multi frequency vlf

[paystreak]

Nokta Macro just introduced their multi frequency detector. How is multi frequency done? Is there a separate circuit on the pcb for each frequency?

What determines frequency for vlf detectors?

[Skippy]

There are a couple of technical articles here from Minelab which cover a bit of this topic. Their sequential multi-freq system ( 3.125kHz & 25kHz ) used in their BBS/FBS range is mentioned.
Documents KBA 00 & KBA 01:
https://www.minelab.com/usa/support/...-base/articles

Modern machines like the Nokta SMF ( and Equinox, Deus2 , etc ) 'simplify' the electronics hardware. The raw signal from the coil is amplified , perhaps a few hundred times, then fed into a high-speed high-resolution analogue-to-digital converter, read by a fast microprocessor. All subsequent analysis is done in software - demodulation, filtering, mixing, decision-making, audio output generation, LCD driving, etc etc.
The micro also generates the waveform that is transmitted by the coil, it just gets amplified before being applied directly to the coil.

[paystreak]

Fascinating. Thanks for the info. So, the Legend just be a "souped up" Simplex+ with a faster processor with more cores??
It would be interesting if somebody does a side by side tear down of these two detectors.
Thanks again,
paystreak

[Skippy]

Here's your Simplex 'teardown' , thanks to 'MD-Hunter' , in Ukraine. The micro is an ST-Micro STM32F401.

https://md-hunter.com/nokta-simplex-...w-in-pictures/

And at MD-Arena ( same guy? )
https://md-arena.com/razborka-nokta-...alloiskatelya/

STM micro:
https://www.st.com/content/st_com/en...f4-series.html

My hunch is that while designing the SMF machine, they realised they could easily make a stripped-down single-freq machine using the same basic design, that would be low-cost.

[Carl-NC]

Nokta Macro just introduced their multi frequency detector. How is multi frequency done? Is there a separate circuit on the pcb for each frequency?

What determines frequency for vlf detectors?

Paystreak, the answers to these questions could easily fill a book. And there are different ways to do MF so there's not just one answer. But done right, MF is not just a souped-up SF. Most of these newer machines are direct-sampling designs so the "separate circuit for each frequency" is all in software.

[ivconic]

Paystreak, the answers to these questions could easily fill a book. And there are different ways to do MF so there's not just one answer. But done right, MF is not just a souped-up SF. Most of these newer machines are direct-sampling designs so the "separate circuit for each frequency" is all in software.

Here is me messing around AD9833, trying to see what can be done with it.
I simply lined up 4,8 and 16kHz in sequence.
First part of video, transition without a pause and second part with 100uS pause.
Obviously this is not done right.

[ivconic]

And then i used only 8kHz and 16kHz.
16kHZ is with phase 90 degree relative to 8kHz.
And than becomes interesting:

[ivconic]

@Carl,
Arduino UNO + AD9833 = 22mA at 5V.

[Carl-NC]

The trick is switching frequencies while driving a search coil; you want to do it at the zero crossings of the coil current. Even then it can take several cycles to settle out.

[ivconic]

The trick is switching frequencies while driving a search coil; you want to do it at the zero crossings of the coil current. Even then it can take several cycles to settle out.

I will have to make some TX buffer and output and hook at this.
AD9833 allows me to adjust frequency, waveform and phase.

[Sean_Goddard]

Two words....

FOURIER SYNTHESIS.

Done by me back in 1989...

[ivconic]

[ivconic]

Goats have certain idea about it...

[Sean_Goddard]

I posted the document on this forum of all to see around 5 years ago.

Bob Podhrasky at Garrett reviewed it as did Guartel AND Whites in the US and UK AND CScope. ALL said at the time it was not possible, but now with DSP it is.

Nokta have the documents at the moment