Some technical questions about VLF and Coils

[fluxx]

Hi,

There's probably answers to my questions somewhere, but I haven't been able to find them anywhere. If someone could point me to places where I could find them, or simply answer, that would be great.

Ok, first questions about coils.

1. Why is the RX coil shifted in resonant frequency to the TX coil. Shouldn't it be tuned to the same resonant frequency as the TX? If the RX coil has a big Q-factor, won't the signal be totally filtered? Is it because of the needed phase shift for the discriminator? How accurate must the resonance be? What's the mechanism here?

2. I made a DD coil for my IDX-PRO (DIY version), but my shield was made of aluminum foil. I was able to null it with a scope, but the sensitivity was terrible. I then removed the shield and the sensitivity is even worse! When I put a piece of metal, the RX amplitude doesn't change much at all, unless I get the metal really close. Moreover, the iron is more responsive than aluminum. Is the amplitude of the RX signal the only way to detect metal. The way I understand, the coil is in balance and the RX is zero until a metal comes along. That breaks the balance and the RX jumps to non-zero value. I have matched the inductances (more or less), tuned the capacitors to right frequencies, and it simply doesn't work. The RX signal barely changes when I bring a metal object to it. It changes more if I move the coil, which throws it out of balance, but metals just don't do much.

3. How is shielding made with graphite different than say, aluminum foil wrapped around everything? I'm guessing it has to do with graphite being non-detectable by the coil, as it has low conductivity. Is this correct? Does the graphite not affect the depth of detection?

4. How important are the exact inductances of the coils if the caps are tweaked so that the resonant peak is at the right frequency?

5. How important are the resistances of the coils? I'm guessing they should be as low as possible, is that correct?

6. Is there a place I could read about coils, so I could make them better. I'm begining to think this one is not going to work no matter what I do to it, but it's killing me as to why that is. I don't want to make another one that doesn't work and basically do trial and error until I get it right and finally give up and just buy one.


That's my questions so far, I hope I don't bother someone for answering the same questions all over again, but if you can point me to where this has already been answered, I would be very grateful.

Here is my experimental DD coil for IDX: http://i.imgur.com/QjJmm1X.jpg if it looks bad, it's because I've already tried lots of combinations and none work. It is now just loops of wire and cables, but I did make it properly, with shielding and everything else.

[kt315]

try to order INSIDE THE METAL DETECTOR copy. this is a NICE BOOK to read at first, for from a cooker experimenting increasing power supply voltage
in Surf PI trying to encrease the depth to doc guru with PHD in the electronics.
to start from something is primary step.... later with more knowledge all your q's will be out themselves.

Anatloy

[fluxx]

Thanks for the answer. I did some googling, it looks like it's a pretty good book. I have no doubt it is worth the money it costs. Although for my standards, it is definitely not cheap. I will try and get it, but until then, I would like to know some answers to my questions, so I can go on with the project. Any help would be appreciated. Thanks.

[Silver Dollar]

Coils are tricky. I haven't been successful yet either and recommend buying rather than building.

There are a few things I have learned.
TX wants the lowest resistance for more power but your oscillator has to be able to drive it.

RX sometimes likes high resistance having something to do with "Q". I think we want a broad
tuning to allow for no phase change in the signal across different conditions. Likewise the
purposeful detuning of TX to RX frequency is to avoid phase changes in the RX signal.

Shielding is a tricky thing too. Aluminium shield also kills the signal when I try to use it.
I've had some better results with mesh screening and great luck with graphite. The problem
is lots of metal in the shield causes capacitance and is a large constant target for the coil.

I have been able to use a coil originally designed for 6 khz on myBandito at 10 and even 15khz
so I do not believe inductance has to be spot on (at least for Tesoro designs). I have great trouble
getting and keeping the null and have noticed if it is too good the coil is "dead". At the perfect
null the whole circuit resonates and becomes very sensitive but like I say finding that point and
keeping it while the epoxy dries is a challenge.

I believe some of the ease of which the factories make coils is from perfect windings and precise
locations of the coils. I've been thinking of experimenting with electronic tuning to see if it might
make things easier...

[fluxx]

Silver Dollar, this is very helpful to me. You seem to have experience with building your own coils and that's exactly what I want. Tell me please, as I am a beginner in the whole metal detecting thing - can you test the coil effectiveness simply on oscilloscope? Does the amplitude of the RX signal change drastically when the metal is present? Also, can you test the coil without shield like this - provided of course that you nulled it and it is static, ie not moving? Because, that way, my current situation is rather bad, I can only detect big, iron object at few cm air depth, which is really bad. I seem to have had better results with aluminum foil wrapped around windings. I am currently using the PC soundcard as oscilloscope, but I do have the real oscilloscope available as well. In my opinion, the PC is just fine for this, and you can actually hear the 6.59 KHz sound changing in velocity. I use a software oscilloscope that is supposedly able to show 1mV/Div, but I'm not sure how accurate is this. I can achieve a null of a couple of mV pk-pk like this (although, as you said, hard to stabilize and maintain). Still, the sensitivity to metal, at least visually is a few cm at most. Is this normal for the homemade DD? What should I expect in a best case scenario?

Sorry for a long post, looking forward to your answer.
Cheers!

[Davor]

For VLF you need coils rigid. While you might not see a target in proximity with a scope, the Rx will see it even a few decades below any visible indication on a scope. Provided it is well balanced. From what I've seen from a photo, you are not even half way there yet.
There are a few documents on the forum that describe how to make a working coil, so just follow what it says there. Aluminium foil is mostly OK, but you have to apply it with a gap, and also insulate it on top of foil to avoid short circuit loop between Rx and Tx.
Balancing is easy if you are systematic. You arrange coils to get some level of balancing, and fix the outer rims of coils to the base with krazy glue. Then you squeeze here and stretch there to get a better balance, apply more krazy glue at "D" corners. Then fine tune and fix, inch from the corners to the centre. Fine tune and fix 2 inches from corners. When you reach the middle, the coils are already very rigid, so deforming them to reach a better balance is a bit hard. Now, LEAVE it be till tomorrow so that the wires relax, and do final tuning then. You don't have to leave a tuning curl - as manufacturers do - if you are systematic.

[fluxx]

Davor, thanks for your answers. What you describe, I have already made (using aluminum foil, leaving a gap, insulating between coils). I did have the (visible) changes in amplitude a few cm depth, but I thought it was too little, so I thought the aluminum foil is to blame, so I removed it to test it. IMHO, it's even worse and the instability huge now. I didn't want to glue it yet, because these are all preliminary tests, but I will try as you said. I will have to get another plastic surface as a frame, as this one is not rigid at all (in fact, it's got to be the worst material ever for a coil).

That being said, my questions still stand. How much depth should I expect if I do everything by the book? How comparable is it to the factory-made one?

Thanks for the tips, keep the answers coming

Cheers!

[fluxx]

And an additional question: Davor said about some documents describing the making of a DD. I cannot seem to find them anywhere. Can someone give me a link or something?

[Qiaozhi]

And an additional question: Davor said about some documents describing the making of a DD. I cannot seem to find them anywhere. Can someone give me a link or something?

Concentric coil -> http://www.geotech1.com/cgi-bin/page...anar/index.dat
DD Coil -> http://www.geotech1.com/forums/attac...7&d=1325929634

[golfnut]

1. Why is the RX coil shifted in resonant frequency to the TX coil. Shouldn't it be tuned to the same resonant frequency as the TX?


If co-tuned the phase traversal at resonance from -90 to 90 constantly flips and sounds off - too twitchy. It will air test well though.


If the RX coil has a big Q-factor, won't the signal be totally filtered?


It will reject everthing except odd integer multiples of the resonance 3rd 5th etc



Is it because of the needed phase shift for the discriminator? How accurate must the resonance be? What's the mechanism here?


The big phase toggle from -90 to 90 will be out of harms way if the Rx resonance is safely below or above the Tx. You just loose a bit of sense.



2. I made a DD coil for my IDX-PRO (DIY version), but my shield was made of aluminum foil. I was able to null it with a scope, but the sensitivity was terrible. I then removed the shield and the sensitivity is even worse!

Aim for a null <350mV

The phase of the preamp out WRT to the Tx coil volts should look likr this for a DD.







When I put a piece of metal, the RX amplitude doesn't change much at all, unless I get the metal really close.

Thats kinda normal, a good lump close will change the level but is not practical for tuning up.



The way I understand, the coil is in balance and the RX is zero until a metal comes along. That breaks the balance and the RX jumps to non-zero value.


Yes this is right, the rate of change of signal level is important - A static large signal wont go though, ( like the Leaked Tx signal ) the filter amps after the detectors are high gain differentiators centred at 8Hz





I have matched the inductances (more or less), tuned the capacitors to right frequencies, and it simply doesn't work.


They do work..





The RX signal barely changes when I bring a metal object to it. It changes more if I move the coil, which throws it out of balance, but metals just don't do much.

moving the matal across at sweep speed will wive the same effect as moving the coil with a fixed target..



3. How is shielding made with graphite different than say, aluminum foil wrapped around everything?

Aluminium needs a gap - otherwise it sheilds not only the e feild but the H field too !



I'm guessing it has to do with graphite being non-detectable by the coil, as it has low conductivity.


Graphite is ohmic and can work well with no gaps - no large circulating currents as ohmic.




Is this correct? Does the graphite not affect the depth of detection?


If the ohmic value is very low the sense suffers.


4. How important are the exact inductances of the coils if the caps are tweaked so that the resonant peak is at the right frequency?

Not that important, I dint caluculate L values, turns etc. Be mindful you do need a bunch of turns to put out a good feild. Plus you need a solid amount of turns to Rx well.

Too few or too many gives issues.


5. How important are the resistances of the coils? I'm guessing they should be as low as possible, is that correct?

Yes, lots of R on Rx give noise - like a resistor . Tx R gives less Field and wasted battery in the attempt.




6. Is there a place I could read about coils, so I could make them better. I'm begining to think this one is not going to work no matter what I do to it, but it's killing me as to why that is. I don't want to make another one that doesn't work and basically do trial and error until I get it right and finally give up and just buy one.


may be how its wired up - show what u got..

[fluxx]

Wow, thanks, guys! Great answers. Golfnut, so you are saying that the rate of change is actually important, so I don't just measure the signal at RX and if it grows when the metal is present, it is good? I will put the shielding back on tomorrow, and I will try and null it and actually test it on an actual IDX. Just to make sure - you are saying the <350mV is ok on the actual coil? Isn't it a bit too big? Or are you saying it is at the output of the first opamp?

[Davor]

It is at the output of the preamp. Results at a bare oscilloscope are deceiving when you reach fine tuning stage, as the impedance of a probe is often higher, and capacitive couplings play differently. As lower as better, but don't expect to reach 0 because it is near impossible, and not at all necessary. It is important to have it low because of the op amp distortions that spoil detection. Your target may be at nanovolt scale, but if noise injection due to distortion is in microvolt scale, you can't reach it, and your depth suffers.
In air you can expect to detect coins up to 1.5 diameter of your coil. 1 diameter is more realistic in the ground. Everything beyond that requires some serious TLC.

[fluxx]

Thanks for the helpful answers guys! I'll try it out today and post the results! Thanks!

[kt315]

DD Coil -> http://www.geotech1.com/forums/attac...7&d=1325929634

this document has not THE MAIN - Know-How. without know-how every thing is not working (that you have see from topickiller posting). WHERE in coil design is a tuning winding? it is absent. the document is nonsence. a firm coil HAS the tuning winding because you are not able to reach the good balance without the winding.

[Qiaozhi]

WHERE in coil design is a tuning winding? it is absent.

As you showed in your diagram, it is a good idea to leave one loose turn for fine tuning. However, a lot of amateur builders don't do that, and with careful adjustments it is possible to null the coil without it. BUT ... as many have found; once the epoxy has cured, the null is quite often lost. In that case, a tuning winding is definitely recommended.