot of question marks in my mind about this.

[Carl-NC]

mH = millihenries, measure of inductance
uH = microhenries, measure of inductance
I don't know what you mean by LP & RP
Q = quality factor = 2πf*L/R
X = reactance = 2πf*L

Coil values (inductance L and resistance R) are usually an integral part of the design. For example, the TX inductance can be part of an LC oscillator where changing L will change f. Yes, you can change C to get f back to a correct value but you've also changed the ampere-turns of the TX which can affect sensitivity. And changing the R of the TX coil can affect the ground balance point in a VLF. The RX coil is similar; it might be resonated, and even if it's not decreasing the L can decrease sensitivity while increasing the L can cause overload.

You can't just put whatever coil you want on a design and expect it to work as well as the coil the circuit was designed for. In some cases, it won't work at all.

[Daves]

mH = millihenries, measure of inductance
uH = microhenries, measure of inductance
I don't know what you mean by LP & RP
Q = quality factor = 2πf*L/R
X = reactance = 2πf*L

Coil values (inductance L and resistance R) are usually an integral part of the design. For example, the TX inductance can be part of an LC oscillator where changing L will change f. Yes, you can change C to get f back to a correct value but you've also changed the ampere-turns of the TX which can affect sensitivity. And changing the R of the TX coil can affect the ground balance point in a VLF. The RX coil is similar; it might be resonated, and even if it's not decreasing the L can decrease sensitivity while increasing the L can cause overload.

You can't just put whatever coil you want on a design and expect it to work as well as the coil the circuit was designed for. In some cases, it won't work at all.

basically, let's assume that after creating a coil, we measure and note the inductance, resistance and other values. So how do we understand whether we can use this coil or not, that is, whether it is good or bad in terms of performance? How do we check which value by taking the coil L and R values? Or how do we understand which side it should go.
Let's assume;
single frequency 15khz
Resonant tx drive
Coil center tapped or 2-ended classic (does it matter?)
I'm asking for a VLF metal detector.

Note: Carl, I need more than 1 brick to build the wall

[Carl-NC]

What circuitry are you talking about? One that exists, or one that you will design? In a new design the circuitry & coil are designed together. After that, new coils should match the spec of the original coil.

You will see VLF designs that have TX coils ranging from 100uH to 10mH, and RX coils that range from 250uH to 50mH. What you choose for a design depends on the design. A gold VLF (say, 50kHz) will use low inductances. A relic machine (say, 2kHz) can use much higher inductances. But let's say you want to design a generic 10kHz detector. For this, most any inductances can be made to work well so you can pretty much choose whatever you want, but then the circuitry has to be designed around what you choose. If you choose a 500uH TX coil and a 50mH RX coil, that's a very high transformer ratio so the preamp gain will need to be lowered to prevent overload. Whereas a 1mH TX and 4mH RX is a much lower ratio so the preamp will need more gain.

Keep asking questions and I will try to clarify further.

[Daves]

What circuitry are you talking about? One that exists, or one that you will design? In a new design the circuitry & coil are designed together. After that, new coils should match the spec of the original coil.

You will see VLF designs that have TX coils ranging from 100uH to 10mH, and RX coils that range from 250uH to 50mH. What you choose for a design depends on the design. A gold VLF (say, 50kHz) will use low inductances. A relic machine (say, 2kHz) can use much higher inductances. But let's say you want to design a generic 10kHz detector. For this, most any inductances can be made to work well so you can pretty much choose whatever you want, but then the circuitry has to be designed around what you choose. If you choose a 500uH TX coil and a 50mH RX coil, that's a very high transformer ratio so the preamp gain will need to be lowered to prevent overload. Whereas a 1mH TX and 4mH RX is a much lower ratio so the preamp will need more gain.

Keep asking questions and I will try to clarify further.

Let's say we are going to design a new one,
Use: General
Frequency: 15khz, VLF
Method: Direct Sampling

1)What should be the number of turns and wire thicknesses of TX and RX? Or what should be the average coil values? (with which rx gain value 10x, 15x?)
2)What should be the input filter or filters and value (frequency) before the opamp gain at the RX input? (bp, lp, hp.....etc)
3)After determining TX and RX, will I only determine the rx opamp gain, is there anything else?
4)How much does the TX driving method affect the depth? How much difference is there between classic 2-ended coil driving or 3-ended center tapped? (What will change if we try the same tx coil with two different methods)

According to this information, I think as follows;
for average general use
700uH - 1mH TX coil
6mH - 10mH RX coil
Coils in this value range will be sufficient. So how can I improve it to get better results in terms of performance (signal quality, depth, stability, lower noise)? Based on this, how should the input filter and opamp gains be, do some properties of the magnetic field created by the transmit coil change according to the mH or uH value? Accordingly, does the rx coil or input circuit need to meet this (i.e. is there a kind of matching circuit)

[Kev]

Hi Daves,
Too much head scratching.
Build an established design, say TGSL, then tinker with things, work out why the present values were chosen, modify, push to the limit, extract all compromises, refer to Carl's book, fiddle, break things and fix them, then design from scratch, and then make it perform better than commercial in your niche of the woods

Cheers
Kev

[Daves]

Hi Daves,
Too much head scratching.
Build an established design, say TGSL, then tinker with things, work out why the present values were chosen, modify, push to the limit, extract all compromises, refer to Carl's book, fiddle, break things and fix them, then design from scratch, and then make it perform better than commercial in your niche of the woods

Cheers
Kev

I did what you said, these questions come after them.

[Carl-NC]

Let's say we are going to design a new one,
Use: General
Frequency: 15khz, VLF
Method: Direct Sampling

1)What should be the number of turns and wire thicknesses of TX and RX? Or what should be the average coil values? (with which rx gain value 10x, 15x?)
2)What should be the input filter or filters and value (frequency) before the opamp gain at the RX input? (bp, lp, hp.....etc)
3)After determining TX and RX, will I only determine the rx opamp gain, is there anything else?
4)How much does the TX driving method affect the depth? How much difference is there between classic 2-ended coil driving or 3-ended center tapped? (What will change if we try the same tx coil with two different methods)

According to this information, I think as follows;
for average general use
700uH - 1mH TX coil
6mH - 10mH RX coil
Coils in this value range will be sufficient. So how can I improve it to get better results in terms of performance (signal quality, depth, stability, lower noise)? Based on this, how should the input filter and opamp gains be, do some properties of the magnetic field created by the transmit coil change according to the mH or uH value? Accordingly, does the rx coil or input circuit need to meet this (i.e. is there a kind of matching circuit)

So first is to choose the TX & RX inductance. Generally LRX > LTX but it doesn't have to be. And for 15kHz LTX is usually 0.5mH-5mH. Let's choose LTX = 1mH and LRX = 5mH. You can change these but it won't make a whole lot of difference.

Next, decide what coil you want to make. For simplicity let's do a 10" DD and assume the windings have an equivalent circular radius of 100mm. For the TX coil, a coil calculator suggests 43 turns. For the RX coil, it's 92 turns.

Now choose the coil resistance and from that calculate the wire gauge. RTX is the more critical one because it determines how much phase shift you get in the TX current and also how much it will drift with temperature. These issues require more explanation than I have time for right now so we'll just choose RTX = 1Ω. This will produce a TX phase shift of 0.6°. The circumference of the TX coil is about 630mm so the wire length is 0.63*43 = 27m. The wire gauge needed to achieve 1Ω is therefore 20AWG or 0.8mm. The RX coil is (usually) not very sensitive to resistance, so let's arbitrarily pick 10Ω. The wire length is 0.63*92 = 58m so we need 27AWG or 0.355mm wire.

You then design a TX circuit to drive the coil with some voltage and current drive. On the RX side, you design the preamp so that it does not overload on some strongest desirable target at some distance, maybe a large penny at 4 inches. This will depend on the TX drive and the transformer ratio (which we chose to be about 2) so you can see that everything needs to be considered together. There is no best coil design, and once you have a complete design you cannot improve it by just changing the coil.

[Kev]

Ok Daves I understand now sorry.
A valuable methodological exercise.

[ivconic]

So first is to choose the TX & RX inductance. Generally LRX > LTX but it doesn't have to be. And for 15kHz LTX is usually 0.5mH-5mH. Let's choose LTX = 1mH and LRX = 5mH. You can change these but it won't make a whole lot of difference.

Next, decide what coil you want to make. For simplicity let's do a 10" DD and assume the windings have an equivalent circular radius of 100mm. For the TX coil, a coil calculator suggests 43 turns. For the RX coil, it's 92 turns.

Now choose the coil resistance and from that calculate the wire gauge. RTX is the more critical one because it determines how much phase shift you get in the TX current and also how much it will drift with temperature. These issues require more explanation than I have time for right now so we'll just choose RTX = 1Ω. This will produce a TX phase shift of 0.6°. The circumference of the TX coil is about 630mm so the wire length is 0.63*43 = 27m. The wire gauge needed to achieve 1Ω is therefore 20AWG or 0.8mm. The RX coil is (usually) not very sensitive to resistance, so let's arbitrarily pick 10Ω. The wire length is 0.63*92 = 58m so we need 27AWG or 0.355mm wire.

You then design a TX circuit to drive the coil with some voltage and current drive. On the RX side, you design the preamp so that it does not overload on some strongest desirable target at some distance, maybe a large penny at 4 inches. This will depend on the TX drive and the transformer ratio (which we chose to be about 2) so you can see that everything needs to be considered together. There is no best coil design, and once you have a complete design you cannot improve it by just changing the coil.

Very interesting and educational post.
Carl... most "classic" coils have different specs though.
Let's take for example 2 coils from two different manufacturers:
Minelab TS1000 (25cm) or TS800 (18cm), working frequency=5kHz
and
Tesoro series coils for "newer" Tesoro models (Tejon, Cibola, Vaquero etc) which are (22.86 x 20.32cm) 9x8",
working frequency 14kHz,14.3kHz,14.5kHz,14.7kHz,17.2kHz,17.6kHz etc
Why do I suggest these coils? Because they have very similar inductances; TX~1mH and RX~15mH.
Now...See where this is going?
How would you explain the following:
1) Same or similar inductance and thus different operating frequency.
2) The TX/RX ratio in terms of inductance is in both cases higher than the examples you mentioned in the last post.
It's much clearer what the TX coil specs will determine... but less clear what the RX coil specs determine.
Can you elaborate on this?
4) What happens if, for example; RX coil for Tesoro Tejon, we reduce the number of windings and reduce the inductance to 2-5mH?
And vice versa, what happens if we increase the number of windings and inductance to ... say 30-50mH?
I think that everything about the metal detector is more or less clear.
But there are still a lot of "holes" in the information about coils. Such questions should be fully clarified.
Here's an idea for a new chapter in your book. Not to write how to make a coil, but to explain such and similar questions.
One of the questions that has always interested me is the following: I was wondering what top engineers start from when they start a new metal detector project and what about coils?
How do they determine coil specs?
By what criteria are they managed?
Is the final version of the RX coil a "forced" version by the rest of the hardware or is it simply initially set as such and later the rest of the hardware is adapted accordingly?

[Carl-NC]

There is a lot of leeway in choosing coil inductances. 1mH and 15mH would also work in Dave's example. By increasing the LRX you end up with a higher transformer ratio. This increases the return target signal but also gives the same increase on ground signal and EMI, so it pretty much evens out. A higher transformer ratio means the preamp gain must be lower to hit the same overload points. I'm not a fan of designing for a high transformer ratio, it doesn't help much at all and just makes the coil heavier.

You can design for lower inductances. Minelab BBS & FBS uses LTX=500uH and LRX=500uH. A lower LTX creates a stronger B field for the same voltage drive. But a really low LTX (or RTX) makes it difficult to maintain the inductance if you want to make large coils because +/- 1 turn can make a large value difference. Also, as you go up in frequency the inductances tend to decrease because you need to maintain a high enough self-resonant frequency and also to keep the TX field strength up.

I cover all this in ITMD3.

[ivconic]

Thanks, another educational post.
When I want to learn something; first I look at how smarter people than me did it.
I used to be "obsessed" with Minelab Musketeer then Fisher series of detectors and of course Tesoro.
Of course, others too. White's is among the favorites but the older models used coplanar concentric coils which (for me) are more difficult to DIY. The same goes for Garrett coils.
That's why I've only mentioned TS coils and the "newer" Tesoro coils for now because they're easy to replicate.
Now with your explanation, the logical conclusion is that most RX coils could have a much lower inductance if the gain is increased in the front end?
Of course, you need to find the right measure.
What is still unknown is the "rule of thumb" for such calculations.
I don't believe that in serious manufacturers, engineers bother too much winding dozens of experimental coils in order to get the right measurement, there must be some fixed math behind that story?
And while things about TX coils are still much clearer; Things about RX coils continue to cause attention and many questions.
I'm sure the answer to this question is some "simple math". Because I see that there are several very successful coil manufacturers who easily, quickly and mass produce various sizes of coils for many existing models.
...
I am currently "obsessed" with the Minelab Vanquish 340 and its coils. I have no intention of "dissecting" coils, but I am very interested in inductance.
I doubt that all coils for the entire "Multi-IQ" series are very similar if not the same.
But we already know in advance that ML installed something in the coils and that easy measurement from the outside is impossible, I didn't even try.
...
Ok, it will be very interesting to read your new book.

[Carl-NC]

I don't know how engineers originally came up with coil specs. Looking at really old coil designs, it appears they bought into the fantasy of transformer gain because some of them had an extremely high LRX. And then there is one of the Tesoro coils where LTX=6mH and LRX=6mH. No idea how they landed there.

I've told this story before: the White's V3 was required to use DFX coils, where LTX=540uH and LRX=15.7mH. It quickly became obvious that LRX was too high for 22.5kHz but we were not allowed to change it, so we had to add "tweak" components inside the coil to make it work. It also made for a very thick winding in the 4x6DD shooter coil that was difficult to bake out, those coils went bad quite often, sometimes just sitting in inventory. One day I was playing with an Explorer and decide to put the Explorer coil on the V3. The LTX was about the same and the LRX was 1/30th that of the V3, so I had to boost the preamp by 5.5X. The V3 had digital calibration so once I put the coil on all I had to do was re-calibrate. It worked perfectly, same performance.

[Daves]

I don't know how engineers originally came up with coil specs. Looking at really old coil designs, it appears they bought into the fantasy of transformer gain because some of them had an extremely high LRX. And then there is one of the Tesoro coils where LTX=6mH and LRX=6mH. No idea how they landed there.

I've told this story before: the White's V3 was required to use DFX coils, where LTX=540uH and LRX=15.7mH. It quickly became obvious that LRX was too high for 22.5kHz but we were not allowed to change it, so we had to add "tweak" components inside the coil to make it work. It also made for a very thick winding in the 4x6DD shooter coil that was difficult to bake out, those coils went bad quite often, sometimes just sitting in inventory. One day I was playing with an Explorer and decide to put the Explorer coil on the V3. The LTX was about the same and the LRX was 1/30th that of the V3, so I had to boost the preamp by 5.5X. The V3 had digital calibration so once I put the coil on all I had to do was re-calibrate. It worked perfectly, same performance.

Carl I read what you wrote,
1)What happens if TX and RX are 1:1?
2)What filters should be on the RX input? (After adjusting the coil values, which filters should we add to the RX input?)

[Carl-NC]

1) Nothing, that will work as well. Tesoro has a coil with LTX=6mH and LRX=6mH.

2) For a single frequency design you usually bandwith-limit the preamp to the signal of interest.