TGSL Experiments

[simonbaker]

Hi Don , Simon , Ap , WM6 and Jerry ,


Your 1M mod is sure on my list 'to do' ( for a long time ) , but first I wanted to make a basic stable TGS MD.

I live in Rotterdam Holland / the Netherlands / Ap Berg country .

Perhaps I've mis-interpretated ( is this correctly spelled ?: it's hard to write without a spellingscheck) the term 'motorboating' .

It did chatter however continuously during the 'field test' :

- I used a headset which was in my opinion far enough removed , I do remember some posts about the speaker problems.

- The psu dc/ ac convertor was turned of , as Jerry suggested and the ignition of my car was also switched of just in case ...( I whish I had a shed like Jerry , I could drive my car into it for a field test )

- I do think that those vibration influence by wind / shaking hands / or sheep on my coil are bigger as I expected so there's much to win for me overthere

- Ap Berg : ' Veel te koud hier hahaha ' : yes the leads are grounded only on 1 side , in the past I did read about the noise so I wanted to prevent the pcb injecting noise into the leads ( from pcb to pots ) perhaps there is also something to gain overthere , also on my list 'to do'.

- As Simon suggested to shorten the Tx coil : I do get then about 4 - 6 mV ac ...so my quest for the noise source continuous and I will post my findings
I was waiting to pot my coils : I just wanted to make sure that my coils are good enough and the nulling and phase shift is good enough , that's why they are partially glued down with ( I hope ) removeable hot glue

I must be close to the finish and barking at the right tree , at least I hope so .


regards and thanx everybody for the suggestions and feedback ,

Danny
Dennis the Mennis

Lot's of chatter is typical of unpotted coils outdoors where air currents move it.

Motorboating is a more of a steady beeping. I can get it under power lines, or with certain sensitivity settings and "dfbowers" mod.

BTW: did you include the "dfbowers" mod in your PCB (sensitivity threshold range increase)?

I find with the dfbowers mod, I can make "motorboating" if I turn the sensitivity to max, but that is not unexpected. It depends a little on your LM308 chips whether it will happen also.

-SB

[simonbaker]

Hi TGSL friends,

Induction balance coils are a pain to build. You can do everything right, then, while potting, a minute resin shrinkage can upset the balance.
It is much worst in the field. With your perfectly balanced coil you hit an area with highly mineralized soil. The Induction Balance is upset and you loose a lot of depth and sensitivity or the detector may not even work at all.

So what can be done about that?

Working on my Hybrid detector I ran into the same problem. I have found a solution that seems to be the answer. I can now adjust the Induction Balance of the coil in the field. A simple 25 turns pot is the interface. Eventually I intend to change this pot for a digitally controlled pot. Then the adjustment can be automatized with the MCU.

The system is still very crude but it looks very promising. Probably it would also work with a VLF, but I really don't know anything about the functioning of VLF's, so I need your input.

It will take some experimenting.

Anybody willing to give it a try?

Tinkerer

This is an interesting subject. It may be different for IB phase detectors (like TGSL) than for IB PI detectors (yours is the only one so far?).

For TGSL, null point does not change the target signal phase, only the null signal phase. However, the null signal phase can be important for ground balancing. But we have the GB pot adjustment if the null phase shifts.

I am beginning to think there are two main components to null signal phase -- 1) capacitive coupling (cable, coils), and 2) magnetic coupling (mainly shields).

My feeling is that changes to null signal phase due to capacitive coupling are not important -- don't worry about them. That's because ferrite (and ferrite-like ground) will not modulate capacitive coupling, only magnetic coupling. However, I need to investigate that thoroughly before I believe it.

For PI detectors, signal shape seems to be critical for detection algorithms. Signal shape may be more sensitive to slight changes in the coil couplings than TGSL target phase. I don't know.

If I were making a PI detector, I would want to use a computer chip so I could normalize the signal as much as possible to remove effects of calibration problems such as shifting coil nulls, distance to ground, etc. It sounds tricky, which is why I'm not ready to tackle it. But interesting problem.

-SB

[Dennis the Mennis]

The Phase shift of Tx and Rx in relation to the thickness of the shield isolation ?

Experiment with 4 scope image clips added:

The Phase shift of Tx and Rx in relation to the thickness of the shield isolation :

* DD Coil ( model Ivconic '09) parameters : Tx = 5.9 mH and Rx = 6.5 mH
* Default values for capacitors .
* Scope images are measured on output of the LF353 ( pin 7 U101a) .

Why didn't I get the desired (aprox) 20 degrees phase shift between Tx and Rx during nulling with a reasonable residual voltage ?

I've tried a capacitor bank for tuning both Rx and Tx : didn't get the 20 degrees , not even close .

What the heck was going on .... The sheep curse ?

Recently I've tried a thicker isolation between the coil and shield with a great result :
I do get the 20 degrees ( aprox ) phase shift .

For isolation before I've used only a small 3M black isolation tape between coil and shield and sometimes no tape at all but the shield direct on the coil.

I've used a stripped RG59 coax cable , perhaps too thick but it was worth the experiment , anyway now I did get ( almost ) the 20 degrees with just the normal capacitor values for Tx and Rx.

Calculation for usec for 20 degrees = ( 1/ 14000 ) * ( 1/20 ) = 3,57 usec

In clip 1 both coils are unshielded : http://www.youtube.com/watch?v=3hY4s6_cKw4

- Almost 'perfect' phase shift and I used these coils only for test purpose .
Even when the residual voltage rises or lowers the phase shift remains almost the same.
　
In clip 2 only the Tx coil is shielded : http://www.youtube.com/watch?v=zZTDKBd3Gn0

- The lower the residual voltage the bigger the phase shift.
　
In clip 3 both coils are shielded : http://www.youtube.com/watch?v=iJKx8JJ1A8M

- With Aluminium kitchenfoil direct on the coils .
Even more phase shift when the residual voltage becomes lower.
　
In clip 4 both coils are shielded : http://www.youtube.com/watch?v=N7yrp8l_fSU

- BUT between the coil and the shield I've placed some isolation :
I've stripped a RG59 coax cable and used this as isolation between the coil and shield , perhaps way to thick ... ? Try for your self ...
The first part of the clip are the images of the Rx coil direct vs the output the LF353 ,
the second part as shown in the other clips Tx vs output LF353.

Conclusion ( for this pcb and coils ?):

When your unable to reach the 'desired' and stable phase shift and your pcb components have all the standard values : first try the coil isolation ....or keep it in mind at least.

kind regards

Dennis the Mennis


btw I've used a kind of mylar as shield ( stripped from a VGA cable shield ) for the last two coils in clip 4 .... you don't think that this was the solution ..... 　

[lunamay]

hi everyone
i'm still trying to perfor my tgsl but still poor deapth. got few questions to ask.
firstly mayby stiupid question...but is all pots in circuit (i mean sens geb and disc) are lineal or logarytmic:P? i'm asking because i have a strange behaviour when i turn them up or down...also they induct a lot of noise..maybe shealding connection will help

another thing... is there a new nice way to check the coil without a scope? i have only good multimeter without scope function and i hear that voltage drop on rx should be as close to "0" as possibe (maybe i said something wrong but it's a lot of posts to reread:P)

i would also ask for dimensions of coil plastic chasis if someone has the, because i would like to build my own:P

sorry for mistakes..

best regards to all

[dfbowers]

hi everyone
i'm still trying to perfor my tgsl but still poor deapth. got few questions to ask.
firstly mayby stiupid question...but is all pots in circuit (i mean sens geb and disc) are lineal or logarytmic:P? i'm asking because i have a strange behaviour when i turn them up or down...also they induct a lot of noise..maybe shealding connection will help

another thing... is there a new nice way to check the coil without a scope? i have only good multimeter without scope function and i hear that voltage drop on rx should be as close to "0" as possibe (maybe i said something wrong but it's a lot of posts to reread:P)

i would also ask for dimensions of coil plastic chasis if someone has the, because i would like to build my own:P

sorry for mistakes..

best regards to all

I don't think that it matters if your pots are linear or log taper, the adjustment position will be different. I have used both successfully. Pots that are incorrectly adjusted can cause all sorts of strange things. Once I get mine setup right, things work great. Particularly, the GB pot will cause the TSGL to work incorrectly or not work at all.

As far a voltage drop being close to "0" are you referring to nulling?
I check my coils first with a scope. I hook them to a small signal generator and I can determine inductance that way. If you are interested, I can show you how to build a small signal generator from a 555 timer and a handful of components..Very useful for cheap!!

Also, I think I have pretty much nailed the optimum coil shell size. This comes out to be 13 ounces at most. These dimensions are for the INSIDE so it's the size of my wooden former. The distance between ears is 19mm (fits Tesoro rods perfectly). The dimensions below are for Ivconics coils in the "TGSL coil making" .pdf . I also use the attached pattern with a perfect fit.


If you cannot build one I may be able to provide you with one.
Don

[lunamay]

thank You !!

i'have seen your work at youtube and You got nice deapth.

i haven't got any scope but i mesure inductance with my multimeter. unfortunatly i can't mesure the fase shift and see what happens... i was thinking what should i do to calibrate the coil (best distance beetwen rx and tx ) to have max deapth.. also have a problems with noise induction in cable..when i tough it it seems to make a noise.. i checked connections and i'm ussing 4 wire shealded cable...another thing.. what do you think to make a sheald for coil using a aluminium foil from electrolitic capacitors..it is conductive and it's easy to find. I was looking for aluminium tape but i couldn't find it. I have a lot of work with coil..it's my secound but deapth for coil about 9-10 cm is not good;/

soon i will post some pictures

best regards

[simonbaker]

Aziz's latest circuits for his laptop VLF showed me the simple circuit that is like nudging a pendulum that I was looking for.

Well, close -- it is not part of an oscillator, but is driven by any oscillator output (with low impedance I assume).

What's important, if I'm not mistaken, is that it takes advantage of high-Q TX coils, where the TX voltage rises way above the driving voltage, and I assume that provides a higher coil current accordingly.

Of course the issue of whether such a high-Q TX coil makes more difficulties than advantages remains (ref: mikebg), but it is at least a simple circuit to play with if you can find a nice way to drive it (relaxation oscillator?) -- or build it into an oscillator.

On the other hand, maybe the easiest way for us to utilize high-Q coils with the TGSL is a center-tapped coil.

-SB

Aziz - I thought I would study this "high voltage" TX coil circuit again. I was wondering if you had ideas on optimal values for the capacitors for a given frequency (it seems the resonant frequency depends on C3 + C4, so how do we allocate the values?). I would assume the goal is to maximize the current in the coil.

I also have been comparing it to a simple RLC series circuit with basically same values (C = C3 + C4), driving it with the same sinusoidal voltage source. Surprisingly, I get a much higher voltage and current at the coil.

However, what also matters also is the current drain on the "battery" -- which is not modeled explicitly in my Spice circuit. The idea was to simulate the "nudged pendulum" analogy of a resonant circuit to get a high amplitude current with very little steady-state input.

Wondering what your thoughts were on it. Which circuit really is better to take advantage of a high Q TX coil to get the most magnetic field for least battery drain?

-SB

[dfbowers]

Aziz - I thought I would study this "high voltage" TX coil circuit again. I was wondering if you had ideas on optimal values for the capacitors for a given frequency (it seems the resonant frequency depends on C3 + C4, so how do we allocate the values?). I would assume the goal is to maximize the current in the coil.

I also have been comparing it to a simple RLC series circuit with basically same values (C = C3 + C4), driving it with the same sinusoidal voltage source. Surprisingly, I get a much higher voltage and current at the coil.

However, what also matters also is the current drain on the "battery" -- which is not modeled explicitly in my Spice circuit. The idea was to simulate the "nudged pendulum" analogy of a resonant circuit to get a high amplitude current with very little steady-state input.

Wondering what your thoughts were on it. Which circuit really is better to take advantage of a high Q TX coil to get the most magnetic field for least battery drain?

-SB

Hey Simon


I have been looking at that basic Tx scheme for a long time, and seems to be popular with many of the designs on www.md4u.ru. KT315 may be a good one to ask as well as it must be his Tx method of choice. That's pretty much the Tx section of the Volksturm and apparently works well. The only thing odd that I have observed is that when driven by a square wave, the shape of the sine wave doesn't come out very clean, so may have some waisted energy other places than the fundemental frequency. Maybe drive the coil from a sinusoidal source?

I had posted this question before:

http://www.geotech1.com/forums/showthread.php?t=17602

The other big question I have when it comes to this is impedance matching of the coil to get the most energy transfer.
Don

[simonbaker]

Hey Simon


I have been looking at that basic Tx scheme for a long time, and seems to be popular with many of the designs on www.md4u.ru. KT315 may be a good one to ask as well as it must be his Tx method of choice. That's pretty much the Tx section of the Volksturm and apparently works well. The only thing odd that I have observed is that when driven by a square wave, the shape of the sine wave doesn't come out very clean, so may have some waisted energy other places than the fundemental frequency. Maybe drive the coil from a sinusoidal source?

I had posted this question before:

http://www.geotech1.com/forums/showthread.php?t=17602

The other big question I have when it comes to this is impedance matching of the coil to get the most energy transfer.
Don

In my simulation with the square wave driven circuit, the current signal through the coil looks nice and sinusoidal, although the current through the resistor has a spike (which may not be bad). The current spike may be the pendulum "kick" that I was looking for - an efficient way to keep the pendulum going.

I'm thinking that maybe a square wave (with possible duty cycle tweaking) may end up being the most efficient way to drive a TX coil. Even though a sine wave may have a better "transfer" efficiency, creating a high power sine wave (which typically takes class A or AB amplification) may be quite inefficient and you lose much more than you gain. Square wave pulses basically can be created with class C or D amplification which is much more efficient. It may turn out that the higher frequency harmonics of the square wave do not represent power loss - they simply are attenuated by the RLC tank and contribute no current.

The impedance matching question is interesting. I often wondered about putting a transformer between the TX coil and the oscillator, but that might be the same as a tapped TX coil with the tap at a judicious location - also a design consideration. For square-waves, it might mean increasing the square-wave voltage until the law of diminishing returns sets in.

It would be useful to model a battery driven sine generator and pulse generator and then graph the power actually being taken from the battery for a given TX current. That might help put the design tradeoffs in perspective.

-SB

[simonbaker]

In my simulation with the square wave driven circuit, the current signal through the coil looks nice and sinusoidal, although the current through the resistor has a spike (which may not be bad). The current spike may be the pendulum "kick" that I was looking for - an efficient way to keep the pendulum going.

I'm thinking that maybe a square wave (with possible duty cycle tweaking) may end up being the most efficient way to drive a TX coil. Even though a sine wave may have a better "transfer" efficiency, creating a high power sine wave (which typically takes class A or AB amplification) may be quite inefficient and you lose much more than you gain. Square wave pulses basically can be created with class C or D amplification which is much more efficient. It may turn out that the higher frequency harmonics of the square wave do not represent power loss - they simply are attenuated by the RLC tank and contribute no current.

The impedance matching question is interesting. I often wondered about putting a transformer between the TX coil and the oscillator, but that might be the same as a tapped TX coil with the tap at a judicious location - also a design consideration. For square-waves, it might mean increasing the square-wave voltage until the law of diminishing returns sets in.

It would be useful to model a battery driven sine generator and pulse generator and then graph the power actually being taken from the battery for a given TX current. That might help put the design tradeoffs in perspective.

-SB

Following up on subject of coil-driving circuits, I did very informal comparison with LTSpice, and here are some impressions:

I compared simple series RLC circuit driven by sine wave with a series/parallel circuit driven by same wave.

Rv represents the impedance of the voltage source.

The coil resistance is 7 ohms, chosen because that is a practical value you can get with 24 gauge wire. Coil inductance is 6 mH.

Capacitor value Cx is .02uF. Cxa + Cxb = Cx.

The ratio Cxa/Cxb was varied from .1 to 1.0 to see what is best.

The voltage source impedance Rv was varied from .1 to 400 ohms to see the effect.

Conclusions:

The ratio Cxa / Cxb = .1 seems to give maximum current in the parallel tank coil. Not sure if other dependencies.

When Rv is less than 75 ohms, the series RLC circuit has higher coil current, but less Q in the frequency response.

When Rv is more than 75 ohms, the series/parallel circuit has higher coil current and greater Q in the response.

Overall conclusion - the series/parallel circuit probably excels for power efficiency, giving you best coil current and battery life trade-off.

If you have a low impedance voltage source, then the simple series circuit can produce a higher absolute coil current. However, maybe adding an impedance matching transformer could make the series/parallel circuit operate at a higher current point. More studies needed.

-SB

[simonbaker]

Coil resistance at resonance

Following up on subject of coil-driving circuits, I did very informal comparison with LTSpice, and here are some impressions:

I compared simple series RLC circuit driven by sine wave with a series/parallel circuit driven by same wave.

Rv represents the impedance of the voltage source.

The coil resistance is 7 ohms, chosen because that is a practical value you can get with 24 gauge wire. Coil inductance is 6 mH.

Capacitor value Cx is .02uF. Cxa + Cxb = Cx.

The ratio Cxa/Cxb was varied from .1 to 1.0 to see what is best.

The voltage source impedance Rv was varied from .1 to 400 ohms to see the effect.

Conclusions:

The ratio Cxa / Cxb = .1 seems to give maximum current in the parallel tank coil. Not sure if other dependencies.

When Rv is less than 75 ohms, the series RLC circuit has higher coil current, but less Q in the frequency response.

When Rv is more than 75 ohms, the series/parallel circuit has higher coil current and greater Q in the response.

Overall conclusion - the series/parallel circuit probably excels for power efficiency, giving you best coil current and battery life trade-off.

If you have a low impedance voltage source, then the simple series circuit can produce a higher absolute coil current. However, maybe adding an impedance matching transformer could make the series/parallel circuit operate at a higher current point. More studies needed.

-SB

Still studying this series/parallel circuit for use as coil driver. Some of my previous comments were over-simplified in terms of best ratio of Cxa to Cxb. It is a trade-off involving the battery current and the coil current, and depends a lot on the internal resistance of the power supply (battery).

Anyway, I said what the heck, I like the series/parallel circuit, so I thought I'd breadboard one and see how it compares to the spice simulation.

I'm driving it with my signal generator using a square wave from zero to 3 volts at the resonant frequency (I can tune for max coil voltage). My signal generator has internal resistance of 600 ohms.

For my "production circuit", I drive a "push-pull" transistor pair connected to a 3 volt battery supply and drive the resonant circuit with that. A LTSpice circuit is shown.

As an independent test, I drove the resonant circuit directly from my signal generator, to estimate the coil resistance at resonance. A second LTSpice circuit is shown representing that (no transistors).


Findings:

I found the real-world coil voltage is a lot less than the simulated voltage (see photo). I used a very low-resistance coil (24 awg wire), about 6 to 7 ohms, with 5.9 mH inductance.

Simulated: 45 V pp.
Real World: 13 V pp.

I'm wondering why the results were so different. So I speculated that maybe the coil resistance at resonant frequency is significantly more than the DC resistance. To test this, I ran separate simulations with a simpler circuit (no transistors), increasing the coil resistance in steps, then see when the coil voltage matches my real-world results (also without transistors).

With a 7.6 ohm coil, my simulation indicates the coil acts like it has about 27 ohms resistance at resonance.

With a 25 ohm coil, my simulation indicates it acts like it has about 42 ohms at resonance.

Basically this "sucks" because it means that you can't get as much coil current as you would like by using a low resistance coil and this driver circuit. The simulations looked exciting -- using the full driver (push-pull transistors), they implied you could get about 45 V pp from a 3 volt square wave, where I actually get about 13 V pp.

That doesn't mean it is a worthless circuit, but not as exciting as we'd like. Of course you can use the usual methods to increase current, such as a higher voltage source.

But I'll keep checking in case there is something I'm doing wrong with the breadboard circuits.

Does anyone know theory why the coil resistance at resonance is 400% more than the DC value?

-SB

PS. I did a lousy job labeling the signal graphs. The top trace, V(n006), is measured at the transistor collectors. The bottom trace, V(n008 ), is the coil voltage (non-grounded terminal). The simulation and the oscilloscope photo show the same signals.

[simonbaker]

Coil resistance at resonance

As mentioned in previous post, my 7 ohm coil seems to act like a 27 ohm coil at resonance (about 13 Khz).

Where does the extra resistance come from? Is it radiative losses? Is it "skin effect"? Seems much more than I expected.

Does anyone have data on coil resistance at different frequencies?

Regards,

-SB

[simonbaker]

Coil resistance at resonance

I tried an experiment which may have been foolish, comments welcome.

I wanted to test if the "skin effect" is causing coils to exhibit higher resistance (lower Q) at resonant frequency than predicted by the DC resistance -- something that should be improved by using "Litz" wire, theoretically.

I don't have Litz wire, so my (quick & dirty) experiment was this:

I stacked 4 coils in parallel, as close as possible. Each coil had about 25 ohm resistance and about 6.5 mH inductance.

The combined stacked coils, connected in parallel as a single coil, measured approx 6.9 ohms resistance and 4.46 mH inductance.

I then put the stacked coils into a resonant circuit (about 16.7 kHz) and compared the actual coil voltage with an LTSpice simulation. (The simulation used 6.9 ohms for the coil resistance and 4.46 mH for the inductance.)

I expected the combined coils would act more like Litz wire, since each coil has thin wire and less skin effect.

I was wrong. The results were just the opposite.

The combined coils showed an even less-than-expected voltage that was consistent with an even higher-than-expected resistance of about 90 ohms instead of 6.9 ohms. That is 1300% higher resistance than the DC resistance.

This is a worse agreement than for the single 7.6 ohm coil using 24 AWG wire, which acted like about 27 ohms (355% higher than DC resistance).

Conclusion: my stacked coils did not achieve a "Litz wire" effect.

It is also possible that something else is causing the lower-than-expected Q at resonance than skin effect, such as radiation, vibrational modes, maybe even bad capacitors in the resonant circuit, etc.

Whatever is going on is having a huge effect on the predicted Q of the coils, causing much lower Q than expected.

I'd like to learn more about making the highest Q coil possible; how to do it, what are the limiting factors.

-SB

[porkluvr]

As mentioned in previous post, my 7 ohm coil seems to act like a 27 ohm coil at resonance (about 13 Khz).

Where does the extra resistance come from? Is it radiative losses? Is it "skin effect"? Seems much more than I expected.

Does anyone have data on coil resistance at different frequencies?

Regards,

-SB

Would you please explain how your measure statement works? The LTspice 'Help' description is long and involved.

How exactly did you determine your coil resistance being high?

There are a few other things about your simulations that are not exactly obvious, but that will do for now.

[simonbaker]

Would you please explain how your measure statement works? The LTspice 'Help' description is long and involved.

How exactly did you determine your coil resistance being high?

There are a few other things about your simulations that are not exactly obvious, but that will do for now.

Basically I just built the circuit (see below) and compared it with the simulation, using a 3.0 V pp sine wave voltage source with 600 ohm internal resistance.

The data depends on the coil I'm testing.

For example, with 7.3 ohm, 7.45 mH coil, with resonance about 13 kHz, the coil voltage is predicted by simulation to be: 18 V pp.

My actual circuit only has about 7.5 V pp.

I thought it would be interesting to see what coil resistance would make the same 7.5 volts in the simulation.

So, in the simulation, I vary the coil resistance until the coil voltage matches the actual circuit, 7.5 V pp. The resistance necessary was about 93 ohms.

So although the coil DC resistance is 7.3 ohms, at resonance, it acts as if the resistance is 93 ohms.

I'm sorry that some of the measurement numbers in the previous posts I can't replicate exactly, (I was trying so many tests quickly) so I may have drawn some conclusions I shouldn't have. But the basic conclusion is the same -- my actual circuit resonates at a much lower voltage than the simulated circuit. I'm not sure why and I would like to find out the answer.

I'm using a solderless breadboard for the circuit, so perhaps poor contacts or something could be involved????? Or this is just an obvious behavior of real coils???

Regards,

-SB