Frequency Steps

[Danno]

Hi,
There are a few selectable frequency detectors on the market, I was wondering what the optimum ratio between frequencies would be to see a marked change.
For example if you had a detector set on 4.6khz I am guessing you would not see much difference in response until at least double, so about 9.2khz then 18.4khz?

Or as the frequency increases would the ratio be less than double?
I appreciated you have to factor in ground and EMI etc.

Thanks in advance

[waltr]

Check in on the thread about the frequencies used in the NOX detector.
Guessing they have worked out your question to choose frequencies.

[Skippy]

I don't think the Equinox (single) frequencies are well thought out, personally, they just opted for the 2:1 ratio, covering the range the machine/coil works over, so giving them 5 / 10 / 20 / 40 , then threw in 15 as an all-round freq. If they chose to miss off 20 & 40 from the 600 model, they would only be left with 5 & 10, so they had to include another choice, hence the 15.
I think a closer ratio than 2:1 would be better, perhaps 1.6 :1 , giving 6 / 10 / 16 / 25 / 40.

However their multifreq components of 7.8 / 18.2 / 39 (in most modes), with weighting toward 7.8 or 18.2 depending on operating mode, seem to work pretty well.
If I was designing a two-mode multifreq machine, these two 'main' frequencies would be close to my choices. The 7-8 kHz choice has proven to be a good choice for general medium/high conductor targets, eg. it's choice in many current Fisher/Tek machines ( Greek series etc) , and for low-conductor hunting, such as we do in Europe for ancient coins and smaller targets, the 13-18 range has proved itself with the success of machines like the Tek T2 / XP GMP & Deus on 12 or 18.

What freqs a selectable-freq machine runs at is also affected by the circuit topology. Are you using forced drive ( like the Equinox, Whites V3/DFX , Fisher CZ's ) or are you wanting resonant transmission ( like the Deus, and NokMak multifreq machines ) ?

[Math]

Hi Skippy when you say forced drive vs resonant.. is this square wave from an mcu/timer vs colpitts/hartley types

such as Tesoro Silver Sabre etc.. the Deus uses this? and Nokmak resonant multi freq must be difficult to do?

[Skippy]

'Forced drive' is square-wave, normally, from a H-bridge circuit, or something that behaves much like it. I recall Fisher CZ's have half a H-bridge, and large capacitors on the other half ?, but that is a single frequency 5kHz, it may not suit complex waveforms like the V3 / Equinox. And it's possible to use a centre-tapped TX coil and achieve the same effect, I would imagine.
Resonant mode circuits generally switch in additional capacitors to drop down the tuned frequency. Possibly switching in/out resistors too, to control the Q-factor. The Deus uses this method, it's assumed the Nokta Impact ( and hence likely the Makro MultiKruzer) also does this, both using latching relays for switching. It's also viable to switch the inductor - this was considered when we worked on Henrik's 'Deus small coil' project. A bifilar-wound coil could be configured to use one or both windings to give approximately 2:1 frequency change. The actual drive for these switched tuned circuits comes from digital logic ( usually a micro ), so they are not free-running Hartley/Colpitts etc circuits.

[Math]

Thanks Skippy great info.. ive been thinking about different TX drive types recently and multifreq designs..

With the older free running colpitts etc a metal target altering the coil inductance would in turn produce a
freq change of the TX osc..? an advantage esp. as you say in #3 when using higher freqs for low conductor
hunting as this would increase phase shift as well as amplitude? Using the TX signal for the demods also gives extra
effect as its width changes? I think Tesoro in their Silver Sabre design were doing something with the TX shape in the
GEB/Disc sections? Also free resonant designs have full range harmonics.. better for low conductor hunters?

Going to 'part forced' square wave based TX drives like Fisher 1265/6 and Lobo lowers these effects.. theres some loss
of the even harmonics and in fully forced designs as Minelab F/BBS etc these resonant benefits are about lost?
But part and fully forced square drives have their own advantages.. more frequency stability = less noise, larger current changes?

[waltr]

If I was designing a two-mode multifreq machine, these two 'main' frequencies would be close to my choices. The 7-8 kHz choice has proven to be a good choice for general medium/high conductor targets, eg. it's choice in many current Fisher/Tek machines ( Greek series etc) , and for low-conductor hunting, such as we do in Europe for ancient coins and smaller targets, the 13-18 range has proved itself with the success of machines like the Tek T2 / XP GMP & Deus on 12 or 18.

I agree 7kHz and 14kHz for high and low conductors. It was Dbowers that built and compared the TGSL (14.5kHz) to the IDX (7kHz) on a set of buried targets.
The IDX was better on high conductors and the TGSL on law conductor, Both where the same on medium conductors (lead bullets).

Maybe doubling frequency is enough to obtain a significant change of signal.
It is also easier to do in the circuits/processor.

[Math]

Yes i agree too Waltr those are 2 good freqs to use for high and low conductors

Ive just thought on the TX width changing for the demods on a Colpitts/Hartley free running resonant design
..the RX changes too a short time after so not really any extra benefit there..?

But if you were to run a resonant type TX osc and use it to gate/clock an mcu and average that to produce a slowly
changing frequency to follow any drift in the free running TX osc.. (the NCO module on a Pic micro would be
ideal for this) ..then you could use that for the demods which would double the benefits of the free running
designs.. re metal targets altering both the coil inductance and frequency with full range harmonics

[ivconic]

My most desirable targets are ancient coins.
Most made of bronze, rarely from silver and gold.
Many of bronze coins are coated with thin silver layer.
So, i am telling this to point on my next opinion.
As above noted; during the 33 years of coin-shooting mostly; i found out that most beneficial working frequencies for me are around 7kHz.
Let's say 4.6 to 8.7khz range exactly.
That range compromises with all pros and contras and presents best trade off for my kind of searching.
Frequency itself does not mean a thing. It is almost irrelevant if not followed by several other important features at detector.
First of all is GEB feature. Second recovery speed. And third and less important is Discrimination.
But please do keep in mind that this is very particular and special case, my own case, may own preferences.
It should not be understood as general rule.
So when i make detector or pick the detector; i tend to pick one with working frequency within mentioned range.
My XP Deus is stuck to 7kHz for years. Gives me best performances on coins.

[waltr]

Very good observations Ivconic.
What target you want has a lot to do with which detector and frequency.

Here in the USA the most desirable coin (for the greatest number of detectorist) are the US Silver coins. The 4-8kHz range is excellent for these.
For the Nickle coins the higher frequency of 12-18kHz works better.

[ivconic]

Many of the Bulgarian metal detector manufacturers seems also understand this very well.
Since their detectors are having feature to switch between 6-8kHz and 15-18kHz.
This also brings up the nice memories on Fisher CZ5 and later CZ3D.
Its range is 5-15kHz. And it preforms hell good on my desired targets too.
Hell... this could be good idea for those who are still left there in FT!
Take CZ5 and just reduce the "ironmongery" at it. Make it ultra light weight.
No customized COG displays. no membrane buttons, no monkey business.
Just make it "vintage" on handling, few turning knobs and good audio.
No need for jingle&bells. Keep the same CZ5 soul inside.
So nice idea. Almost like a dream.
Problem is there is nobody left there, able to materialize it.
I don't mind putting any criticism here.
Obviously DJ is gone totally senile. And the rest of guys with "diplomas" there are just wasting their and our precious time.
According to their latest products; they exist there only to... exist.
CZ3D was the last real product coming from that kitchen.
Ever since than; only hardest mediocrity left there. No ideas. No clues. Just exist there to collect the fees.
If i was not totally disgusted by DJ character; i would contact him and give him few indeed valuable suggestions.
But anyway it is too late. Eastern Europe enthusiast in meantime evolved in real md developers.
We don't need overseas knowledge anymore. We can do that on our own.
Certain era is ended finally.

[ivconic]

If you ask me; main things today are happening in Bulgaria and Turkey.
Fresh blood. Fresh ideas. Lot of indeed talented people there.
Everything what's really counts is happening there.

[Carl-NC]

Problem is there is nobody left there, able to materialize it.
I don't mind putting any criticism here.
Obviously DJ is gone totally senile. And the rest of guys with "diplomas" there are just wasting their and our precious time.
According to their latest products; they exist there only to... exist.
CZ3D was the last real product coming from that kitchen.
Ever since than; only hardest mediocrity left there. No ideas. No clues. Just exist there to collect the fees.

Ouch.

[ivconic]

Ouch.

That's how we do the motivation speech here in Serbia.
And guess what; here it works!
Don't know if will have any results there...

[Danno]

Check in on the thread about the frequencies used in the NOX detector.
Guessing they have worked out your question to choose frequencies.

Doesn't mean the Equinox steps are optimum though,