VLF frequencies

[pebe]

Hi everyone - I’m new here.

I have not been involved with metal detectors before, but as I am retired with time to spare I thought I would like to have a crack at designing and building one from scratch. I have got hold of an old VLF machine because it provides the necessary basic hardware to experiment on.

But there are several points that I am not sure about, so for starters I’d like to pose these questions.

1. Looking at various schematics in this forum, VLF types seem to operate at frequencies from about 5kHz to about 15kHz. Is there a ‘best’ frequency?

2. Is the phase angle from a given target the same for all frequencies?

3. Is there a chart published anywhere that relates phase to type of target?

Any help would be most welcome.

[Carl-NC]

1. Lower frequencies tend to penetrate bad ground better, higher frequencies tend to be more sensitive to small thin and low conductive targets like jewelry & nuggets. Many detectors have migrated to around 15kHz as a good all-purpose frequency.

2. No, it shifts.

3. Not sure of one.

[pebe]

Thanks Carl. I'll go for 15kHz then.

[technos]

1. Lower frequencies tend to penetrate bad ground better, higher frequencies tend to be more sensitive to small thin and low conductive targets like jewelry & nuggets. Many detectors have migrated to around 15kHz as a good all-purpose frequency.

2. No, it shifts.

3. Not sure of one.

1) I thought that a general frequency was around 9kHz? 15 to 20kHz seems to be the realm of most nugget detectors. There is better iron reject/ground balance at these frequencies. But it would not go as deep if you were "coinshooting" esp. with a bigger coil.

2-3) While it is not fixed, it should be noted that pulltabs seem to give the same response as rings! If pebe wanted an idea of what metal detector theory theory is based on, he should look at some of the patents in that section that have some complete grafts of the different phase angles (and hence targets) of different things. However, it is noted that it's when the "trash" responds in the same way (i.g. a pulltab) is when the detector (and "us") is fooled! .

I think White's detectors are operating at 8.2 kHz, Tesoro at 12 kHz. Garret is around 6.5 KHz. If it was me, I'd build a detector to the same frequency as an inexpensive coil from White's (8.2kHz) -or somebody else for that matter - and buy one so I would not hassle with winding one!

[Carl-NC]

1) I thought that a general frequency was around 9kHz? 15 to 20kHz seems to be the realm of most nugget detectors.

Frequencies have been edging upward lately. White's used to be 6.6kHz; Prizms are 8.2kHz, MXT is 14kHz. Tesoro used to be 10kHz, Tejon series is 17kHz. Newer Fishers & Teks are up around 15kHz I believe.

If it was me, I'd build a detector to the same frequency as an inexpensive coil from White's (8.2kHz) -or somebody else for that matter - and buy one so I would not hassle with winding one!

I also recommend this. A wideband (untuned) coil gives the most flexibility. White's DFX/MXT coils are untuned, as are Prizm. Tesoro coils (I think all) are untuned as well. Use a factory coil to get your circuit working, then use your working circuit to build a custom coil.

[pebe]

Thanks to you both for your suggestions.

The used detector I have bought as a test bed (but not yet received) is a Mictronta 4003. I believe its performance is ‘middling’, and it runs at 15kHz.

I will take some measurements on that and change the coils as necessary. If it looks like the front end setup may be usable, then I’ll keep it. If not I’ll either replace both the coils or keep the TX coil and rewind the RX coil because at the moment it feeds a grounded base transistor. The Rx coil is series tuned so it will be low impedance, but ideally I need a high impedance one to feed an op amp.

The rest of the circuit will probably be gutted because I want to change the TX drive and try a different method of measuring the phase and amplitude of the received pulses. At the moment I am trying to find information on how to interpret the results once I can measure them. The electronics don’t pose a problem, but interpretation of the results may be a long learning curve.

[Qiaozhi]

Is there a chart published anywhere that relates phase to type of target?

Have a look at the U.S. Gifford patent (4,486,713).

[pebe]

Have a look at the U.S. Gifford patent (4,486,713).

Thanks Qiaozhi. That's just the information I was looking for.

[Agraz]

low frequency, and separation goes deeper, not detect smaller objects.

[Agraz]