TGSL Tuning & troubleshooting

[Waikiki_Sweep]

If we building metal detector then we have to use as less metal inside search head as we can.

Good info about mylar. But why better than aluminum tape? Mystery of coils...

If we building metal detector then we have to use as less metal inside search head as we can. Otherwise this metal becomes a target or masking the target.

Mylar contains 100 times less metal then aluminium foil and it is not so conductive.
So we have:

lower phase shift
lower eddy curents
lower inductance
lower capacitance

Also other elements like coaxial cables should be lighter too. (less metal, lower diameters)
Similar effect was with Surp PI project. Aluminium foil shild, bulky coaxial cable were masking targets, decreasing detecting range, prolonging delays, masking gold and small targets.

If someone have wires with lower diameter AWG 34, 40, 44 may be it is good idea to try it for TGSL because by lowering amount of metal we may have more sensitivity and detection distance, more stronger phase shift and discrimination for smaller targets.
Instead of PI detectors (several Ampers in pulse) TGSL using much less current (miliamperes) so higher resistance of wires is not so impotrant.

[simonbaker]

Good info about mylar. But why better than aluminum tape? Mystery of coils...

[Dennis the Mennis]

Mylar shielding as standard for coil shielding material

After several tests I am sure that it is not necessary to alter TGSL schematics elements values.

( No needs to add capacitor 4700 pF across TX coil.)

Also once I have removed aluminum tape shielding situation gets better significantly.
Now I can detect small aluminum pull-tab from coca can at 25-30 cm even in high EMI noise.

Shielding was made from aluminum tape according all instructions regarding gap, grounding and insulation but it fails.

Aluminum tape was looking so thick and may be it was significant influence on inductance or phase.

With shielding:
1. chattering or motorboating in all range of sensitivity for all metal mode and half range of sensitivity for discrimination mode.
2. 10-15 cm for 1eu coin 15-20 cm for pull-tab (High rise building EMI noise)
3. Very sharp + - 1 mm. nulling position where MD detecting (approx 85% of full swing of preamp output that you can see without nulling).

Without shielding:
1. chattering half range of sensitivity for all metal mode and no chattering for any position of sensitivity for discrimination mode.
2. 20-25 cm for 1eu coin 25-30 cm for pull-tab (High rise building EMI noise)
3. Wide range of coils nulling position (from zero to signal full swing of preamp output) allows to detect with moderate results.

So much difference in characteristics with shielding and without it.

Also it is impotrant to have a set of targets from different metals for testing because in some phase positions you can detect some metals (aluminum) for a very high distance and can't see some coins at all.

I will try mylar shielding (survival blanket from Wall-Mart: $3.30. I have checked that it gives 300 ohm for 1 meter for 2 cm stripe).

Also may be some layer of insulation between coil and shielding is necessary to lower capacitance and inductance of shielding.

Some questions may arise regarding TL071 (I have received in Silverdog kit instead of LM308. I am not sure that it is good replacement because I can see -1.0 .. -1.5 mV shift at the output of it so it may reduce sensitivity. I will try to sent some negative voltage to sensitivity pot to increace range of sensitivity settings. (Say connect -5V with 100K resistor to lower end of sensitivity pot)

I will continue experimenting. Thanks to all.

Perhaps you can use the Mylar shielding from ( for example) an old VGA cable.
I stripped down one and it was enough for one coil !!

I had the same problems with my Aluminium shielding : I was trying for ages to get everything right .

When I tried the Mylar shielding( as DFBowers decribed in his manual ) lots of problems disappeared ... , don't forget the naked wire .

So all credits to DFBowers !!! Mylar should be the material for shielding !!!!

The TL071's I've tried didn't have a big difference compared to the LM308's under workshop conditions .

The conditions you've described for shielded- and unshielded coils match with my experience ... but the only the shielded coils do matter ( less sensitive ( to noise) ).

The chatters you have in AM mode are still there in DISC mode but most likely they are nulled like in this example were also the target presented is disc'd out or nulled:



kind regards ,

Dennis the Mennis

[Wiltran]

TGSL

Waikiki, I bought mylar for shielding at a dollar store for a buck!
Its called Hologram gift wrap. you get 14 sq. ft. in the pack. I checked it with an ohm meter, and its very thin - will probably work perfectly.

[Waikiki_Sweep]

Thick aluminum tape shielding was a source of noise :-(

After several tests I am sure that it is not necessary to alter TGSL schematics elements values.

( No needs to add capacitor 4700 pF across TX coil.)

Also once I have removed aluminum tape shielding situation gets better significantly.
Now I can detect small aluminum pull-tab from coca can at 25-30 cm even in high EMI noise.

Shielding was made from aluminum tape according all instructions regarding gap, grounding and insulation but it fails.

Aluminum tape was looking so thick and may be it was significant influence on inductance or phase.

With shielding:
1. chattering or motorboating in all range of sensitivity for all metal mode and half range of sensitivity for discrimination mode.
2. 10-15 cm for 1eu coin 15-20 cm for pull-tab (High rise building EMI noise)
3. Very sharp + - 1 mm. nulling position where MD detecting (approx 85% of full swing of preamp output that you can see without nulling).

Without shielding:
1. chattering half range of sensitivity for all metal mode and no chattering for any position of sensitivity for discrimination mode.
2. 20-25 cm for 1eu coin 25-30 cm for pull-tab (High rise building EMI noise)
3. Wide range of coils nulling position (from zero to signal full swing of preamp output) allows to detect with moderate results.

So much difference in characteristics with shielding and without it.

Also it is impotrant to have a set of targets from different metals for testing because in some phase positions you can detect some metals (aluminum) for a very high distance and can't see some coins at all.

I will try mylar shielding (survival blanket from Wall-Mart: $3.30. I have checked that it gives 300 ohm for 1 meter for 2 cm stripe).

Also may be some layer of insulation between coil and shielding is necessary to lower capacitance and inductance of shielding.

Some questions may arise regarding TL071 (I have received in Silverdog kit instead of LM308. I am not sure that it is good replacement because I can see -1.0 .. -1.5 mV shift at the output of it so it may reduce sensitivity. I will try to sent some negative voltage to sensitivity pot to increace range of sensitivity settings. (Say connect -5V with 100K resistor to lower end of sensitivity pot)

I will continue experimenting. Thanks to all.

[Dennis the Mennis]

Dear Simon

1. Does the sensitity pot affect the cracking sounds?
No ( Noise could be that high that the sens pot has no effect )

2. Are the cracking sounds constant, or only when detecting a target?
constant ( Motorboating ? see example )

3. Where are you testing? Are you far away from all EMI sources?
Indoor ( go outside !)

Hi MD107 ,

Perhaps another thing to check is to get away from all indoor noise , go outside far way from all EMI sources !

You would be surprised !

In my case the noise in my house is very inpredictable and the noise average is around 30 -40 mV top top (I think it's too high) , even in my garden [10 meters away from my house] there is still a high noise ...

I have a small advantage of having an USB scope ; here is an example of noise measured on the outputs of the LM308's.
Please note that in my case ( the TGS ) the system starts to crack around 25/30 mV top-top and starts to alarm around 40mV t-t



I even caught 'motorboating' in a graph : due to the inpredictable noise you get 5 or 6 alarms a second continuously ( in this specific example)



Are you also sure you have to correct voltages as Simon asked you in previous posts : very important !

Would you measure the voltages on C12 and C15 (relative to ground) for these cases?

1. Voltage on C15 (disc channel) for:
a. Disc pot at min setting.
b. Disc pot at middle setting.
c. Disc pot at max setting.

2. Voltage on C12 (ground channel) for:
a. GB pot at min setting.
b. GB pot at middle setting.
c. GB pot at max setting.

Total six measurements.

Regards,

-SB

kind regards ,

Dennis the Mennis

[metaldetector107]

TGSL-EDU

Dear Simon

1. Does the sensitity pot affect the cracking sounds?
No

2. Are the cracking sounds constant, or only when detecting a target?
constant

3. Where are you testing? Are you far away from all EMI sources?
Indoor

[simonbaker]

Dear Friends

I have built my second TSGL-EDU, this time performance very good
30cm for silver ring, but same problem i have like first device.
i cant make it quite it cracks please help.

kind regrads

1. Does the sensitity pot affect the cracking sounds?

2. Are the cracking sounds constant, or only when detecting a target?

3. Where are you testing? Are you far away from all EMI sources?

Congratulations on making an MD with 30 cm depth!

Regards,

-SB

[porkluvr]

Dear Friends

I have built my second TSGL-EDU, this time performance very good
30cm for silver ring, but same problem i have like first device.
i cant make it quite it cracks please help.

kind regrads

See what happens to insert a 10pf capacitor across U101A pins 1~2.

Also, what kind of FETs are you using? Wrong FET might not turn off completely. That would be bad.

What is magnitude of negative power supply voltage?

edit: Of course you're using quality capacitors for C9 and C10? (NPO ceramic, or else film?)

[porkluvr]

I have cracks and detection range so poor too. When I have added capacitor 4700 pF across TX coil cracks stops and detection range rises from 10 cm to 25 cm. Frequency drops from 14,500 to 12,500 or so. May be it is high EMI noise in my apartments and change in frequency and phase has helped to reduce it. I need more experiments outdoor somewhere in low noise place with the capacitor and without it.

If you lower tx frequency and it increases detection range, what happens if you put tx frequency back where it should be and then replace C6 with slightly lower value? (First must measure C6.)

Also, read my next post.

[metaldetector107]

TGSL-EDU

Dear Friends

I have built my second TSGL-EDU, this time performance very good
30cm for silver ring, but same problem i have like first device.
i cant make it quite it cracks please help.

kind regrads

[Waikiki_Sweep]

When I have added capacitor 4700 pF across TX coil cracks stops and detection range rises to 25 cm

I have cracks and detection range so poor too. When I have added capacitor 4700 pF across TX coil cracks stops and detection range rises from 10 cm to 25 cm. Frequency drops from 14,500 to 12,500 or so. May be it is high EMI noise in my apartments and change in frequency and phase has helped to reduce it. I need more experiments outdoor somewhere in low noise place with the capacitor and without it.

[simonbaker]

voltage measurements

Dear Simon

its 15n

kind regards

Would you measure the voltages on C12 and C15 (relative to ground) for these cases?

1. Voltage on C15 (disc channel) for:
a. Disc pot at min setting.
b. Disc pot at middle setting.
c. Disc pot at max setting.

2. Voltage on C12 (ground channel) for:
a. GB pot at min setting.
b. GB pot at middle setting.
c. GB pot at max setting.

Total six measurements.

Regards,

-SB

[metaldetector107]

tgsl

Dear Simon

its 15n

kind regards

[simonbaker]

Dear Simon


c6 0v
c12 650mv
c15 116mv

rx 6.30 26ohm
tx 5.85

thanks

Thanks, those look OK.

Can you measure the capacitance of C6?

Regards,

-SB