Spectra V3

[Davor]

Considering the comparator being in charge of supplying the final push-pull, I'd say it is as crude as it gets: a zero crossing of a sum of signals triggers the comparator. It is binary. Point is that the error induced this way is AM modulation of each carrier that is at least the lowest frequency apart, so the only requirement for picking the frequencies is to keep all the criss-crossing modulation products away from each of the carriers. Having a typical MD RX bandwidth well below 100Hz it is not too difficult to achieve.
The other option is having all frequencies in a perfect harmonic relationship, only avoiding the 2nd harmonic, and it seem to be the case here.

I might make a simple spice sim of this case. Or not.

[Sergey_P]

Distinctive feature of the V3 - availability mode correlation S=(S(f1)+S(f3))*S(f2)

[ODM]

Having dived into russian and gotten the schematic, it looks like there's no frequency summing magic involved, just a level translation as the signal is coming from a microcontroller's digital output pin. If it's a 2.5kHz square wave, it'll have a strong third harmonic at 7.5kHz and a relatively strong ninth at 22.5kHz.

It would be interesting to see if there were pulse trains which would emphasize some harmonics in order to pour energy into them, there probably are. As for the distinctive features - I guess any metal detector with capability for receiving multiple frequencies at the same time, either through analog or digital means, could do the same math. Getting those frequencies separated in the first place costs effort.. and components.

[Sergey_P]

http://whiteselectronics.com/images/stories/products/v3_story_booklet.pdf


WO8704801A1 ( http://md4u.ru/download/file.php?id=6952 )

[Aziz]

http://whiteselectronics.com/images/stories/products/v3_story_booklet.pdf


WO8704801A1 ( http://md4u.ru/download/file.php?id=6952 )

Thanks Sergey,

I didn't know the story booklet about the V3.
I like White's honesty about the multi-frequency story. All in all the Spectra V3 is a nice detector.

I did a two-frequency VLF. Your LC network for three-frequency VLF is interesting too, which I might test it one day.

Cheers,
Aziz

[Aziz]

Oh yes,

the H-bridge in the TX driver makes fully sense using either a rectangular or triangular current exciting. You get odd harmonics (sin(wt), sin(3wt), sin(5wt), ...)
1*2.5 kHz, 3*2.5 kHz (=7.5 kHz), 9*2.5kHz (=22.5 kHz)
Yep, makes sense.
Aziz

[Sergey_P]

square wave: A(3)=1/3*A(1); A(9)=1/9*A(1). Harmonics in V3 - aligned

[Aziz]

square wave: A(3)=1/3*A(1); A(9)=1/9*A(1). Harmonics in V3 - aligned

Yep, I've left the coefficients.

How it is aligned? In the coil with an LC tank?
(Curious and interested to know and learn.)
Aziz

[Sergey_P]

I am lazy, do calculations - themselves

http://i49.tinypic.com/hwmw4o.jpg
(coil has several resonant circuits )

[Aziz]

I am lazy, do calculations - themselves

http://i49.tinypic.com/hwmw4o.jpg
(coil has several resonant circuits )

Thanks, I know it now.

I'm keen on trying it too after changing my multi-frequency controller (Hmmm, Davor's ultra-low noise amplifier isn't leaving me alone.. ) .

Cheers,
Aziz

[Aziz]

I found it difficult to follow the Russian topic! (BTW, I don't speak Russian).
Here is an online translation:
http://www.microsofttranslator.com/b...D94%26t%3D8074

Aziz

[Carl-NC]

It would be interesting to see if there were pulse trains which would emphasize some harmonics in order to pour energy into them, there probably are.

There are. The three frequencies have closely equalized energy levels.

[Skippy]

Would this help?http://forums.whiteselectronics.com/...0&d=1347182782

[Aziz]

Here is the H-Bridge output: (digital)
http://www.wolframalpha.com/input/?i...from+0+to+4e-4

(enter at input line: plot (sgn(sin(2*pi*2500*t) + sin(2*pi*3*2500*t) + sin(2*pi*9*2500*t))), t from 0 to 4e-4 )

Here is the, what the TX coil makes at the end (coil current):
http://www.wolframalpha.com/input/?i...from+0+to+4e-4

( plot (sin(2*pi*2500*t) + sin(2*pi*3*2500*t) + sin(2*pi*9*2500*t)), t from 0 to 4e-4 )

All three frequencies overlayed:
http://www.wolframalpha.com/input/?i...from+0+to+4e-4
( plot (sin(2*pi*2500*t), sin(2*pi*3*2500*t), sin(2*pi*9*2500*t)), t from 0 to 4e-4)

Cheers,
Aziz

[Taktyk]

HI,

How much current flows through the coil DFX?

Below is a oscillograph for the coil current... This is BM950 for W5900/6000. Here it is almost 1A.

Regards
Taktyk