IGSL

[Davor]

It actually looks as a common mode pickup of mains hum. Now, as there are a few options for connecting a coil to the preamp, and there are also a few options for coil realisation, I can only wonder what situation you have there. Whatever you have, using a device away from mains voltage, and on batteries, should eliminate this kind of hum. I can hear your device being a bit nervous, and sound seem binary. All of this will get much quieter outside.

[sini]

You are absolutely right,
in my home is to much interference signal. I am ready to do real tests on field, but in my area there are few interference signals - 750 kW, 639 kHz radio transmitter 8 km from my home, main train corridor, etc..
Thank you for your help

[Davor]

Here is a kind of interim solution to a better IGSL front-end with minimum intervention to the original layout. It considers single ended coil connection to the non-inverting side of a preamp. Because all the resistors are reduced by factor of 4.7, and there is no resistor in series with the non-inverting branch, the input referenced noise drops from over 6nV/sqrt(Hz) to 3.2nV, some 6dB improvement.

The inverting side is configured as a band pass filter, so 50Hz is attenuated for some 40dB more, which is a good thing.

Noise figures assume an op amp with ~7nV/sqrt(Hz) input noise. It could be improved with a better op amp, and by further reducing resistors. Because input DC bias is derived through equal resistors (47k), even OP27 that is notorious for it's problem with common mode noise will perform well.

[golfnut]

myth busted

On the nulling,


I did a check of mine and it was poor at 200mV - thing is, Ive been finding small roman bronze with it - so how does that work.

I also did an air test with some ferrite on my coil to have an 800mV null and it was the same air test distance as the 200mV null case.

S

[Davor]

Assuming an op amp provides 0.002% THD (LM833) and your fundamental is at 200mV, and assuming the most distortion power is shared between 2nd and 3rd harmonics, you may expect some 1/100 000 of 3rd harmonic in output.

Now, consider air signal of 200mV meeting a ground signal of equal amplitude, both submitted to a 1/100 000 3rd harmonic mechanism, after some tedious calculation you get ~240nV of the 3rd order product reflected back to the carrier vicinity. You want to know what it means when referenced to the input, so you need to divide these numbers with gain (47) and you obtain 5nV.

Considering noise over a 10Hz bandwidth exceeds 5nV by a safe margin, you may consider your 200mV a tolerable figure, but only in case of LM833 or better op amp.

NE5532 is the same topology and you can expect comparable results so it should work the same here.

LF353 is the standard front-end of TGSL, and it has 10 times THD of LM833, hence it's sensitivity will drop at smaller values of residual "air signal" than your IGSL with LM833.


THD is typically expressed for 1kHz, and tends to rise rapidly past 1kHz.

..............................................

And now something completely different...

You may consider AM noise of a well built (Tx) oscillator somewhere at -140dBc/Hz at 10Hz due to components noise contribution only, so you would have plenty of headroom before you match input noise. But somehow you are barely within the limits of input noise with 200mV carrier (e.g. amplitude as seen from the preamp input). Let's see why...

Noise performance of LC oscillators is related to the supply rail voltage purity, hence it must be superb, or otherwise...

Let's ponder a second on it. I found that a simple 7805 regulator gives some 40uV of noise in audio band. Thus it is divided by sqrt(20000) and scaled for 7808, normalised to 200mV amplitude, reduced for gain, and we get some ~10nV/sqrt(Hz), again very close and barely exceeding input noise performance. Right here it is clear that a switching regulator would be a killjoy at this place. It is also clear that halving air signal residual would reduce AM noise injection by half.

Filtering a power supply rail using a low noise reference, such as red LED, would improve power rail noise by factor of 10 or so. IMHO it would be a game changer in PI vs. IB depth battle.

[Turjake]

Hello,

My IGSL TGSL is suffering from weak iron channel: moderate sized iron bolt gives clear tone to 20cm then prr prr sound to 25cm. Everything else works fine and depth for 1euro coin is 35cm. I have made stability mod, changed to bipolars and added capacitors to -6.2V and 8V rails around U5. These made the detector much more stable and added little depth but weak Fe channel remains. I have checked components in the Fe block but cant find faults. Is the pin swap mod inevitable at this point? Any troubleshooting tips appreciated. I have scope also but don't have reference to working Fe channel. Thank you.

[Davor]

Fe targets work more intense when pointing at the coil, not laying down. Also there are many targets you assume are made of iron, but are in fact poor conducting alloys.

Pin swap will enable straightforward discrimination, and possibility of joining Sens. control, but it will not influence sensitivity. If you are certain there is lack of sensitivity you better seek the cause elsewhere.

I think pin swap is an excellent mod because it makes Fe and non-Fe discrimination completely independent. I so much like this way of independent MD interface that I intend to keep this approach in my future builds as well.

[liekwang]

?Hai. Davor is you've got a board layout for the file the top side of its component parts..??

[Davor]

Point is - I did not make any changes on the board. Of all my mods the pin-swap is the most demanding, as it requires swapping adjacent pins on two chips.

I'm about to make a few additional changes on my IGSL by autumn and then it will be ready for PCB editing. I'm very happy with IGSL's ability to discriminate both iron and coloured metals so it is very helpful for recovering artefacts (see a picture of a chisel I found a few days ago), but I'm not happy with it's GB range (I want it to be beach-ready), and I'm not too happy with it's Rx front end.

Do you need board layout for the current version of IGSL? You can buy a board at Silverdog's shop if that's more convenient to you.

[der_fisherman]

Point is - I did not make any changes on the board. Of all my mods the pin-swap is the most demanding, as it requires swapping adjacent pins on two chips.

I'm about to make a few additional changes on my IGSL by autumn and then it will be ready for PCB editing. I'm very happy with IGSL's ability to discriminate both iron and coloured metals so it is very helpful for recovering artefacts (see a picture of a chisel I found a few days ago), but I'm not happy with it's GB range (I want it to be beach-ready), and I'm not too happy with it's Rx front end.

Do you need board layout for the current version of IGSL? You can buy a board at Silverdog's shop if that's more convenient to you.

That could be very old indeed. It depends on the metal, the ground where it was buried (how "preserving" the ground is") and whether or not Roman ruled there at any time.......Have you any thoughts on its age?
I find long nails from time to time in the forest near where I live, which is very close to the "Limes" Wall in Germany, one of Emperor Hadrian's works.
The nails appear to be hand forged and the thought is that they were used to hold the wooden Limes wall together....I actually re-bury them now as there are SO many around where I search....

[Davor]

Because there is some structure from ~14th century near by, I guess it must be somewhat younger than that. Other than that there is no way establishing it's age.
Anyway, having both iron and coloured metals tones simultaneously makes it much easier recognising recent thrash from genuine artefacts.

[simonbaker]

In order to detect two metals "simultaneously", I would think they would have to be slightly separated and the signals received at separate times -- otherwise, the two signals would combine into one. Is that not true?

Regards,

-SB

[golfnut]

As the IGSL has 2 receivers which are gated at different times it is capable of resolving coincident target phases.


IDX, TGSL etc if you get an equivalent mass/size of Fe and non Fe, as the phases of the targets give one pos response and one neg response - then the two signals, which are on the same channel - sum to zero and nothing is heard. Some clad coins do this.

S

[Davor]

It simply discriminates two separate quadrants, and these quadrants may overlap creating a "third tone" that is very useful for recognising foil, small gold, and tin plated thrash, as it does not lose sensitivity at the discrimination border. The quadrants can be set apart and then you have a gap in between, or a notch, but this feature is less useful than the previous one. With my gain stage mods the recovery speed becomes so fast that you can hear both tones in succession with iron cored coins, making them easily distinguishable.

I like these features. IMHO it is much better being aware of iron than obliterating it by more common single channel discrimination. A large iron artefact is a good find.

[simonbaker]

It simply discriminates two separate quadrants, and these quadrants may overlap creating a "third tone" that is very useful for recognising foil, small gold, and tin plated thrash, as it does not lose sensitivity at the discrimination border. The quadrants can be set apart and then you have a gap in between, or a notch, but this feature is less useful than the previous one. With my gain stage mods the recovery speed becomes so fast that you can hear both tones in succession with iron cored coins, making them easily distinguishable.

I like these features. IMHO it is much better being aware of iron than obliterating it by more common single channel discrimination. A large iron artefact is a good find.

I agree what it does, and to elaborate that: the coil only detects a single signal with a single amplitude and phase (assuming targets are not spatially separated).

A piece of silver directly on top of a piece of iron will give a signal -- but is that signal unique in some way to identify the two metals? I think not -- the signal could be the same as a single piece of some other metal. You cannot separate phases the way you can separate frequencies, because there are many combinations of two signals which can sum to the same amplitude and phase.

So if you set your disc knobs for "overlap", then a target in the overlap area could be: 1) iron, or 2) silver, or 3) iron + silver. You don't know which.

If you are looking for both iron and precious metals, you don't need discrimination. Why would you need to know? You're going to dig anyway.

That is "devil's advocate" reasoning of course. In reality, I do see some advantage to knowing, particularly when evaluating the size of a target. A 2 meter iron target may mean junk (like a buried car) whereas a 2 meter silver target could be very interesting, and at the same time a small iron target may be interesting. So the extra info is useful there. Also, you "hear" the amount of junk in your search area, which can be useful for guiding your search.

But I think some people may get the impression that the "two-tone" beep indicates two different metals overlapping, which is possible but not guaranteed. Just checking to make sure that is correct understanding.

-SB