PI discrimination

[josephrz]

I think I read somewhere that the slope of eddy current decay is a function of the target's resistivity. If this is true, couldn't I take two samples of the signal a fixed time apart and from those sample/holds use the delta to infer a likelihood of the target being ferrous?

[Sean Goddard]

Re: PI discrimination

Josephrz, YES in a way I believe it is, BUT...

Try this, I posted it on a forum a while back and I would like some people to try it out.

Firstly, build your TX using a PUSH-PULL BRIDGE type system as in high output audio amplifiers.

Use a microprocessor (PIC) to measure the voltage at the points you want over THREE SUCCESIVE TX PULSES. Then average these responses (no idea what equation you have to use, but that's up to you).

Next, and IMMEDIATELY after the LAST of the three pulses, fire off a FOURTH POSITIVE going pulse, then, without allowing any time for decay, fire off a REVERSED POLARITY pulse.

Measure your two sample points (dv/dt) again, and COMPARE it with your average fro mthe three positive pulses.

What you SHOULD see, is that the MORE FERROUS your target, the MORE dissimilar will be your +/- decay pulse readings. The more ALIKE the +/- pulses, the LESS FERROUS your target.

What this system is doing is using the RELUCTANCE or the "resistance" of the target to being magnetised. As ferrous materials GENERALLY have higher reluctance than non-ferrous materials, then you will find that this will give you a PI with a rudimentary discriminate. Of course ferrite will STILL be a major problem.

DON'T FORGET to diode protect your output stage as the back EMF's of this system can reach 2KV! Also, remember to take into accoutn the decay time of t he field around your coil when transmitting the negative pulse immediately after the positive pulse. The latent decay time of the coil, or more specifically it's INDUCTANCE becomes CRITICAL, and you want to keep this as low as possible.

It works in theory, and it works on my bench, and, AGAIN, I give it to the world to investigate the possibilities. IF anyone makes a commercial machine using this system, then PLEASE remember who gave you this idea, and a reasonable contribution to my "beer & mortgage fund" would be gratefully accepted.

The last time I posted this idea, Eric Fooster sent me a personal email syaing that he wanted to try it. Don't know how far he got with it.

Eric, if you're out there still, drop me an email, we still need to meet up sometime.

[Sean Goddard]

Re: PI discrimination

Forgot to mention, you MAY have to characterise the WHOLE decay curve, but you could integrate this in the PIC mathematically.

seeing as how ferrous materials decay FASTER, INITIALLY than non ferrous, you could alsso include this in your equeation.

Either way, I cant see that it's that difficult to provide a good ferrous disc using this system.

Of course, Eric will come along and prove me wrong (no offence meant Eric) but if it REALLY is this simple I'm surprised that no one else has come up with this idea along time ago, then again, new inventions are made every day, and I DO recognise Eric as THE foremost authority on PI's in the known universe!

[josephrz]

Re: PI discrimination

Thanks, Sean, for your recommendation.

[Eric Foster]

Re: PI discrimination

No Sean, you could well be right. Whites took out a patent on using a reverse reset pulse to do ferrous discrimination, some years ago. They never used it in a production model though. Don't know why.

I may have been in at the start of PI development and made it acceptable as a treaure hunting detector, but there are many clever people out there now, who are doing things beyond my capabilities. That was one of the reasons why I started the PI forum, so that people could build on the work that I had done.

Eric.

[Sean Goddard]

Re: PI discrimination

Hi Eric, Well it wouldn't be the first time I've been wrong ;-)

I honestly didn't know about the Whites patent, but I think it shows I might be on the right track. Not bad for a Newbie who's trying lots of "What if's" out here.

I've still got a few ideas up my sleeve that I can't post, but one day I will, if I ever get commercial on the design.

I'm not afraid to have others learn and improve my designs, after all, that's how I got started.

Come to think of it, isn't that how we ALL got started?

The future is not as far away as we think it is, and I think there will be some very interesting developments soon.

With the combined efforts of people like those on this forum, the boundaries get pushed back a little with every post placed here.

[john]

Re: PI discrimination

If what i have read in the patents office of the usa, i think you might be advised to look at the Minelab patents. I am mot saying that their machines are pulse but a browse of their patents might enlighten you.

john.