TGSL troubleshooting questions

[Jerry]

I have built TGSL and I experimented with the coils, especially with TX coil. I tried 2mH coil on TX, and also tried 11mH 4 TX coil. RX picked up all those, pretty similar results with classic standard TX output (oscillator block). i had 10 cm range from coil on 5 US cents and maybe 15 cm for 5 dinars. My conclusion is that no matter which coil for TX is used , nothing 's changed except frequency, with higher inductance i got lower frequency 10 KHz on oscillator and with 2mH i got about 20KHz, also sound were changing on the speaker towards lower and higher. I also changed RX block same as IGSL, but i kept same transistor in oscillator 2N2970. Detection distance stayed same. I wonder how someone can reach 30- even 35 cm in the air with that configuration even with default TGSL (not modified to IGSL) Is that really true or someone is playing with results?

Two thoughts: First, I wonder if you are able to ground balance out a small piece of ferrite with all of the coil combinations? If you cannot ground balance the detector, you will not get maximum performance.

Second, the location where you are doing you testing really needs to be quiet EMI wise or you just are not going to get a good air test.

I have built two TGSL's, both Ivconic and Edu versions and I get about 30 - 35 cm air tests on coins with either. Things like soda cans, much further. However this only translates to about 12 to 16 cm of actual depth in the dirt that I have here. I have found things deeper once in a while, but that is not typical.

Jerry

BTW: I wound up a variety of coils for testing purposes last winter but thus far have not done the experiments. Your experience got me to thinking about that again. Will post some results when I get them.

[zoomtronic]

I have built TGSL and I experimented with the coils, especially with TX coil. I tried 2mH coil on TX, and also tried 11mH 4 TX coil. RX picked up all those, pretty similar results with classic standard TX output (oscillator block). i had 10 cm range from coil on 5 US cents and maybe 15 cm for 5 dinars. My conclusion is that no matter which coil for TX is used , nothing 's changed except frequency, with higher inductance i got lower frequency 10 KHz on oscillator and with 2mH i got about 20KHz, also sound were changing on the speaker towards lower and higher. I also changed RX block same as IGSL, but i kept same transistor in oscillator 2N2970. Detection distance stayed same. I wonder how someone can reach 30- even 35 cm in the air with that configuration even with default TGSL (not modified to IGSL) Is that really true or someone is playing with results?

[der_fisherman]

The very front end of IGSL still sucks. It is designed as a differential amplifier, and a noisy one. By simple reconfiguring the front end resistors it could become a non-inverting amp with some 10dB less noise. This simple reconfiguration does not satisfy my needs as I am using a differential mode coil, hence I need a differential mode improvement. I have a candidate up my sleeve, but until I test it - it stays with me.

The other thing to apply is a wide range separate GB channel. There are 4 channels in all available in IGSL, and my mods so far use 3 channels. 2 of which are discrimination channels for Fe and Cu, and the 3rd channel is all-metal/GB combo. I realised that joint GB and all metal is not optimal solution, at least not for difficult terrains and especially not for beaches, so I'll use the remaining channel for a dedicated GB function. I expect to have much better discrimination, and GB ready for every imaginable terrain.

I just wanted to say simply thanks for all your work. Once the design is fully finished, I want to update mine to that standard, its pretty useless at this time, so I bought a complete detector in the UK last year....but I want my own built one to work as well. Till then its in the loft!!

Regards

Andy

[Davor]

The very front end of IGSL still sucks. It is designed as a differential amplifier, and a noisy one. By simple reconfiguring the front end resistors it could become a non-inverting amp with some 10dB less noise. This simple reconfiguration does not satisfy my needs as I am using a differential mode coil, hence I need a differential mode improvement. I have a candidate up my sleeve, but until I test it - it stays with me.

The other thing to apply is a wide range separate GB channel. There are 4 channels in all available in IGSL, and my mods so far use 3 channels. 2 of which are discrimination channels for Fe and Cu, and the 3rd channel is all-metal/GB combo. I realised that joint GB and all metal is not optimal solution, at least not for difficult terrains and especially not for beaches, so I'll use the remaining channel for a dedicated GB function. I expect to have much better discrimination, and GB ready for every imaginable terrain.

[sini]

One question for Davor. You talked about Rx side improvement, especially with IGSL. I know IGSL schematic with your modifications. Are these modifications what you are talking about ? Do you know any next IGSL improvement ( on the edge of possibilites ) ?

[Davor]

Or make. Comes up to the same thing. Front end must be adjusted for a given coil.

[der_fisherman]

I have tried to find musketeer coil making on the forum, and i got nowhere, can someone please direct me where is that thread.
Thanks in advance

I think you are meant to buy a finished search head from Musketeer, I believe a slightly different set up and/or parts are needed if I remember correctly for that search coil.

[Davor]

You can find Mustafa's solution at 12'th page of IGSL topic, http://www.geotech1.com/forums/showt...61-IGSL/page12

In short, Tx at about 1-1.2mH <3ohm. Rx at about 16mH <50ohm.

[zoomtronic]

I have tried to find musketeer coil making on the forum, and i got nowhere, can someone please direct me where is that thread.
Thanks in advance

[Davor]

Up to about 100mV and using a low THD op amp you are OK, so yes, 120mV is still OK. If you notice lack of depth you'll know where to fix it.

Is IGSL Musketeer coil much more sensitive than IGSL TGSL ?

Numbers say it should be, but I didn't compare them so it is a definite maybe
The strongest side to the Musketeer coil is a Tx at ~1mH that allows for much stronger current (and field). Rx side of the most detectors is a place of possible significant improvement, especially with IGSL. The front end of IGSL sucks. Big time. It is over 10dB noisier than achievable with the components used in it's design, and a performance bottleneck.

[Jerry]

If i can not get lower than 120mv on pin 7, is the problem in the coil? 'cause Rx is 6.55mH Tx is 6.05mH, and i suppose nulling is directly connected to the detection distance from coil?

The voltage at pin 7 has been amplified X 47 from the coil input. So with 120mv on pin 7, the actual null voltage at the coil should be about 2.5 mv which should be good.

Jerry

[zoomtronic]

If i can not get lower than 120mv on pin 7, is the problem in the coil? 'cause Rx is 6.55mH Tx is 6.05mH, and i suppose nulling is directly connected to the detection distance from coil?

[sini]

Thanks, Davor,
very well explained. I will try to re-null my coil by small piece of foil ( I will try to achieve zero voltage on preamplifier output ). You helped me much to underestand principle of nulling. Is IGSL Musketeer coil much more sensitive than IGSL TGSL ?

[Davor]

There are several things operating at angle, and it is good to know what is refereed to. We are talking about coil balance...

At a coil you have only Tx and Rx windings arranged to balance by geometry means. It reduces a so called "air signal" to a minimum. This signal is induced from Tx to Rx windings exactly as transformers do, but balanced in a way that it is at a minimum. In case Rx coil is not loaded at all, the "air signal" voltage would be exactly in phase with Tx voltage, or counterphase, but not at angle. Because of capacitive coupling etc. there is a very small component that is orthogonal to the Tx phase that appears at Rx coil, and it is a reason you can't achieve zero volts balance. At least not by squeezing and stretching the coils, playing with a curl of wire, or a piece of ferrite. All these methods are tuning the component of the "air signal" that is in phase with Tx.

You can introduce a piece of foil in vicinity of your coil, provoking induction of "air signal" that is at angle against Tx by means of eddy currents. Thin foil eddy current response is near ideal 90° so it is capable of nulling your coil deeper than achievable by stretching/squeezing alone.

The deepest minimum is a key to the rig sensitivity and stability, so it is important to get balance as deep as possible. The static phase is irrelevant for balance.

20° phase shift is a consequence of a loaded coil near resonance. a well balanced coil will remain well balanced regardless of the coil loading and the phase shift that changes with it. Because of loading, all responses are rotated by additional 20° or so, and it is only a function of resonance and loading - not balance. The 20° shift is relevant only for a waving ferrite response (e.g. ground balance).

A well balanced coil has the deepest minimum at any arbitrary angle.

[sini]

Thank you, Davor.
20 cm maximal air detection distance is without chatters. More sensitivity cause audio oscillation ( periodic false detection ). I use IGSL with TGSL coil : Tx 5.7 mH, Rx 6.2 mH . Oscillation frequency is 14.8 kHz. Tomorrow I will play with threshold. I think I rebuild IGSL TGSL coil to IGSL Musketeer coil ( Musketeer coil should have better detection distance). Now I use chips same as described in IGSL schematic from Ivica with pin swap in Fe channel. Davor, please, can you explain me sentence: "By geometry (or a small piece of ferrite) you can null only the components at phase with Tx, but not those at angle". My nulling method is to find minimal residual voltage with proper 20 degree phase shift between Tx and Rx. Thanks