TGSL Experiments

[Dennis the Mennis]

Hey Danny -

I was a little hasty with this request. I think the outputs of the LM308s are not really what we are interested in.

We really want to see something closer to the output of the SD mosfets with very little filtering, just enough to show the shape of the target phase pulse as the target passes our coils.

That would require a special circuit I think, although perhaps just looking at the voltage on C15 and C12 with a sensitive scope (probably needs broadband amplification) might be close enough.

Regards,

-SB

Hi Simon ,

Allready made some clips on the outputs of the LM308's Sundaynight or must I say Monday morning ...

I will post them asap , I'm on a short break right now and the GPRS Internet connection
on my provider is terrible slow , I can't upload them with this kind of speed.

I will try to catch some of the other signals when I'm back home.

kind regards ,

Danny

Dennis the Mennis

[simonbaker]

Hi Simon ,

Allready made some clips on the outputs of the LM308's Sundaynight or must I say Monday morning ...

I will post them asap , I'm on a short break right now and the GPRS Internet connection
on my provider is terrible slow , I can't upload them with this kind of speed.

I will try to catch some of the other signals when I'm back home.

kind regards ,

Danny

Dennis the Mennis

Thanks DtM! Those will still be interesting to look at. I'm interested in how the shape of the pulse changes with depth, and also the spectrum.

-SB

[simonbaker]

measure ground effects

Hi Danny, I keep thinking of more experiments that we need your USB scope to help with.

dfbowers was mentioning that we don't understand the "ground signal" very well.

So it seems it would be a good idea to measure the raw signal from the ground/soil and take some data.

I know this wouldn't be easy to do right away, because you would need to hook the probes to a working TGSL and carry a laptop around! Just something to think about for the long term.

Cheers,

-SB

[kacike]

Coil for TGSL - wanted

Don, I wonder if it is possible to get a complete 20cm coil from you, I am in Spain you could fed-ex it over. Send me a PM if you would, letting me know costs etc. regards kacike

[Dennis the Mennis]

YouTube clips of the target response at different depths

Thanks DtM! Those will still be interesting to look at. I'm interested in how the shape of the pulse changes with depth, and also the spectrum.

-SB

Hi all ,

Finally back in town and finished the YouTube clips .

These clips show the target response ,of an one Euro coin , at different depths on the outputs of the LM308's.

The TGS is connected to a potted coil ( wet grass mod ) with a Belden 8723 cable , AM mode , sens at max .

I've made a 2 meter long sling , connected it to the ceiling light switch (pull cord) and placed the 1 euro coin in a plastic tube.

It was pretty fast clear that this setup was too light : the speed of the coin reduced too fast and made a piece of heavy wood underneath the tube.

This gave a better 'continuous' speed wich is allmost impossible ...

The first couple of samples at the lowest depth give a double beep .

I've started with scope view only for couple of samples with some measurements at the bottom for channel a and b and let the coin pass a couple of times ,
starting point each time the same .

Did this 'procedure' again for the scope and spectrum view combined.
In the spectrum view the response stays around the 5 Hz.

The average responses at the different depths : ( calculated by the scope after some capture counts)

5 cm = 3 v pp
7,5 cm = 1,6 v pp
10 cm = 0,75v pp
12,5cm = 400 mV PP
15 cm = 200 mV PP
17,5cm = 120 mV PP
20 cm = 75 mV PP
22,5cm = 65 mV PP
25 cm = 35 mV PP
27,5cm = 30 mV PP

I hope this test is a little representative , offcourse the reponse depends also of the target speed

Please set your YouTube settings to HD and Full Screen for best results

5 cm : http://youtu.be/fYJMOuhKpmM?hd=1

7,5 cm : http://youtu.be/40yblftMv04?hd=1

10 cm : http://youtu.be/olyvEAUgw18?hd=1

12,5cm : http://youtu.be/bA-USpuxZnk?hd=1

15 cm : http://youtu.be/AjMeVsD84mk?hd=1

17,5cm : http://youtu.be/PDwAh3T1ueM?hd=1

20 cm : http://youtu.be/PwCsuL8CWP0?hd=1

22,5cm : http://youtu.be/gooaZpW3RYI?hd=1

25 cm : http://youtu.be/EzhTG2Bx-ho?hd=1

27,5cm : http://youtu.be/j6otSxzkxSk?hd=1

Simon , I'm so glad you love overkill


kind regards ,

Danny

ps : is it normal YouTube sends bills ?

[simonbaker]

Hi all ,

Finally back in town and finished the YouTube clips .

These clips show the target response ,of an one Euro coin , at different depths on the outputs of the LM308's.

The TGS is connected to a potted coil ( wet grass mod ) with a Belden 8723 cable , AM mode , sens at max .

I've made a 2 meter long sling , connected it to the ceiling light switch (pull cord) and placed the 1 euro coin in a plastic tube.

It was pretty fast clear that this setup was too light : the speed of the coin reduced too fast and made a piece of heavy wood underneath the tube.

This gave a better 'continuous' speed wich is allmost impossible ...

The first couple of samples at the lowest depth give a double beep .

I've started with scope view only for couple of samples with some measurements at the bottom for channel a and b and let the coin pass a couple of times ,
starting point each time the same .

Did this 'procedure' again for the scope and spectrum view combined.
In the spectrum view the response stays around the 5 Hz.

The average responses at the different depths : ( calculated by the scope after some capture counts)

5 cm = 3 v pp
7,5 cm = 1,6 v pp
10 cm = 0,75v pp
12,5cm = 400 mV PP
15 cm = 200 mV PP
17,5cm = 120 mV PP
20 cm = 75 mV PP
22,5cm = 65 mV PP
25 cm = 35 mV PP
27,5cm = 30 mV PP

I hope this test is a little representative , offcourse the reponse depends also of the target speed

Please set your YouTube settings to HD and Full Screen for best results

5 cm : http://youtu.be/fYJMOuhKpmM?hd=1

7,5 cm : http://youtu.be/40yblftMv04?hd=1

10 cm : http://youtu.be/olyvEAUgw18?hd=1

12,5cm : http://youtu.be/bA-USpuxZnk?hd=1

15 cm : http://youtu.be/AjMeVsD84mk?hd=1

17,5cm : http://youtu.be/PDwAh3T1ueM?hd=1

20 cm : http://youtu.be/PwCsuL8CWP0?hd=1

22,5cm : http://youtu.be/gooaZpW3RYI?hd=1

25 cm : http://youtu.be/EzhTG2Bx-ho?hd=1

27,5cm : http://youtu.be/j6otSxzkxSk?hd=1

Simon , I'm so glad you love overkill


kind regards ,

Danny

ps : is it normal YouTube sends bills ?

OMG totally outstanding!

Really excellent Danny, that is a huge work!

I learned a lot from seeing those plots. The pulse shape looks very similar to some LTSpice simulations I did using a trapezoid input pulse into the first LM358. It confirmed the usefulness of the simulation.

I was surprised how the pulse shape and duration did not change much with depth.... however, I'm kind of glad it doesn't, since it means our electronics don't have to be tuned radically for different depths.

My intuition was that a slower responding circuit might be better for deeper targets. Now I wonder if it would help that much.

The shallow-depth pulse does have a different symmetry around the 0 volt axis (negative pulses larger than positive pulse), presumably because the feedback diodes in the amp circuits limit the positive-going pulse. At deeper depths and weaker signals, there is more vertical symmetry - so it works out fine.

One thing I have long wondered is if there is any practical way to trigger on the leading negative-going pulse instead of the positive pulse, because that would give a less delayed beep. However, because there are two negative pulses, it is a messy circuit to try to design, I have worked on it. For deeper targets, the positive pulse looks cleaner anyway. Just a thought.

I'm not sure what to make of the frequency graphs, other than what you pointed out. It might be interesting to do an FFT of just the pulse waveform by itself, a shallow one without much noise.

I think if you can ever do the same experiment probing the voltages on C12 and C15 (probably need wideband amplification), then the FFT might tell us some more info.

Well, I've got to say you outdid yourself with those outstanding graphs, DtM! Thanks, I'll probably refer to them in the future. That is a good setup you made, maybe a video of your swing would be entertaining.

Oh yes, I envied the thoroughness... you took a load of good data.

Cheers,

-SB

[Dennis the Mennis]

Hi Danny, I keep thinking of more experiments that we need your USB scope to help with.

dfbowers was mentioning that we don't understand the "ground signal" very well.

So it seems it would be a good idea to measure the raw signal from the ground/soil and take some data.

I know this wouldn't be easy to do right away, because you would need to hook the probes to a working TGSL and carry a laptop around! Just something to think about for the long term.

Cheers,

-SB

Hi Simon ,

No problem , I'll keep it in mind , the wheather is getting better and better for a 'field trip''.

Any special requests and testpoints in mind if the moments there ?

I can make a setup from the boot of my car or from the picknick table where I've been before , only there is high pull tab and other polution risk ....

[Dennis the Mennis]

OMG totally outstanding!

Really excellent Danny, that is a huge work!

I learned a lot from seeing those plots. The pulse shape looks very similar to some LTSpice simulations I did using a trapezoid input pulse into the first LM358. It confirmed the usefulness of the simulation.

I was surprised how the pulse shape and duration did not change much with depth.... however, I'm kind of glad it doesn't, since it means our electronics don't have to be tuned radically for different depths.

My intuition was that a slower responding circuit might be better for deeper targets. Now I wonder if it would help that much.

The shallow-depth pulse does have a different symmetry around the 0 volt axis (negative pulses larger than positive pulse), presumably because the feedback diodes in the amp circuits limit the positive-going pulse. At deeper depths and weaker signals, there is more vertical symmetry - so it works out fine.

One thing I have long wondered is if there is any practical way to trigger on the leading negative-going pulse instead of the positive pulse, because that would give a less delayed beep. However, because there are two negative pulses, it is a messy circuit to try to design, I have worked on it. For deeper targets, the positive pulse looks cleaner anyway. Just a thought.

I'm not sure what to make of the frequency graphs, other than what you pointed out. It might be interesting to do an FFT of just the pulse waveform by itself, a shallow one without much noise.

I think if you can ever do the same experiment probing the voltages on C12 and C15 (probably need wideband amplification), then the FFT might tell us some more info.

Well, I've got to say you outdid yourself with those outstanding graphs, DtM! Thanks, I'll probably refer to them in the future. That is a good setup you made, maybe a video of your swing would be entertaining.

Oh yes, I envied the thoroughness... you took a load of good data.

Cheers,

-SB

I'm not sure what to make of the frequency graphs, other than what you pointed out. It might be interesting to do an FFT of just the pulse waveform by itself, a shallow one without much noise

Hi Simon and all others ,

I will make a FFT of one puls waveform and expand it ........ the moment my pcb is working again .

I suffered from some falses for a long time due to - or caused by vibrations in the control wires ( pcb to front ) especcialy the wires to the AM / DISC switch .

At first they were twisted , changed that to 'normal' parallel running wires but the problem remained .

Changed the wires the a shielded cable ( each wire individual shielded , same cable I've made for the ( to do ) test between coil and pcb .

It seems that the falses are at least 90% reduced but this could also be caused by the short circuit I've had on my pcb while resoldering the cable...... The + 8V rail was directly connected to the output of the LM358

Same strange thing happened when I connected an other type of earplug : this caused also some falses or motorboating : changed the (twisted) wires to single shielde wire.
Don't know the results yet

So back to the repair bench.

I think if you can ever do the same experiment probing the voltages on C12 and C15 (probably need wideband amplification), then the FFT might tell us some more info.

I've tried that in the past and tried it recently again : I think I need at least a canonball to get some response .
Or an 1 euro coin at few millimeters distance .

I have no means for amplication but I'll give it a shot .

Any ideas how I get that canonball hooked up ?

Well, I've got to say you outdid yourself with those outstanding graphs, DtM! Thanks, I'll probably refer to them in the future. That is a good setup you made, maybe a video of your swing would be entertaining.

Oh yes, I envied the thoroughness... you took a load of good data.

You're welcome ,

I knew you'd like the overkill ... this one took a lot of time in total and time flies when you're having none


kind regards ,

Danny

[simonbaker]

Hi Simon and all others ,

I will make a FFT of one puls waveform and expand it ........ the moment my pcb is working again .

I suffered from some falses for a long time due to - or caused by vibrations in the control wires ( pcb to front ) especcialy the wires to the AM / DISC switch .

At first they were twisted , changed that to 'normal' parallel running wires but the problem remained .

Changed the wires the a shielded cable ( each wire individual shielded , same cable I've made for the ( to do ) test between coil and pcb .

It seems that the falses are at least 90% reduced but this could also be caused by the short circuit I've had on my pcb while resoldering the cable...... The + 8V rail was directly connected to the output of the LM358

Same strange thing happened when I connected an other type of earplug : this caused also some falses or motorboating : changed the (twisted) wires to single shielde wire.
Don't know the results yet

So back to the repair bench.




I've tried that in the past and tried it recently again : I think I need at least a canonball to get some response .
Or an 1 euro coin at few millimeters distance .

I have no means for amplication but I'll give it a shot .

Any ideas how I get that canonball hooked up ?



You're welcome ,

I knew you'd like the overkill ... this one took a lot of time in total and time flies when you're having none


kind regards ,

Danny

Yeah, sometimes a PCB gets touchy and they are the devil to know why. I keep trying though.

Not sure which PCB you are using. Some TGSL PCBs had a long, snaking ground trace between the speaker ground and the battery ground. I think the silverdog pcb has a better, more direct, arrangement... anyway, you could try a jumper wire from the speaker ground to the battery ground (yes it is a ground loop, but worth a try) to see if you get less motorboating.

Also, of course the sensitivity pot can affect motorboating, especially if you have the "dfbowers mod" that gives more range to the sensitivity pot.

But sometimes it is something else and can be a mystery.

As for the "cannon ball" -- I was thinking maybe rig up a simple LM358 wideband amp, simple 2 resistors, etc. to amplify the signal on a little breadboard, maybe 100x gain, with big cap to take out DC bias from input. Fun stuff.

Keep up the good work, we'll see...

Cheers,

-SB

[Farside]

Hi all,
Is anyone aware of a discussion around highly optimising the TGSL?

I'm thinking about using the TGSL to target something very specific and reject everything else.

[simonbaker]

Hi all,
Is anyone aware of a discussion around highly optimising the TGSL?

I'm thinking about using the TGSL to target something very specific and reject everything else.

I think you would need to be specific and probably do some actual experiments. There are many ideas but I would have to try them out to trust what is best for a particular target/environment.

The original TGS I think has a "notch" filter for ignoring metals on both sides of your chosen metal. You can find the circuit at the top of the Tesoro Golden Sabre thread. Of course you run the risk of missing good targets whenever you use discrimination, and notch DISC probably misses the most.

I had a theory that a "slower" response (with slower sweep), such as eduardo's choice of resistors and caps in the TGSL, might help with deeper targets -- I don't think anyone has demonstrated this.

Ivconic's IGSL may be very useful also since you can still "hear" certain rejected metals and gives you more information about the signal.

hugo also has a "Two tones TGSL ID" thread for a simpler-to-build but not quite as flexible way as the IGSL to use different tones to characterize a target.

Coil size also can be used to favor certain targets over others -- smaller coils for smaller shallower targets, bigger coils for larger, deeper targets.

Just some ideas. Feel free to be more specific as to what you are trying to find.

-SB

[Farside]

Thanks for he post.

I have in mind locating native silver deposits (placer nuggets and near surface veins). Not typically how silver is naturally found I know...

I think the best approach would be like you say - obtain a collection of nuggets from the area and tune the MD from there.

[simonbaker]

Thanks for he post.

I have in mind locating native silver deposits (placer nuggets and near surface veins). Not typically how silver is naturally found I know...

I think the best approach would be like you say - obtain a collection of nuggets from the area and tune the MD from there.

OK -- my suggestion would be to not use a notch DISC because there probably isn't much trash where you'll be looking. If there is a lot of trash, then notch is OK and ignore the rest of this message.

I would experiment to determine the best DISC pot setting to maximize the DISC channel signal for silver. This would require putting an oscope probe on the LM308 output and swinging a silver target over the coil and turning the DISC pot for maximum pulse. You could also try AM (All Metal) mode and see which gives a larger LM308 output signal.

However, the overall sensitivity of the TGSL is limited by the GB channel signal -- which typically is much less amplitude. I would experiment with turning the GB pot to try to get a greater signal for silver (looking at the other LM308 output).

Of course this will throw off the ground balance. However, in some soils, you might be able to tolerate some ground-balance "detuning" in favor of more sensitivity to silver.

You could even experiment with disabling the GB channel by jumpering one of the filter op amp inputs to pin the GB LM308 output high (watch out you don't draw extra current). This would create a super-sensitive detector optimized for silver, but it would also easily give false signals to ground variations and probably iron. You could play with the sensitivity control to see if you can find a happy medium that is more sensitive to silver than the standard TGSL configuration. Probably not too practical.

You will of course detect some other metals as well with a little less sensitivity, so you'll probably end up having to throw back a whole bunch of gold nuggets that get in your way...

Just another idea to play around with.

-SB

[Farside]

Geat Response thanks Simon! i think I'm going to print this out and file it away in my TGSL folder :-)

[golfnut]

IGSL which is similar to TGSL -


If you get reasonable ground balance - Silver should be detectable at approx 2x air distance that of gold. In my experimentation.

Generally, booms in on Silver at the expense of gold - which seems to be 'phased, further away.


If the GB is set for NO - Falsing on ground - you almost loose gold, sometimes Nickel, but keep some silver..

S