GPX Series Operating Frequency

[AuMan]

To energise the coil with current your detector is transmitting an RF signal to the coil and this is what I want to understand.

It isn't putting an rf frequency into the coil It is turning the coil on for 240uS reaching a current of about 2.5A in the coil, which then switches off and the magnetic field collapses back into the coil in about 3~5uS, energising the target.
From this point, the coil needs to settle to no signal ASAP before the first sample is to be taken. This is where the coils resonant frequency comes into play along with the detectors damping resistors. The coil should settle to 0v and not oscillate.(A vlf will continue to oscillate)
Just note that when ground is added in the coil will no longer settle to 0v, the 0v is in air, no metal.

The NF coils from memory when I measured them ages ago were about 750khz.
The ML commander I think were about 500khz.

The Nugget finder coils are a fast coil and as far as what the detector see's they are perfect coils.

Search for the sd2000 schematic. In there there is a transmit pulse train with all the pulse lengths. The GPX is not that much different.

Cheers Mick

 

[Deepseeker]

Awesome explanations guys! I find this stuff totally amazing :Y:

 

[user 12139]

good stuff AuMan, I did the resonance test again, turns out I was a little wrong in my post above. Using a VNA I measured my NF 15" Evo and a ML 11" mono, both came in just under ~1000Khz (the test was done inside, at such low frequencies the wavelength is huge so I'd expect more accurate results testing outside).

If I measured two identical coils, and one came in say 250khz lower than the other, from a performance standpoint what would you expect to change? Could you provide feedback to this question please for both PI's and VLF's.

I do have some papers relating to fast coils and the design of, I need to read it properly!

Thanks again AuMan for coming back on this post, I have a better understanding now thanks to you. Cheers


AL.

 

[user 12139]

for this who wish to view the SD2000 transmit timings

https://reader009.docslide.net/reader009/html5/20170926/55219a0b497959b2608b45ed/bg3.png

 

[Ded Driver]

BigL, just curious do you own a RF Analyser (VNA) or where did you get access to one, did you find it for hire or purchase? or you work in an industry/business that has these?

 

[user 12139]

BigL, just curious do you own a RF Analyser (VNA) or where did you get access to one, did you find it for hire or purchase? or you work in an industry/business that has these?

I own the VNA, I used it in amateur radio. As the need for them is very specific I doubt very much you'd find one for hire. I could probably point you to where you "may" borrow one, PM me

 

[Ded Driver]

cheers BigL, I don't need one, was just curious as I know they have specific uses & not many people outside R&D or commercial biz use them. I wasn't aware of amateur radio use. I also know they can cost BIG $ ... small handhelds not so bad, but Ive seen pro units for just shy of $30k.
awesome bit of kit but beyond my needs. :Y:
.
would be very interesting if you did a set of waveforms for a GPX as you did for the SD .. just to compare changes between the 2 series.

 

[AuMan]

Hi Ded,

Just for reference that diagram is from here> http://www.minelabmods.com/joomla3/index.php/news/120-sd2000-circuit-diagrams

BigL, I'm not sure what actual difference a faster/slower resonant frequency coil will make on a PI other than initial settling time. I would have to test them side by side on the bench and in the field to know for sure.

Coils for detectors are a bit of a black art!

As for VLF detectors, it will change the operating frequency, but other than that I don't know as I don't play with them.

Cheers Mick

 

[Ded Driver]

cheers for the additional reference Mick :Y:

 

[user 12139]

Hi Ded,

Just for reference that diagram is from here> http://www.minelabmods.com/joomla3/index.php/news/120-sd2000-circuit-diagrams

BigL, I'm not sure what actual difference a faster/slower resonant frequency coil will make on a PI other than initial settling time. I would have to test them side by side on the bench and in the field to know for sure.

Coils for detectors are a bit of a black art!

As for VLF detectors, it will change the operating frequency, but other than that I don't know as I don't play with them.

Cheers Mick

thanks Mick, yes I was thinking the same result with a PI.. it'd probably resemble a slow coil. Your dead right about black art, I know from my time building phrased arrays that sometimes what doesn't makes sense results in positives, but there's always a negative too, its a balancing act.

I guess most who've read this post realise my intentions is to build a few coils for both PI & VLF.. will be starting with one for my NOX 800, not interested in the target ID working at this stage as thats a completely different ball game, Im guessing I'd need Minelab to get onboard with the circut for that. Anyway, I'm confident I can build a ripper gold coil for the NOX as all the Minelab NOX manufactured units are a broadband coil whereas mine will be purely for single frequency gold use at 20 or 40 Khz., one or the other, not both. My aim is to try the Super-D design which are reported by Minelab to benefit depth. This will have 3 benefits over stock NOX coils 1) it won't be broadband which should benefit somewhere, I'm guessing sensitivity 2) better depth compared to a DD and 3) using a larger aperture design will open the target sweet spot. It will be interesting to see the outcome compared to stock NOX coils.

As for my PI 4500, I also plan to try a Super-D for it.

I'm guessing a mono coil would not work on VLF due to the timing pause required for mono receive like the PI's have, whereas a DD design has individual TX & RX coils, can you tell me if this is correct please?

Thanks AL.

 

[AuMan]

thanks Mick, yes I was thinking the same result with a PI.. it'd probably resemble a slow coil. Your dead right about black art, I know from my time building phrased arrays that sometimes what doesn't makes sense results in positives, but there's always a negative too, its a balancing act.

I guess most who've read this post realise my intentions is to build a few coils for both PI & VLF.. will be starting with one for my NOX 800, not interested in the target ID working at this stage as thats a completely different ball game, Im guessing I'd need Minelab to get onboard with the circut for that. Anyway, I'm confident I can build a ripper gold coil for the NOX as all the Minelab NOX manufactured units are a broadband coil whereas mine will be purely for single frequency gold use at 20 or 40 Khz., one or the other, not both. My aim is to try the Super-D design which are reported by Minelab to benefit depth. This will have 3 benefits over stock NOX coils 1) it won't be broadband which should benefit somewhere, I'm guessing sensitivity 2) better depth compared to a DD and 3) using a larger aperture design will open the target sweet spot. It will be interesting to see the outcome compared to stock NOX coils.

As for my PI 4500, I also plan to try a Super-D for it.

I'm guessing a mono coil would not work on VLF due to the timing pause required for mono receive like the PI's have, whereas a DD design has individual TX & RX coils, can you tell me if this is correct please?

Thanks AL.

BigL, I can't help you on the VLF side of things as I will most likely lead you astray!
If you haven't already join up to geotech there is so much info on vlf detector theory etc as well as pi.
https://www.geotech1.com/forums/forum.php

As for a Super D coil for a GPX, that could work. There are separate preamps for both the Mono and DD inputs from the coil plug and when in mono mode you receive from both preamps, cancel the outputs of the preamps are subtracted and DD only 1 preamp is used.

Cheers Mick

 

[user 12139]

Seperate preamps, that's interesting Mick and very good to know, thanks for that.

I'm back at work for a few days and head out for a swing with a work colleague for a week or so thereafter so will follow things up on the Super-D a bit further down the track when it's too cold to do anything outside. I'll keep this post updated as things progress.

Thanks everyone for the feedback, it's put me on the right track, some very helpful info and I do appreciate it very much.

Thanks AL

 

[gef12]

unlike VLF types which operate in the frequency domain (x) KHz .... PI types work in the Time domain ( pulses / sec )
for info

 

[user 12139]

this covid thing has done wonders with my understanding of PI's

Turns out we were all in the ball park to some extent with how a GPX works.

My testing confirms the GPX pulse output is consistent in frequency, it transmits two bursts of it per cycle. The two bursts differ in amplitude and the delay between them is consistent.. as you'd expect.

Its obvious the brains lay in RX algorithms

 

[user 12139]

unlike VLF types which operate in the frequency domain (x) KHz .... PI types work in the Time domain ( pulses / sec )
for info

thats right.

VK.. AOCP?

 

[gef12]

unlike VLF types which operate in the frequency domain (x) KHz .... PI types work in the Time domain ( pulses / sec )
for info

thats right.

VK.. AOCP?

yes VK8.. Yes AOCP erfect: