Vallon VMH3CS Mine Detector

[Ferric Toes]

I've ordered a LME , will have a play.

Some time ago, I managed to get some 8 pin DIP boards with 2 LME49990 mounted. This could be used then as a dual stage preamp which has definite advantages. If I remember correctly, the current consumption was 30mA for the two which was too much of an increase over the NE5532 which I otherwise use.

Eric.

[Ferric Toes]

Hi Eric

Made a chart to compare the differences with 1k and 25k pulse rate. Wasn't expecting average current to be lower at 25k. Signal is proportional to peak current. Do you have a formula for reduction in noise for pulse rate increase? 10k/1k, 25k/1k

12uS is much too short for a useful TX pulse. With a 300uH coil, all you have is is a sharp sawtooth current waveform in which the eddy currents at switch on will seriously cancel those at switch off. I would increase the TX to at least 50uS which will then put up your current draw to respectable values. TX pulse width should be at least 5 x the TC of the objects you are looking for. 50uS is still short but OK if you are looking for tiny objects or thin gold chains.

There is no need to go to extreme pulse rates except for special applications, as I was doing. For coins, rings etc 5 - 10k pps is sufficient but keep the TX current down to manageable values, and a low current pulse is easier to switch off quickly for short delays.

In designing a detector for a wide range of applications/object sizes, I usually use a contant current PSU circuit for the TX. Shortening or widening the TX pulse, and/or changing the pulse rate always gives the same current draw from the batteries.

Eric.

[daverave]

12uS is much too short for a useful TX pulse. With a 300uH coil, all you have is is a sharp sawtooth current waveform in which the eddy currents at switch on will seriously cancel those at switch off. I would increase the TX to at least 50uS which will then put up your current draw to respectable values. TX pulse width should be at least 5 x the TC of the objects you are looking for. 50uS is still short but OK if you are looking for tiny objects or thin gold chains.

There is no need to go to extreme pulse rates except for special applications, as I was doing. For coins, rings etc 5 - 10k pps is sufficient but keep the TX current down to manageable values, and a low current pulse is easier to switch off quickly for short delays.

In designing a detector for a wide range of applications/object sizes, I usually use a contant current PSU circuit for the TX. Shortening or widening the TX pulse, and/or changing the pulse rate always gives the same current draw from the batteries.

Eric.

Hi Eric...If i were to use a TX frequency of say 10 khz would that be good for most gold rings as im not really interested in finding coins...also what is the ideal TX width and sample width at say 10 khz to find most size gold rings ????

[6666]

Some time ago, I managed to get some 8 pin DIP boards with 2 LME49990 mounted. This could be used then as a dual stage preamp which has definite advantages. If I remember correctly, the current consumption was 30mA for the two which was too much of an increase over the NE5532 which I otherwise use.

Eric.

Hi Eric, I managed to find some mounted on 8 pin dip boards, but only ordered one just to have a play, but 30ma is a bit high.

There are quite a few variations on the market, but the one I chose uses the MT3608 converter.

My MT3608 converters have arrived, nice little board will have a test of them later today.

[Teleno]

what is the ideal TX width and sample width at say 10 khz to find most size gold rings ????

Gold rings are bulky, high inductance and medium resistance targets. They're extremely easy to detect because their time constants are medium-long. No special measures need to be taken, no short delays or pulses. Just make sure your pulse is long enough, say > 50us.

A different matter is hunting for tiny gold, such as nuggets or chains. Low inductance and high resistance means short TC, refinements are required.

[daverave]

Hi Teleno....in my last detector i used a tx pulse of around 40 uS and a tx frequency of over 3 khz and i can get good depth on very small gold ring and chain and on my 9ct gold wedding band i also get good depth but alas i cannot test on 14/18/22 ct gold cause i dont have those at hand to test...need to find some

[Ferric Toes]

Gold rings are bulky, high inductance and medium resistance targets. They're extremely easy to detect because their time constants are medium-long. No special measures need to be taken, no short delays or pulses. Just make sure your pulse is long enough, say > 50us.

A different matter is hunting for tiny gold, such as nuggets or chains. Low inductance and high resistance means short TC, refinements are required.

Gold rings do vary a lot though, depending on the Carat rating. 24 is obviously best, but many rings are made of lower carat gold and are less detectable. The alloyed metal is usually silver and/or copper and even 10% of either causes the conductivity to drop through the floor, giving the ring a much faster decay. It seems strange that both silver and copper are higher conductivity than pure gold, but when alloyed with gold the end result is a poorer conductor. Platinum rings are also worth finding, but this metal has a conductivity of only 16% that of pure copper, so again, depending on the cross section, the decay can be quite fast.

Gold nugget hunting is interesting as again, it depends on where the nugget originated. Western Australian nuggets are less detectable that nuggets from the Victorian goldfields. This is because WA nuggets have a small amount of alloyed silver which, size for size, reduces the conductivity. Their appearance is also slightly more 'brassy' than the Vic. counterparts.

Eric.

[daverave]

Hi Eric....very interesting what you said about different carat rating....i use a sample pulse of only around 13 uS and my depth on very tiny gold rings and chains is very good...im not sure about 14/18/22/24 ct as i dont have any above 9ct....would it be better to say increase my sample to around 20 uS so to be better on larger rings ???

[Ferric Toes]

My MT3608 converters have arrived, nice little board will have a test of them later today.

Yes, there are a few problems I have come across though. The multiturn trimmer for adjusting the output voltage has quite a bit of backlash, so turning it one way, and then back, the action is not immediate.

Mine will not run on one of my bench power supplies, but the other (older) one it is OK. Batteries work OK

Small box shielded with copper tape seems effective in stopping radiated noise but I had a bit of a problem with conducted noise. Some good quality capacitors on input and output leads fixed that, together with a linear regulator and low esr electrolytic cap on the final output.

What I am doing here is taking the three series Li-ion batteries and setting the converter output to 15V. This I then regulate to 12V for the TX circuit. I need to do this so that the TX voltage remains constant over the range of battery voltage otherwise the calibration of my viscosity meter will not hold.

Eric.

[Ferric Toes]

Hi Eric....very interesting what you said about different carat rating....i use a sample pulse of only around 13 uS and my depth on very tiny gold rings and chains is very good...im not sure about 14/18/22/24 ct as i dont have any above 9ct....would it be better to say increase my sample to around 20 uS so to be better on larger rings ???

Hi David,

I presume you mean sample pulse delay? If so, I would stick with the 13uS as you are sampling higher up the decay curve with any target. What might be beneficial for larger rings is to increase the TX pulse width to, say, 100 or even 150uS. This may mean dropping the pulse frequency otherwise you will be pulling too much current from the batteries and/or some internal parts might overheat. Staying at 13uS delay, this other change should not affect the detection of small or less conductive rings.

Eric.

[Teleno]

What might be beneficial for larger rings is to increase the TX pulse width to, say, 100 or even 150uS. This may mean dropping the pulse frequency otherwise you will be pulling too much current from the batteries and/or some internal parts might overheat.

By using cosntant current Tx and a balanced coil you can sample during on-time and no time constants will be atenuated. Soil response will then be the closest to 1/t.

[daverave]

Hi David,

I presume you mean sample pulse delay? If so, I would stick with the 13uS as you are sampling higher up the decay curve with any target. What might be beneficial for larger rings is to increase the TX pulse width to, say, 100 or even 150uS. This may mean dropping the pulse frequency otherwise you will be pulling too much current from the batteries and/or some internal parts might overheat. Staying at 13uS delay, this other change should not affect the detection of small or less conductive rings.

Eric.

Hi Eric....not the delay but the 1st sample pulse width ive set to around 13 uS and very good on very small rings and chains....My tx pulse width is only about 42 uS...the response on small gold is very good....but very poor on silver coins....i wonder would it be a good idea to have a switch for the tx pulse width say 50 uS for small gold and say 100-150 uS for larger gold items...and select whatever gold type your looking for ????

[Ferric Toes]

Hi Eric....not the delay but the 1st sample pulse width ive set to around 13 uS and very good on very small rings and chains....My tx pulse width is only about 42 uS...the response on small gold is very good....but very poor on silver coins....i wonder would it be a good idea to have a switch for the tx pulse width say 50 uS for small gold and say 100-150 uS for larger gold items...and select whatever gold type your looking for ????

Early PI detectors in from 1966 to the mid 1970's had wider TX pulses, longer delays, wider sample pulses, and pulse repetition rates of 100 or less. Typically 350uS TX, 75 -100uS delay, 25uS sample width and 86 pps. They were very good at finding old copper pennies, ha'pennies, threepenny bits, silver sixpences and men's chunky signet rings. Not good on cupro-nickel coins and poor or ladies wedding bands. Advertised features were 15in on an old penny, great insensitivity to silver paper and cigarette foil wrapping, plus salt wet beaches had little effect.

Ideally, the TX width should be 5 x the TC of the highest conductive target you wish to find although 3 x will still give 95% maximum excitation. A short delay time is still beneficial, unless you want to reject poor conductors. Keep the repetition rate as high as you can conducive with battery life and any temperature issues. Early PI's had to be swept slowly because of their low pulse rate and it would be counterproductive to return to that.

Eric.

[daverave]

Hi Eric...i think that i may incorporate a tx pulse switch and choose the type of gold im looking for on the day.....like high/low power switch.
detectors like the whites tdi pro even though they use a lot of power still can be very deep on low conductive items like nickles and still find higher conductive items like heavy rings at good depth...this detector seems to give the best of both worlds on low and high conductors.

[green]

Hi Eric....not the delay but the 1st sample pulse width ive set to around 13 uS and very good on very small rings and chains....My tx pulse width is only about 42 uS...the response on small gold is very good....but very poor on silver coins....i wonder would it be a good idea to have a switch for the tx pulse width say 50 uS for small gold and say 100-150 uS for larger gold items...and select whatever gold type your looking for ????

What is your sample pulse delay time?