EXTRA_EXTRA_DEEP DISCRIMINATING DETECTOR

[Tinkerer]

The Tinkerers IB-PI forum has been asleep for some time. But with the new year, we also have a new design.
The TINKERERS_XXD is the medium power, medium depth model. XXD stands for EXTRA, EXTRA DEEP. (XXXD will be the very, very deep one)
The power consumption is near 1A. With today's excellent light weight LI-ION batteries, this seems to be an acceptable power consumption.
There are 2 standard coils. A 500mm diameter DD coil and a 500mm dia. concentric coplanar coil. Both coils use the same coil housing.

The discrimination for FE is good. The T_XXD, discriminates a steel crown cork at full detection depth, in any position.

How deep is XXD? 50cm or 20 inches for a Dime (18mm) or a Nickel (21mm) Both coins are cupro-nickel, but of slightly different alloy. So get ready for a lot of deep digging.

The T_ XXD is optimized for nonmagnetic metals. It does also find iron in the all metal mode, but to a relative lesser depth.

Tinkerer

[Tinkerer]

TINKERER_XXD_DISCRIMINATING PCB

I am still learning how to use this CAD software.

Here is a screen shot of the PCB. Single sided.

Tinkerer

[WM6]

I am still learning how to use this CAD software.

Here is a screen shot of the PCB. Single sided.

Tinkerer

HI Tinkerer

Excellent project.

What about possibilities of mono-coil design?

[Tinkerer]

HI Tinkerer

Excellent project.

What about possibilities of mono-coil design?

It is possible, if you would like to spend some time on it, I give you the information on how to do it. Which CAD software do you use?

For myself I need to finish this project, before I start another one, (like mono coil)

All the best

Tinkerer

[WM6]

It is possible, if you would like to spend some time on it, I give you the information on how to do it. Which CAD software do you use?

I use very basic and simple Circuit Wizard, but after testing different one I cannot found better to suit my needs. It is multitasking, easy to use, simulation and PCB routing feature incorporated too.

[IBGold]

Hi Tinkerer,
What is involved in the Mono conversion, coil wise no problem for me if you want to contact me direct Moodz can give you my personal contact details.
Regards, Ian.

[Tinkerer]

Hi Tinkerer,
What is involved in the Mono conversion, coil wise no problem for me if you want to contact me direct Moodz can give you my personal contact details.
Regards, Ian.

Hi Ian,

Your coil building perfection is legendary. There is absolutely no doubt that coil quality is extremely important for a good detector.
The Tinkerers_XXD model uses a very unsophisticated coil at present. Many of the usual problems of the PI coils have been taken care of in the processing circuit. But who knows how much the performance could be improved with a top quality coil.

To produce the coils in series, they will need to be built with some precision for repeatability.

I mentioned above, that I think that a mono coil version of the Tinkerers_XXD could be built. I have not tried that yet and have no idea if the sensitivity and depth and discrimination can be the same. But yes, I think it can be done.

Mono coil. Everybody wants a mono coil detector.
Let's see if we can agree on a definition for mono coil.
Can we say that any coil that is not induction balanced, is a mono coil?
Can Moodz's differential coil be called a mono coil?
Can separate TX and RX coils that are not balanced, be called a mono coil?
Dual field coil?

I am really itching to give the mono coil version a try, but I must finish the IB version first. At present the signal processing is done in analog, but the timing is done with the PIC. I also want to control the audio with the PIC. This will take me a long time because the programming language learning curve is very steep for me.

Digital control and signal processing can take the design to a much higher level.

I would gladly accept help for any and every part of the design and am willing to share all of the design's secrets with the helpers.

All the best
Tinkerer

[Aziz]

Hi Tinkerer,

there is no need for a mono coil design. The IB configuration can compensate the losses and has some superior benefits over a mono coil design.

Good to see, you have removed all the resistors in the LC tank path. The 100k resistor does not contribute much to the losses.

Aziz

[Tinkerer]

Hi Tinkerer,

there is no need for a mono coil design. The IB configuration can compensate the losses and has some superior benefits over a mono coil design.

Good to see, you have removed all the resistors in the LC tank path. The 100k resistor does not contribute much to the losses.

Aziz

Thanks for the feedback.

Your help with the LC tank has been very valuable. Your help is much appreciated.

I find the IB configuration combined with the differential amplifier does an excellent job. But people keep asking for a mono coil design.

The 100k resistor is more of a place holder. Maybe it will help to reduce some of the Mosfet switching noise. There are probably better Mosfets available than the one I use, but I don't know how to chose the parameters for low switching noise at 5A coil current.

All the best

Tinkerer

[Aziz]

Hi Tinkerer,

one of the benefit of using a differential front-end is the following:

You can quickly change the configuration into an anti-interference coil design, which will gain 20-30 dB more SNR.

Just try this:
Make an exact copy of the receive coil (same parameters, same geometry, same inductivity, same resistance, etc. but both "receive" coils operated without center-tap = two mono receive coils). Put the new coil 1 meter apart from the receive coil and feed it with the correct signal polarity to the other differential input (change winding orientation if necessary) . So any EMI picking from the receive coil is cancelled from the other coil.

You can also fine balance the IB coil configuration, if you move the other coil closer or farther from the receive coil. To get more depth, the EMI cancelling coil should be as far as possible of course.

I am not joking. Just try it. This will surprise you, if you can cancel most of the common mode EMI noise (inductively coupled EM noise).

Aziz

[moodz]

Hi Ian,

Your coil building perfection is legendary. There is absolutely no doubt that coil quality is extremely important for a good detector.
The Tinkerers_XXD model uses a very unsophisticated coil at present. Many of the usual problems of the PI coils have been taken care of in the processing circuit. But who knows how much the performance could be improved with a top quality coil.

To produce the coils in series, they will need to be built with some precision for repeatability.

I mentioned above, that I think that a mono coil version of the Tinkerers_XXD could be built. I have not tried that yet and have no idea if the sensitivity and depth and discrimination can be the same. But yes, I think it can be done.

Mono coil. Everybody wants a mono coil detector.
Let's see if we can agree on a definition for mono coil.
Can we say that any coil that is not induction balanced, is a mono coil?
Can Moodz's differential coil be called a mono coil?
Can separate TX and RX coils that are not balanced, be called a mono coil?
Dual field coil?

I am really itching to give the mono coil version a try, but I must finish the IB version first. At present the signal processing is done in analog, but the timing is done with the PIC. I also want to control the audio with the PIC. This will take me a long time because the programming language learning curve is very steep for me.

Digital control and signal processing can take the design to a much higher level.

I would gladly accept help for any and every part of the design and am willing to share all of the design's secrets with the helpers.

All the best
Tinkerer

Hey Tinkerer .... I will code your DSP in a DSPIC since I think the task will be quite similiar to what I have already for mono differential ... however I still dont understand what your circuit is ... so it is hard to base code on that.

Regards,

moodz

[IBGold]

Hi Tinkerer,
Any help I can give you you are welcome I have been playing with some new designs over the last months now all in testing phase one of which is an induction balanced coil that I call a DO/DO it induction balances to within 2 micro-volts but when trying to set the RX damping I get strange waveforms and not acceptable sensitivity the D and O coils are in series Aziz will understand which coil I am talking about so I am about to scrap that design also on test is a non IB coil with a single TX coil and a Duel-Field RX coil I am also having problems with this coil although both work on my test detector a Modified SD2000 they don't show any advantages over other coils so I may scrap this design also but the last new design which Aziz also knows about is a AI coil which I call my butterfly coil and this shows good sensitivity and quietness on the bench just have to do field testing on it could be worth a try on your machine but it is definitely a home brew option to do properly as the time taken to make it would make it non viable commercially I have yet to build a differential coil but will when Moodz is ready.
Tinkerer as I said any thing I can do on coils feel free but on the electronics side I only know enough to be dangerous and all in the analog domain when components had legs but am willing to give anything a try.
Regards, Ian.

[Tinkerer]

Hi Tinkerer,

one of the benefit of using a differential front-end is the following:

You can quickly change the configuration into an anti-interference coil design, which will gain 20-30 dB more SNR.

Just try this:
Make an exact copy of the receive coil (same parameters, same geometry, same inductivity, same resistance, etc. but both "receive" coils operated without center-tap = two mono receive coils). Put the new coil 1 meter apart from the receive coil and feed it with the correct signal polarity to the other differential input (change winding orientation if necessary) . So any EMI picking from the receive coil is cancelled from the other coil.

You can also fine balance the IB coil configuration, if you move the other coil closer or farther from the receive coil. To get more depth, the EMI cancelling coil should be as far as possible of course.

I am not joking. Just try it. This will surprise you, if you can cancel most of the common mode EMI noise (inductively coupled EM noise).

Aziz

In certain regions people complain a lot about EM noise caused by lightning etc. I think that the anti interference method you explain above, should help a lot.

I am definitely going to try it.

Tinkerer

[Tinkerer]

Hey Tinkerer .... I will code your DSP in a DSPIC since I think the task will be quite similiar to what I have already for mono differential ... however I still dont understand what your circuit is ... so it is hard to base code on that.

Regards,

moodz

This is very kind of you. I will make a description of my signal processing circuit and send it to you together with the schema. It is a bit different from traditional PI.

For starters, I use a differential input preamp. Nobody does that with PI. However, some time ago you posted your first post about a fully differential front end.

I took a fraction of your idea, the differential preamp, and started experimenting with that.
Over the months, this has evolved into a new way of capturing and processing the target response signal, with very interesting results.

The way I see it, we can all gain by working together. Each of us has some ideas, each only has a limited amount of time to spend on the projects. If we pool our time and ideas and individual knowledge, we get much faster to our goals.

All the best

Tinkerer

[Tinkerer]

Hi Tinkerer,
Any help I can give you you are welcome I have been playing with some new designs over the last months now all in testing phase one of which is an induction balanced coil that I call a DO/DO it induction balances to within 2 micro-volts but when trying to set the RX damping I get strange waveforms and not acceptable sensitivity the D and O coils are in series Aziz will understand which coil I am talking about so I am about to scrap that design also on test is a non IB coil with a single TX coil and a Duel-Field RX coil I am also having problems with this coil although both work on my test detector a Modified SD2000 they don't show any advantages over other coils so I may scrap this design also but the last new design which Aziz also knows about is a AI coil which I call my butterfly coil and this shows good sensitivity and quietness on the bench just have to do field testing on it could be worth a try on your machine but it is definitely a home brew option to do properly as the time taken to make it would make it non viable commercially I have yet to build a differential coil but will when Moodz is ready.
Tinkerer as I said any thing I can do on coils feel free but on the electronics side I only know enough to be dangerous and all in the analog domain when components had legs but am willing to give anything a try.
Regards, Ian.

Hi Ian,

thank you for your offer. We both seem to enjoy experimenting with new ideas as a hobby, even if there are sometimes slim chances for commercial future of our projects. Financing our hobby is of course a problem in itself.
I think we can consider your modified SD2000 as a bench mark, to test other detectors against.
The way I see it, an ideal coil, is a coil that is perfectly matched to a certain detector and brings it to it's best possible performance. You have developed that into a fine art.
I spent a lot of time tinkering with PI-IB coils, so maybe I can be of some help with your IB problem. I don't know anything about the functioning of the SD2000, nor do I know what a DO/DO coil is so you will need to supply some scope pictures and description of the problem as well as of the way you need the signal where it enters the SD2000 circuit.
We can work on it here on the forum or by email, whichever you prefer.
I will send you my email address.

The TINKERERS_XXD coil (500mm OD, about 300uH, is made of a TX coil of low DC resistance, anywhere below 500mOhm will do, including the cable. At present I am using 16 Turns of #14 AWG solid wire, in a tight bundle and 3 legs of #14 AWG multi strand wire, hand twisted, for the feed cable. The third leg is for the Bucking coil, 4 Turns 4x#19AWG magnet wire, laid on the RX coil.

This crude setup gives me the results mentioned above, even inside the house, with re-bars in the concrete walls, floor and ceiling and all sorts of equipment, large freezer, computer, fluorescent lights etc. running less than 10 feet from the coil.

Ah, the RX coil is 250mm ID, 22 turns of #26 AWG magnet wire, center tapped. The leeds, A, B and ground are #22AWG, solid, hand twisted.

Shielding is graphite composition home made paint.

So, yes, there is definitely room for improvement. Litz wire would reduce the partial response signal of the eddy currents generated in the solid wire of the coil.

All the best

Tinkerer