TGSL Tuning & troubleshooting

[Stefano]

I've checked components around pre-amp and they're ok.
So I've tried to re-adjust coil nulling: I've made a 7 turns-5cm loop and connected in series with rx coil.
Then moving it I' ve verified that I'm able to change phase shift and tension, but noticed that the minimum voltage doesn't correspond to the correct phase shift.
I've found a position where I get 4us delay in the rx signal respect to tx which should be 20 degrees (70/360*20)=3.8 but neither in this position nor in any other, I'm able to get a full good working. I mean I have always not a single but a double or even triple sound if target is swept slow and near the coil, and big iron is not completely rejected in discrimination mode.
Noticed also that ground connection of Rx coil shifts a bit the phase:
increments if connected to purple wire and decrements if connected to grey wire.
Detection depth is bad as before too.

So:
Is the method I used with the coil correct to fix the phase and offset of possibly not good original coil?

Is it correct a time difference of 4us for a DD coil (measured at output of pre-amp) for a required phase diff of 20°?

Which side of Rx coil should I connect to ground?

Should a DD coil give just a single tone for each target sweep or is it normal have a double tone?

[Qiaozhi]

I'll check again components value. If the phase shift is more than 20 degrees the wrong component should be around the amplificator U101a, since I scope its output on pin 7. Is that correct?

The TX-RX phase shift is determined by the LC tank in the TX and RX circuits, and by the coil alignment. If you are using an original Tesoro coil, then there's not a lot you can do there, but you could check the TX and RX tuning caps, and the components around the pre-amp. If everything looks correct, then try a different coil.

[Stefano]

The only thing that concerns me is the initial phase-shift between TX and RX. It appears to be greater than 20 degrees.

I'll check again components value. If the phase shift is more than 20 degrees the wrong component should be around the amplificator U101a, since I scope its output on pin 7. Is that correct?

[Stefano]

I have some more pictures.
Only a strange problem in triggering with discrimination at 0% present only when triggering with Ch.1 Was ok if I set trigger on Ch2.
Anyway I'll post here what I saw with trigger Ch.1

[Qiaozhi]

From your photos everything looks more or less ok. The only thing that concerns me is the initial phase-shift between TX and RX. It appears to be greater than 20 degrees.

If it is indeed correct, then the problem must be further along in the design. Presumably you have carefully checked all your component values?

[Stefano]

This is the my situation at the moment:
The sensitivity is too low, and adjusting the sensitivity trimmer doesn't change actual sensitivity much. I get 13-14 cm for a 10 euro cent and 5 US cent (1981) and 17 cm for 1 euro coin.
In both all metal and discrimination mode I get a single sound when the target is far from the coil, but 2 or sometimes 3 sounds if the target is swept near the coil.
Medium size iron target (like a pair of scissors) are not completely rejected when discrimination is at minimum.
All coins are completely rejected with discrimination at 100%.

[Stefano]

I 've added the missing oscillator Fet amplitude stabilization circuit to TGSL and honestly now the output of U101blooks much better but still not perfect, so I added also a 18pF in parallel to R17. I've used a J310 because was the only fet I had at home today.
This is the waveform now.

[Stefano]

have you made a connection in the search head between the TX and RX coils?

Yes, connected purple wire to screen inside the coil shell.

[Qiaozhi]

It's a DD coil. (Laser widescan). I just removed the center tap on it.
I don't understand your statement Qiaozhi, I thought that the phase difference should be always the same, no matter what coil I connected..., anyway looking back to the image I posted I've just realized that they are actually showing a wrong tx-rx phase difference. I know that they should be 200° shifted and they are not.

The 200° phase shift is only correct for a concentric coil. For a DD coil it should be 20°. Your TX-RX phase relationship looks ok to me, but there are some other things you need to check:

1. All targets (ferrous and non-ferrous) should cause a phase-shift to the right, relative to the TX signal.
2. Ferrous targets should cause an increase in amplitude.
3. Non-ferrous targets should cause a decrease in amplitude.

The above is correct for a DD coil. Everything is reversed for a concentric.
Also (since your coil had a center-tap, which has been removed) have you made a connection in the search head between the TX and RX coils?

[Stefano]

Stephan did not say outright that his connection was backward but I read his statement as an inferral because of how it was worded.

I hoped my english was better

If it's a DD coil, then the TX-RX waveforms look correct. If it's a concentric, the RX connections need to be swapped over.

It's a DD coil. (Laser widescan). I just removed the center tap on it.
I don't understand your statement Qiaozhi, I thought that the phase difference should be always the same, no matter what coil I connected..., anyway looking back to the image I posted I've just realized that they are actually showing a wrong tx-rx phase difference. I know that they should be 200° shifted and they are not.
While the very first of my pictures was just to show you mostly the nulling of the coil, image #1 is actually wrong.
As I wrote, the circuit is not working good with either of rx polarity, but that images unfortunately don't help you to help me.
I'll take new pictures with the correct 200° phase difference a.s.a.p.

One thing I have not clear yet: what should I get at the output of U101b pin1: a sine or a square wave? Should it shift rotating the discr. pot.? How much?

[Qiaozhi]

Absolutely! Tx should be in the region of 5.6mH, and rx is about 6.2mH.

Stephan did not say outright that his connection was backward but I read his statement as an inferral because of how it was worded.

If it's a DD coil, then the TX-RX waveforms look correct. If it's a concentric, the RX connections need to be swapped over.

[porkluvr]

I think you meant mH, not uH.

Absolutely! Tx should be in the region of 5.6mH, and rx is about 6.2mH.

Stephan did not say outright that his connection was backward but I read his statement as an inferral because of how it was worded.

[Qiaozhi]

On my 8" concentric coil, RX is 6.2uH and TX is 5.6uH. You should try reversing your TX and RX coils. That will bring your frequency up if you have them backwards.

I think you meant mH, not uH.

[porkluvr]

On my 8" concentric coil, RX is 6.2uH and TX is 5.6uH. You should try reversing your TX and RX coils. That will bring your frequency up if you have them backwards.

[Stefano]

With standard I mean around 20 ohms the rx and tx resistance (I don't remember exactly) but I remember exactly 5.97 mH and 6.57 mH measured with the homebuilt LCF meter project published on Elektor.
I think too that there's something strange around U101b, also because I was reading the "TGSL Experiments" thread which I had strangely not noticed before where this issue and the lack of the oscillator stabilization components is discussed.
One more doubt is about the not perfect linearity of the tx signal waveform.
I'll make more tests in these holiday days.
Certainly a bunch of reliable waveforms of a working tgsl would be useful...