Multi-Frequency

[Math]

those TX frequencies... carrier 856hz (171hz x5) - 1.712khz - 13.7khz are plan A
carrier 686hz (171hz x4) - 1.373khz - 22.0khz are plan B

[Skippy]

That 2nd choice, with a ratio of 16:1 frequencies seems a bit extreme. Have you intentionally chosen it that high for a technical reason? I see that there's merit to not having them close , 3:1 seems to work ( Fisher CZ ) but isn't there a practical limit in the other extreme, where there's nothing to gain, but problems to overcome, such as keeping a coil adequately nulled at two very different frequencies ?
As you will have seem from my previous posts, the coil is one of the tricky aspects of a homebuild multifreq machine, and if existing commercial coils were useable ( with modifications perhaps ) the project has a greater chance of success.

Also: Do you think you low choice of carrier freq could cause any problems, such as requiring a slow coil sweep speed to function properly?

[Math]

hi Skippy and thanks for posting, yes the plan B frequencies are a break from Minelabs x8.. ive chosen x16 to arrive at 22khz as i feel x8 and 11khz would be too low for sensitivity to low conductors.. thats a good point about the coil nulling at 2 frequencies far apart im hoping the multiple of 4 ratios to the coil resonant frequency helps.. this is a totally ‘start from scratch’ experimental multi frequency project for me and i agree with you re the coil building but in trying to improve on Minelabs design this project needs new coil specs for reasons that i will explain as i go.. im prepared to detail fully the coil building with diagrams and photos as i progress and how it can be done without much expense and minimum equipment

[Math]

ok coil design for an 8” DD for plan A design..

aiming for the most sensitive (sensitive to low conductor phase (R component) and amplitude (X component) changes) coil running 13.7khz for our high frequency part of this design..
ive found in my previous single frequency designs that designing a coil with a resonant frequency 4x above our transmit frequency and using a 1:4 TX:RX inductance ratio is best.. this should also work well for our low frequency 1.712khz part being 2x4 ratio below 13.7khz

using the formula from my first post, arriving at required Rsum (Rtx x Rrx) for the coil of 220 ohms
and using 180mm winding diameter for each coil of the 8” DD, ive chosen 0.31mm (29awg) enamelled wire for the TX
and running this through the online multi layer air core coil calculator...
found here.. https://www.circuits.dk/multilayer-air-core-inductor/

to arrive at TX coil, 1mH, 53 turns, 8 ohm, 29awg (0.31mm) RX coil, 4.6mH, 113 turns, 27.6 ohm, 31awg (0.25mm)
these are initial design figures and will vary a little when the coil is built, they dont need to be exact

this is a lower TX power design with the TX coil at 8ohm compared to 2ohm for the Minelab Sovereign

the idea is to use a TX sine wave voltage.. for depth and discrimination benefits

[Math]

just read part about carrier freq Skippy great question.. im planning on a few changes from the Sovereigns design to increase sweep speed.. im hoping that carrier frequency at around half the Sovereign freq isnt going to affect the sweep speed drastically.. i think its going to be build it and see!

[Math]

im thinking of a 15v TX sine wave.. the next steps are to design the TX drive and power supply, these are areas im not particularly good at and need to study more

the pic16f18426 has a complementary waveform generator CWG with input from the DSM that can be used for the TX drive

[Math]

anyone tried switching in extra capacitance for a lower frequency TX, know if it would work?
8.2uF polyfilm in this case for 1.7khz

[Carl-NC]

anyone tried switching in extra capacitance for a lower frequency TX, know if it would work?
8.2uF polyfilm in this case for 1.7khz

For a single-f detector with selectable frequency, it's easy. XP Deus does this, likely others as well. Trying to do it real-time in a MF design is exceptionally difficult, you have to precisely switch on the current zero-crossings. White's (Don Shoemaker) tried this years ago, got a couple of patents out of the effort but no products.

[Math]

thanks Carl i thought the Deus must do this but was a bit doubtful about operating in real time
ill have to rethink the design - it looks only possible the squarewave way Minelab use
i have a block diagram image of the idea in full but cant seem to add it to the reply
does an image need to be below a certain file size?

[Math]

its worked as attachment.. i couldnt see it in preview mode

[Skippy]

Your attachment is only 400 pix wide, it's almost impossible to read. Maybe re-size it to an appropriate size ( eg 1024 pix wide) and see if you have any success.
[ it seems to be a common problem, maybe George/Carl can look into it... ]

[Math]

i should really try this before asking this question and thats what id normally do but ive just got the pic this week and not setup yet.. i think i need to update mplab with the new chip linker? file as my version is a few years old now, ive no wifi yet to do this on my pc im just working from my mobile phone personal hotspot to ipad at the moment..

ok whats the 1mH TX waveform going to look like using only 133nf across for the 13.7khz half?
my best guess is a rocky mountain 1.7khz first half if at all..

[Math]

hi Skippy i took a photo with my phone then emailed to my ipad to post.. i choose full file size in order of MB when i emailed it thinking it would give the best resolution

[Skippy]

Yeah, that's the problem, I suspect, if you try and upload a picture 4 times wider than the forum page, it's not having it, and resizes it, rather drastically. The pics I've attached were all resized to about 900 pixels or less width beforehand, and they loaded OK, and at that size they can even be embedded directly into the post.

[Math]

ok thanks ill try emailing across to my ipad again at the lowest file size, im not sure if theres a resize option.. ill check