TGSL coil + audio problem

[Davor]

At the very moment I have a simple "true balanced op amp" configuration, which is not much of a pre because of not-so-good common mode performance, but it provides equal loading of the coil halves. Obviously a centre tap works it's miracle. It is also very noisy, and input referenced noise is ~10nV/sqrt(Hz). Point is that most of the "lightweight" detectors also have a mid range op amp at their inputs and all of them have very modest noise performance.

I'm planning to replace the existing pre with something truly balanced and very low noise. Funny thing is that BC337 will, given the chance, go below 1nV/sqrt(Hz) because of their Rbb being at only about 30ohm. In balanced configuration, and keeping the components and currents the same, the input impedance is doubled. Hence, the noise is sqrt(2) bigger than the single ended configuration, however, balanced configuration is worth it.

I'll go with this circuit ...

[Xavier]

If you leave your coils unshielded, you will have all kinds of problems when you go to use it. Even a blade of grass touching the coil can cause a false signal, especially if the grass is wet.

I'm going to agree and disagree with you on this one Don. I see no purpose what so ever in shielding a TX coil, on the other hand shielding the RX coil makes a lot of sense.

Hi Davor

Thanks for sharing that with us I will give it a try.

[dfbowers]

Yes, a shielded Rx coil does make more sense. Give the unshielded Tx coil a try and let us know how you make out. But, it's impossible to go backwards though once you pot your coil!! And.. potting epoxy is not cheap, not to mention the week or so you put into building your coil. My suggestion is to not cut any corners.
I have built almost 20 DD coils now, and I only threw away the first 2.

Best practice is to shield both unless your are using a driven Tx. This is not the case with the TGSL.

[Davor]

I'm going to agree and disagree with you on this one Don. I see no purpose what so ever in shielding a TX coil, on the other hand shielding the RX coil makes a lot of sense.

Hi Xavier,

I'm generally in favour of unshielded coils, but ... shield is not as bad as I initially thought it would be. In a VLF coil it would shift a phase by a few degrees, and the most important thing about it is that both Rx and Tx are shielded, because that phase shift gets compensated. I made some simulation that I will not share here simply because it would require too much time explaining what it shows, but a short version is that you are better off if both coils are shielded, or both unshielded, but not the cross breeds of these.

Shielded coils need less care and feeding, but I believe it is worth pursuing the unshielded path.

BTW, I know you are quite active at thunting - your avatar is different from the others.

[6666]

Hi Davor
thanks for sharing your work, that's very kind of you.
what type of capacitor is C3 ?
3x 1u ?

[mikebg]

MORE SIMPLE CIRCUITS

There are more simple circuits with LM386, NE592 or uA733 for input symmetry.
http://www.geotech1.com/forums/showp...67&postcount=2
http://www.geotech1.com/forums/showt...745#post108745

Despite written by me against the shielding in that threads, I use shielded search head housing to avoid variable mutual capacitances between RX and TX networks. They are variable because movement of operator and search head movement relative to ground.
LM386 no need of grounded coil lead for bias, but NE592 and uA733 need grounded to intermediate voltage for bias:
http://www.geotech1.com/forums/showt...782#post145782

Here is the circuit with input symmetry reinvented by revered Thomas Scarborough:

[Davor]

Yes, this is a fine configuration and it's a pity LM386 has input referenced noise well over 100nV/sqrt(Hz). For a simple configuration you'll hardly find anything so simple and right to the point. But there is noise

[Xavier]

Hi Xavier,

I'm generally in favour of unshielded coils, but ... shield is not as bad as I initially thought it would be. In a VLF coil it would shift a phase by a few degrees, and the most important thing about it is that both Rx and Tx are shielded, because that phase shift gets compensated. I made some simulation that I will not share here simply because it would require too much time explaining what it shows, but a short version is that you are better off if both coils are shielded, or both unshielded, but not the cross breeds of these.

Shielded coils need less care and feeding, but I believe it is worth pursuing the unshielded path.

BTW, I know you are quite active at thunting - your avatar is different from the others.

I hadn't thought of the phase shift in both coils. I have noticed that there is about 100uH difference once the coil was shielded, of course this will vary with the type of shielding you use, in my case it was ordinary tin foil. I will use Don's way for making my next coils (Varnish as I wind) once dry I will see what the inductance is but it should have come down a bit and that's OK because the shielding will bring it back up again. I found some 2n4392 in my junk box which I replaced for the 2n4391 that I had in the detector and there is a definite improvement but I'll have to get some 2n4393 as these are not going to do the job. Now to make a new coil former first then turn out some new coils LOL

thunting.com is my relaxation therapy after a hard days work, also a motivation for me to make good working detectors. The one bad quality that I have is I do not except failure.

All the best
Xavier

[Davor]

The phase shift I am refering to is a consequence of eddy currents in the aluminium foil shielding. I did not try it in simulation yet, but instinctively I expect some resistive shielding would be a better choice as there are not much eddy currents.

[Xavier]

Has any body ever used carbon fiber in their coils?

[dfbowers]

Has any body ever used carbon fiber in their coils?

Hi, Xavier. I did experiment with carbon fiber at one time as an epoxy filler. I abandoned the idea because it's conductive! As long as you do not form a complete loop you may be OK.

[Xavier]

Hi Don

I read somewhere that you could make the coil shell out of carbon fiber, it would be an expensive shell though (here in SA it's very expensive) and like you say it's conductivity is a problem even with fishing rods. If the shell were open you would probably not need any shielding and this would make life simpler but I can't help to wonder at what cost?