REVERSE DISCRIMINATION

[daverave]

ive been reading about reverse discrimination and i wonder if its really a good way to discriminate seeing it just uses a later sample...any thoughts on this please.

[Teleno]

ive been reading about reverse discrimination and i wonder if its really a good way to discriminate seeing it just uses a later sample...any thoughts on this please.

It's pseudo-discrimination by estimating the decay time constant (tau). High conductance (silver, copper, gold, aluminium) means longer decay.

[kt315]

Re: Pulsed IB-VLF Detector instead of PI-Detector.
Posted by: Dave Emery (---.sip.clt.bellsouth.net)
Date: March 3, 2005 07:13PM
Aziz,
Eric Foster made a true PI with a balanced searchcoil which could discriminate over twenty years ago. The problem is that the reactive component suffers from ground effect. Using filters such as a motion VLF will result in a major reduction of depth over a regular PI. In short it has been done and may well prove to be a method for the future once the problems involved with such designs are solved. The dream is to find a reliable method of making a regular PI discriminate like a VLF using only a mono coil.
----------
Hi Dave, When you say "discriminate", do you mean determining if the target is ferrous or non-ferrous, or are you referring to distinguishing between different kinds of non-ferrous targets, such as zip-tabs and coins?
Allan
-----
Allan,
I am talking about a fully variable discrimination control such as the control found on a VLF. Maybe even a notch discriminate control to boot! The detector Eric made was ferrous / non ferrous only but it goes to show you how far ahead he has been for so long. He made that detector over twenty years ago!!!

[daverave]

im really thinking about ferrous discrimination which is the biggest problem for a pulse machine to reject...when i get a target i normally turn the delay control up some what...and if the signal is still strong then i see it as junk like iron...but i know its not that black and white.

[Teleno]

im really thinking about ferrous discrimination which is the biggest problem for a pulse machine to reject...when i get a target i normally turn the delay control up some what...and if the signal is still strong then i see it as junk like iron...but i know its not that black and white.

In PI, ferrous/non-ferrous discrimination can be done.

Paramagnetic materials (ferrous) increase the coil's inductance L. Diamagnetic (Copper, Silver, Gold) decrease L.

A simple inductance measurment can produce a discriminating output. For example, you can use a "test period" in which you measure the coil's inductance using a constant voltage source. You measure the time it takes for the coil to reach a test current I. This time is proportional to the inductance. Or you measure the coil's natural frequency making it oscillate. You compare it to the measurements in the absence of a target.

[Sean_Goddard]

Yes this CAN be done and it's not as difficult as you think.

For Eric's version, he missed using a reference pulse BEFORE the main Tx to "sound out" the coil. I would not seek to demean his Genius in any manner, Eric thinks in a very linear fashion (God Bless him) and sometimes that can be a curse rether than a blessing.

BIG HINT: Think B-H hysteresis curves. I'll leave the rest for you to experiment with but I will say I've got a method which shows GREAT promise in the field. Go look at some of my other PI related posts to see what I mean.

[Teleno]

A pulse like this:

[daverave]

as far that i understand i thought ferrous disc could only be done using 2D coils...i know detectors like the chance and others using microprocessors can in some way discriminate out ferrous...and from what i can see from some of the eric foster designs he uses extra samples to get discrimination in some way...not sure if any can discriminate to full depth ???

[daverave]

is that waveform back emf curve without damping resistor ???

[Teleno]

The rising edge is constant current, the falling edge is underdamped. After the rising edge you measure target rsponse (using IB Rx coil), during oscillation you measure frequency for discrimination.

[daverave]

so you still need two coils to be able to do this ??? if so what discrimination is availiable with a mono coil ???

[Teleno]

so you still need two coils to be able to do this ??? if so what discrimination is availiable with a mono coil ???

You were given lots of hints.

[daverave]

i really do not want to use two coils to get ferrous discrimination...just wanna know if possible with mono coil....i know eric foster designs used extra samples but i wonder how good they were at rejecting iron of whatever size.

[kt315]

a time to do new project or to dream?

[Teleno]

i really do not want to use two coils to get ferrous discrimination...just wanna know if possible with mono coil....i know eric foster designs used extra samples but i wonder how good they were at rejecting iron of whatever size.

http://www.geotech1.com/forums/showt...189#post209189