Concentric phase realationship ?

[mykaitch]

Thanks all, I already put both TGSL files in my prior post. Listened to all the advice and went and got a bit of ferrite - I knew the lead from my old glass monitor would come in handy (big ferrite ring). Broke a bit off about the size of a dime and did as said. Guess what? It worked great in fact so good that following up with the 'Tesoro pump' revealed good GB too. Even thought about making it fixed. I could only get a ten turn GB pot but all is well. Need to replace some turf in the garden so I am going to plant some coins (if SHE doesn't catch me)

Did find that the cap that couples the osc to the GB op amp (180pf) is quite important. I used a 100pf for some reason but added another (200pf now) and it was way better. Found Sense adj much better with pot low end to -v via 2k2 rather than to 0v via 200R.
Also 40M across AUTO stopped a tendacy to drift. In effect it becomes a long time constant Auto but it works well-- I have made it switchable.

[Koala]

I can't go back to a DD.

Only have to dig 8"x8" holes

Run less discrimination but still dig less iron


Performance is at least as good if not better.

[dfbowers]

Well, some hours later. Yes my trigger was correct etc. In the end I accepted that Tx and Rx were in phase. Never did have a problem with nulling BTW - 4mV seems pretty good. Slightly changed the Rx tuning - Tesoro use 33n and I used 22n which fixed just about everything even though , in theory, the res f of the Rx is 13KHz. Intermittent fault which had me scoping all sorts of things .... was a wire on the speaker .... enough said. I'll post pics when I've shrunk them a bit. Thanks for all the replies.

This is my 'ref'[ATTACH]35497[/ATTACH][ATTACH]35498[/ATTACH]

How do I delete my old uploaded files ?

I have had pretty good success with concentric coils, especially with the TGSL project. I get a bit more depth and improved pinpointing over the DD style coils, plus the audio seems more crisp. With mine, the TX lags the RX by about 20 degrees.
Don..

[mykaitch]

I don't know anything about the TGSL project. I can tell you than in an induction balance, with the thing at the best achieveable null by just moving wires around, you're eliminating the reactively (direct field) coupled component and reactive components from things like capacitive coupling and eddy currents, leaving you with a residue of resistive phase signal that's 90 degrees away from the reactive direct coupling signal. That residue will usually exhibit a lot of harmonic distortion and noise, often so much that you can't even say what phase the fundamental is. Once you've built the searchcoil and have it in use, the reactive null will drift anyhow, it's inevitable. So when you 'scope it later, what you see will not be of the amplitude and phase that you saw when you originally nulled it.

In order to determine the phase response of the metal detector, you need to introduce lots of reactive imbalance and then draw your conclusions from that. When constructing a concentric searchcoil from scratch, begin by omitting the feedback winding from the transmitter circuit and observe the receiver coil waveform on a 'scope. Of course it'll usually be in the several volts or more range, you can't operate the receiver of the metal detector that way. In the case of a searchcoil that's already built, the best thing to do is to hook it up to the circuit it was designed to be used in, and determine how the demodulated signals respond to ferrite as long as you're not saturating the receiver signal chain.

The most common convention with respect to receiver coil wiring, is to have the receiver be in approximate voltage phase with respect to the transmitter. However the reverse arrangement can also work, and there may be machines out there designed around the inverted polarity.

That' sall I wanted to know, thanks!

[mykaitch]

Finally...

Well, some hours later. Yes my trigger was correct etc. In the end I accepted that Tx and Rx were in phase. Never did have a problem with nulling BTW - 4mV seems pretty good. Slightly changed the Rx tuning - Tesoro use 33n and I used 22n which fixed just about everything even though , in theory, the res f of the Rx is 13KHz. Intermittent fault which had me scoping all sorts of things .... was a wire on the speaker .... enough said. I'll post pics when I've shrunk them a bit. Thanks for all the replies.

This is my 'ref'TGSL_101 _PART1.pdfTGSL_101 _PART2.pdf


How do I delete my old uploaded files ?

[Nupi]

So much knowledge and experience I have not but I have experimented a lot with this matter.
It really excites me, the pattern of the electromagnetic field of a DD coil is fascinating. (But how exactly does that happen?)
I did a simple measurements and a DD coil in the past of more than 120 points with a ferrite bobbin from 4mm to 2mm, to get a better picture of its radiation pattern. That was very surprising results, stunning.

But maybe you can open a topic about this, (The function of the coils for use in the metal detector). I am convinced that there are many questions and we can learn a lot from you, at least I do.
@mykaitch,thank for allowing my (off topic) comments.

[Dave J.]

Sorry, I know of no such book. In the end it's just high school physics, but in high school they didn't teach us how to apply that knowledge to induction balance searchcoil design. I've been at this for 35 years and I'm still discovering new aspects of it.

[Nupi]

I don't know anything about the TGSL project. I can tell you than in an induction balance, with the thing at the best achieveable null by just moving wires around, you're eliminating the reactively (direct field) coupled component and reactive components from things like capacitive coupling and eddy currents, leaving you with a residue of resistive phase signal that's 90 degrees away from the reactive direct coupling signal. That residue will usually exhibit a lot of harmonic distortion and noise, often so much that you can't even say what phase the fundamental is. Once you've built the searchcoil and have it in use, the reactive null will drift anyhow, it's inevitable. So when you 'scope it later, what you see will not be of the amplitude and phase that you saw when you originally nulled it.

In order to determine the phase response of the metal detector, you need to introduce lots of reactive imbalance and then draw your conclusions from that. When constructing a concentric searchcoil from scratch, begin by omitting the feedback winding from the transmitter circuit and observe the receiver coil waveform on a 'scope. Of course it'll usually be in the several volts or more range, you can't operate the receiver of the metal detector that way. In the case of a searchcoil that's already built, the best thing to do is to hook it up to the circuit it was designed to be used in, and determine how the demodulated signals respond to ferrite as long as you're not saturating the receiver signal chain.

The most common convention with respect to receiver coil wiring, is to have the receiver be in approximate voltage phase with respect to the transmitter. However the reverse arrangement can also work, and there may be machines out there designed around the inverted polarity.

Hi Dave, (maybe slightly off topic) but I have some time looking for a advanced book which dealt with all aspects of different types of metal detector coils. Do you maybe know a interresting book on this subject?
thanks

[Koala]

Don't have a problem with the null. My Q is if I look at the tx waveform and the Rx they are both exactly in phase but the TGSL project shows a screen shot with a phase shift which the article says must be there.

There is always a phase shift that alters with your bucking coil and or Rx capacitor. Are you sure you are not using a trigger on both the Tx and Rx as this will cause them to follow each other perfectly. You only need to trigger only on the TX.

[Dave J.]

I don't know anything about the TGSL project. I can tell you than in an induction balance, with the thing at the best achieveable null by just moving wires around, you're eliminating the reactively (direct field) coupled component and reactive components from things like capacitive coupling and eddy currents, leaving you with a residue of resistive phase signal that's 90 degrees away from the reactive direct coupling signal. That residue will usually exhibit a lot of harmonic distortion and noise, often so much that you can't even say what phase the fundamental is. Once you've built the searchcoil and have it in use, the reactive null will drift anyhow, it's inevitable. So when you 'scope it later, what you see will not be of the amplitude and phase that you saw when you originally nulled it.

In order to determine the phase response of the metal detector, you need to introduce lots of reactive imbalance and then draw your conclusions from that. When constructing a concentric searchcoil from scratch, begin by omitting the feedback winding from the transmitter circuit and observe the receiver coil waveform on a 'scope. Of course it'll usually be in the several volts or more range, you can't operate the receiver of the metal detector that way. In the case of a searchcoil that's already built, the best thing to do is to hook it up to the circuit it was designed to be used in, and determine how the demodulated signals respond to ferrite as long as you're not saturating the receiver signal chain.

The most common convention with respect to receiver coil wiring, is to have the receiver be in approximate voltage phase with respect to the transmitter. However the reverse arrangement can also work, and there may be machines out there designed around the inverted polarity.

[mykaitch]

Don't have a problem with the null. My Q is if I look at the tx waveform and the Rx they are both exactly in phase but the TGSL project shows a screen shot with a phase shift which the article says must be there.

[Koala]

https://www.youtube.com/watch?v=TVrISTKtxAw

Got to admit I did not bother with the phase while doing a Tesoro concentric.

Nulled on the scope then set the detector at 0 discrimination. Using a small gold ring, cut half, Victoria Penney, piece of ferrite and a small fibula. Kept testing until best distance with ferrite still Nulled. Best results was with an extra turn on the bucking coil. Ferrite Nulled right in the middle of the ground balance range and a small rusty forged nail still discriminated out.

The biggest change in distance was the small fibula. From memory it nearly doubled from the null point to best distance.

[mykaitch]

Looking at the TGSL project as my ref there is a clear phase lag in the Tx. My Null point is 4-5mV. Are you saying that the in phase Tx/Rx I see is okay? Reading volts won't show this, it has to be done with a 'scope.

[Dave J.]

The voltage on the unloaded receiver coil will normally follow the voltage on the transmitter coil (with no feedback coil). But, the receiver voltage will usually lead the transmitter voltage by a few degrees of phase due to the resistance of the transmitter coil.

The feedback ("bucking") coil winding has to produce a magnetic field opposed to that of the transmitter in order to approximately cancel the voltage induced in the receiver coil by the transmit field. The residual imbalance voltage (typically a few millivolts or less) may be of any phase and usually exhibits a lot of harmonic distortion.

[mykaitch]

Duh...

what IS the relationship ? I measure my tx rx as exactly IN phase (before amp).
Thanks.
Still chasing GB issue...

Seems that I need about 90 Degrees shift ??? Mine is wrong, is that it?