DD or OO Coil Overlap Calculation

[hobbes_lives]

Does anyone know if there is a calculation to approximate the overlap of DD and/or OO coils to obtain null? I imagine it could be a calculation based on a percentage of the flat surface area of the coils, or perhaps a percentage of the distance between the magnetic center of the coils, or ??? I think it would be helpful when calculating the size of coils for a particular housing, and vice versa.

[WM6]

Pretty impossible, except in very coarse approximation, because coils are not the same, coil shape are not the same, coil overlaping are not the same, Tx frequencies are not the same, coil shielding are not the same, etc.

For some idea better to use picture or draving of tunned coil than calculate percentage.

[Jose]

DD coils on the machines PI, you have to cancel to 0 volts?
When I try to ring to "0 volts", the sensitivity is very poor.
Setting the cancellation of 100 or 200 mV, is observed decay curve and the TX pulse and the sensitivity is correct.
Okay this, or is there something wrong with the construction of my coil?
Thanks
Jose

[Skippy]

I imagine if you tried a quick scaled-down mock-up of your search coils the overlap could be easily found. Then scale up the results for the real coil. Drive the TX with a sig-gen, look at the Rx output on a scope, move the coils for a null. You'll need to shield the Rx at least, but aluminium/aluminum foil tape should be ok for a simple test like this. Hope this helps.

[IBGold]

To give you an idea on a 410mm O.D. DD coil the overlap on the coils centerlines is 32 mm after a time you get a feeling on what it should be but it is still trial and error testing.

Regards, Ian.

[Jose]

Thanks for the replies.
DD coils, 37 cm, Blindage Scotch tape 24.
Continue experimenting.
Regards, Jose

[Qiaozhi]

The reason you might want a rough calculation for the coil overlap is when you come to make the coil former. At the beginning you only know the overall diameter of the search head. But how much bigger do you make each D-shaped coil so that it will fit correctly into the shell?

Assuming the required area of the overlap is in some way related to the area of the entire search head, let's do a simple calculation:

Let be the area of the search head:



and be the area of the overlap:



(where r is the radius of the coil shell, and x is the width of the overlap)

Also, assuming k is a constant that defines a linear relationship between
and , we have:



Hence:



Using the real measurements supplied by IBGold, we have r=205 mm and x=32 mm.
Substituting these values into the equation gives a value for k of approximately 0.1.

Re-arranging the equation for x, and checking the result using the values for r and k:


If we agree the constant k can be fixed at 0.1, then the equation can be simplified to:

<---------- final equation

(where d = the diameter of the coil shell)

Performing a second test for a 285 mm diameter coil gives:



and for Jose's 38 cm coil we have:

[Skippy]

I think simple linear scaling up/down of a given searchcoil dimensions is all you need to do. Changes in operating frequency, number of turns, wire gauge etc should matter very little.

[hobbes_lives]

Qiaozhi is right on with what I'm trying to calculate. Perhaps if we can get some more real-world values, we can determine if the calculations will work as an approximation, and narrow k down to a relatively precise value. Nice work Qiaozhi.

The thing that throws me off is a coil like the Sunray S-12. The overlap looks like it is very narrow.

[Carl-NC]

The amount of overlap depends on the exact geometry of the coils. Induction balance occurs exactly where the total TX flux cutting through the interior of the RX coil equals the total TX flux cutting outside the RX coil. This is best solved by modeling software (like what Aziz has written) but practically can be empirically determined and probably broadly estimated.

Five case are pictured below. The easiest case is for a OO coil (A), where the overlap can probably be figured up mathematically on paper. Coil B is an elliptical 00 coil, and the overlap will depend on the eccentricity. An interested coil designer could surely find a relation between eccentricity and overlap empirically.

Coils C-E are DD, but with minor variations. The tightness of the interior bends will matter (C vs D), and bowing the overlaps (E) will also matter. However, the amount they alter overlap might not be significant, I don't know. What's not shown (F) is the elliptical DD coil where overlap will again depend on eccentricity.

Bottom line is do what Qiaozhi did... use a known working coil to estimate similar design geometries.

- Carl

[der_fisherman]

Good diags.

I am pretty sure that even when building coils on a DIY basis, the overlap will still be within probably 1mm of the same place each time.
Assuming same MD, same frequency, same method of making the coils (same former) etc...
What I have never actually tried is to see if the null stays the same over a range of frequencies, I would imagine that within the area of frequencies that we commonly use for MDs, that a null would be the same.....Can anyone comment on that please, someone who has done just such an experiment?

regards

Andy

[Carl-NC]

At high frequencies capacitive coupling will move the null point around. A clever designer can compensate for this. I've designed a coil which holds a pretty deep null from 1kHz to 50kHz, without compensation.

[satdaveuk]

I imagine if you tried a quick scaled-down mock-up of your search coils the overlap could be easily found. Then scale up the results for the real coil. Drive the TX with a sig-gen, look at the Rx output on a scope, move the coils for a null. You'll need to shield the Rx at least, but aluminium/aluminum foil tape should be ok for a simple test like this. Hope this helps.

Hi Skippy
Actually doing it your way can get you near as dam it on setting up DD coils I normally do it that way then tempory fixes then the final phase shift to complete the job.
Anyway nice to see your still around mate
Regards
Dave

[der_fisherman]

At high frequencies capacitive coupling will move the null point around. A clever designer can compensate for this. I've designed a coil which holds a pretty deep null from 1kHz to 50kHz, without compensation.

Who uses high frequencies for MDs?
regards
Andy

[Skippy]

50 kHz would be good for nugget-hunting, and would also find 'micro-jewellery' well. I imagine a broadband null would be needed for multi-frequency machines, which, even if they don't receive all the frequencies, still transmit them.