The Moodz detector lives !

[simplejohn]

HI Moodz, yep the bipolar one,

Also on the vids noticed that the audio is predictably unstable... Target is clearly heard though, Is this a result of EMI in the surbs?

[6666]

I'm interested in bipolar to.

[Orbit]

Hello Moodz by batteries using your pi and which consumption ?

[moodz]

Hello Moodz by batteries using your pi and which consumption ?

I use LIPO batteries and its a 4 cell so it runs between 16.5 volts charged to 12 volts discharged. The current is 330 mA at 16 volts and 430 mA at 12 volts.

[moodz]

Aha!
I wondered for some time how you do it, but it is obvious now. That's how it goes with all nice things.
I noticed a method for direct X sampling by detecting the change of voltage at a snubber. With flyback controlled by snubber you get preferably a flat top voltage, and it is dependant upon the changes in permeability of the coil surroundings, or a presence of ferrous objects. The way you do it enables using the same front end for both X and R.
I observed the same effect with active damping using a MOSFET as a constant current device. Initial overshoot is inevitable, as the effective resistance at the beginning of Rx period is too much for a meaningful damping, but after the other hump it stabilises at proper resistance values and effectively it is faster than the critically damped coil. Detecting the first hump for X component is brilliant.
Well done

It works really well ... here is a movie .... a ferrite target ( first half of video ) vs a 20 cent coin - non ferrous / magnetic ( second half of video ).

https://www.youtube.com/watch?v=8r9iW-qbIrY

[Orbit]

Thank you Moodz so I thought about consumption !

[Polymer]

Hi WM6 ... I have grown tired of secret squirrel detector projects so very shortly I will gather the documents and publish my design for a kickass PI detector.
This detector can resolve X and R signals from a MONO coil in the time domain ... and using only two samples ( this provides ground rejection AND opens the door to discrimination on a PI )

See Pix below ....

Hi moodz, you really got me howling with laughter with the " ...secret squirrel... " expression. I've never heard the expression before and luv it.
It also hits the head of the nail from my perspective. Way to go!

I'm looking forward to your project coming out into the light and hopefully there will be pcb's and/or pre-programmed cpu's.

Thank you for bringing this out.

Godspeed to you!

Have a great weekend, Polymer

[Geo]

i am interesting also for this project!!!

[soverey]

I'm also interested in this project, thank you!

[Achillion]

Moodz, I'm not sure how you establish your discrimination but I picked up something from the web that may be of use.

It's based on sampling around the zero crossing and integrate a section of the X and R component of the signal.

Well, probably best way to explain is to have a look at the attached picture

[moodz]

Achillion, I have seen that one before but thanks alot anyway .. if it is the one I am thinking of with the box car integrator in software it is OK in air but needs more work to track the ground and still discriminate. The difficult part is doing it with weak targets eg at long range or small targets.

[moodz]

10 mm aluminium foil target test ....

https://www.youtube.com/watch?v=a7vaqyAi6uI

For the 11 inch mono its not too bad detecting the 1cm x 1cm aluminium foil target at 30 mm.

[moodz]

discrimination waveform is improved ( again )

...and here are the discrimination results. Detector signal chain is set up to reject ferrite ( X ) and respond to conductive ( R ).

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

Four targets are used in order
1. Ferrite ( touching coil )- strong X response ...very little R response from detector audio.
2. Copper at 40 cm - strong R response ... very little X response
3. Steel at 30 cm - About equal X and R response
4. foil at 5 cm - some X and some R. ( X due to inductive surface effect ... eddy current cannot go deep in foil )

The detector is a lot more sensitive than those ranges ... had to move targets close enough to demonstrate effect on waveform which is then amplified / sampled.

by sampling an X and R channel in conjunction with a ground tracking means .... target discrimination and ground rejection is achieved with very high signal to noise ( ie discrim at depth )

Clearly a new age of discriminating, sensitive and deep punching ( works with monos ... ) PI detectors is possibly within reach.

moodz

PS .. the noisy junk on the waveform was due to a dry joint ... now corrected ( but I could not be bothered reshooting the video :-) )

[6666]

https://www.youtube.com/watch?v=a7vaqyAi6uI

For the 11 inch mono its not too bad detecting the 1cm x 1cm aluminium foil target at 30 mm.

Says
This video is private.

[moodz]

Says
This video is private.

whoops .. try now.