TESORO GOLDEN SABRE

[adem]

thanks IVCONIC

[Dennis the Mennis]

. . . _ _ _ . . .

???
That seems strange. Usually, when all done properly , TGSL should detect 1e coin easily at 31-32cm distance in air (Disc mode, Sense to max, no chatters, ferrite rod 10cm long totaly rejected)..
Be sure that you used LF353, 2x LM308, etc...etc... TX transistor can be 2N2907 or BC327 (many other similar too).
Negative voltage can varry from -6.00 to -6.30V.
???
Strange!?
I only got low distances with not so good nulled coil so far. Doubt in coil remains in your case too.
What about ferrite? Take 10cm long ferrite rod and check if it is fully rejected up to 5cm away from coil surface.. Than check with some small and thin silver coin is it accepted at the same adjustments when ferrite rod is rejected... Inform us here.
Cheers!

Hi ,

I've changed my ceramic caps into the polyester caps but the problem still remains : poor detection distance - max 15 cm on a 2 euro coin ...
I've bought , as you mentioned , a 10 cm ferrite rod and it was totaly rejected : poor detection distance.
Changed the BF245C into B type , no progess ( only Tx voltage dropped so changed it back ).
Used LF353 , LM308 ( even tried TL 081 )2N2907 etc everything according
original.

Max suggested to make another set of DD 25cm coils : problem remains.
The coils are for TX 6,08 mH / 21,7 ohm / 94 windings and for Rx 6,53 mH / 22,8 ohm / 101 windings of AWG 30.( L values according the multimeter and given the values they should be reasonably )
The coils are constrained , shielded , with airgap , correct shielded cable used , ground Rx in search head , isolated in overlap points ...

I've tried to gain the same images as shown on page 57 posted by Max and used these as a lead , I don't know where else to start.
Used in Tx 4,7 nF // 15nF to reach 14,5 Khz and for Rx only 10nF.

scoop image attached of :
- TX - nulling signal direct on Rx leads ( not connected to pcb)
- Tx
- Coils
- Testpoint Schematic for videos

Please find here link for video showing :

Nulling direct on coil (pcb disconnected):
http://www.youtube.com/watch?v=rXLodliOgzc

Nulling on TP2 (see testpoint image attached) :
http://www.youtube.com/watch?v=SyQTNIxPSyI

Nulling on pin6 input LF353 (coil connected to pcb to compare with firdt video):
http://www.youtube.com/watch?v=AyK02Zk1PLY

Nulling on TP4 ( the influence (?) of nulling on DISC LEVEL):
http://www.youtube.com/watch?v=xg615rcykg8

DISC potmeter from CounterWise to CounterClockWise on TP4 :
http://www.youtube.com/watch?v=o8pYD7AQQhs

Nulling on TP5 (the influence(?) of nulling on GroundBalance):
http://www.youtube.com/watch?v=Pn--c863r1k

GEB potmeter from CW to CCW :
http://www.youtube.com/watch?v=RtPP2Qe67jk

Nulling on TP6 switch in ALL METAL position:
http://www.youtube.com/watch?v=u1VqMiKEhS0

Nulling on TP6 switch in DISC position:
http://www.youtube.com/watch?v=Urx633REO8s

DISC potmeter from CW to CCW on TP 6 switch in DISC pos :
http://www.youtube.com/watch?v=VILCDHE5kJg

Nulling on TP7 ( the influence (?) of nulling on GEB ):
http://www.youtube.com/watch?v=htUUla-vk-E

GEB potmeter from CW to CCW on TP7:
http://www.youtube.com/watch?v=Mkzde5mX-c4

TP8 ferrite rod rejected :
http://www.youtube.com/watch?v=PeE4AL4kKh4

TP8 detecting 2 euro coin at max 15 cm :
http://www.youtube.com/watch?v=gB1GnSdN3fs

TP10 detecting 2 euro coin at max 15 cm , >20 cm no response :
http://www.youtube.com/watch?v=sdTP34u3pM8


Is there a possibility that when even the nulling seems fine , the sensitivity is poor ?
So coil related problem)

Anyone any suggestions what could relate to this problem of the poor detecting depth ( in air) ?

I've payed most attention on detection in ALL METAL mode , didn't experiment much (yet) in other modes.

Is this a example that nulling isn't only about residual voltage but perhaps most important the influence of the phase shift on serveral testpoints?

When nulling on scoop 30mV is on MultiMeter 000.0 mV
" " 80mV is on MultiMeter 1mV
The residual voltage mentioned in serveral post ( 4 -12 mV depending on type MM) is pretty high with scoop measured direct on coil ?




regards Dennis the Mennis

[Dennis the Mennis]

[quote=Qiaozhi;102430]It is unlikely that the low sensitivity is caused by using ceramic capacitors throughout the design. Polyester caps would be preferred in the TX and RX circuits, but not essential. Ceramic caps are best for coupling, decoupling and filtering.If you change the TX and RX caps for polyester types, I doubt you will notice any difference.

Hi Qiaozhi ,

I guess you were right ...(?)

regards

Dennis the Mennis

[Qiaozhi]

It is unlikely that the low sensitivity is caused by using ceramic capacitors throughout the design. Polyester caps would be preferred in the TX and RX circuits, but not essential. Ceramic caps are best for coupling, decoupling and filtering.If you change the TX and RX caps for polyester types, I doubt you will notice any difference.

Hi Qiaozhi ,

I guess you were right ...(?)

regards

Dennis the Mennis

I'm not going to say, "told you so".

Looking at the values for your coils and the tuning caps used, I have calculated your TX frequency to be 14.5KHz, which you have confirmed is correct. However, your RX circuit is tuned to 19.7KHz, which in my opinion is too high, as Tesoro circuits are usually off resonance by around 9%. In this case you should be using a calculated capacitor value of 15.5nF to tune the RX to 15.8KHz. I would suggest trying a 15nF cap instead of the 10nF and see if there is any improvement.

[porkluvr]

Darn, Qiaozhi. You beat me to it.

LTspice shows almost 3x increase in received signal voltage by changing to 15n, with a 6.53mH receiver coil.

Sorry about the wild goose chase, Dennis.

[ivconic]

I'm not going to say, "told you so".

Looking at the values for your coils and the tuning caps used, I have calculated your TX frequency to be 14.5KHz, which you have confirmed is correct. However, your RX circuit is tuned to 19.7KHz, which in my opinion is too high, as Tesoro circuits are usually off resonance by around 9%. In this case you should be using a calculated capacitor value of 15.5nF to tune the RX to 15.8KHz. I would suggest trying a 15nF cap instead of the 10nF and see if there is any improvement.

Very nice calculator here!

http://www.whatcircuits.com/lc-reson...cy-calculator/

[ivconic]

" ..Nulling on pin6 input LF353 (coil connected to pcb to compare with firdt video):
http://www.youtube.com/watch?v=AyK02Zk1PLY .."

I guess those "notches" (kind a distortion in sinus) are caused by bad quality of RX cap ?
Because i usually have there perfect sinus without any distortions...

I am using PP caps at TX and RX whenever i can find proper ones.

[ivconic]

Off resonance:

TX = 6mH + 20nF = 14.53kHz
or.......... + 22nF = 13.85kHz
RX = 6.5mH + 15nF = 16.12kHz

1.59kHz or 2.27kHz off resonance....

[ivconic]

" ...Is there a possibility that when even the nulling seems fine , the sensitivity is poor ?..."

"Nulling seems fine" but you must understand that coils nulled well on one frequency are not nulled well on another (different) frequency. Residual voltage rised if frequency changed.

I think Qiaozhi already gave here good hint. Keep RX coil resonance within +1.8 - +2kHz than TX coil.
And use pp caps.

BTW thank You very much on your effort making those clips!
Best Regards!

[ivconic]

Qiaozhi and others;

i was thinking to experiment with serial LC at RX input? To make it more "wide band". What do you think about that?
BTW i always asked myself why at some designes we can see serial and at some parallel tuned coils? What are differences? (generally i think i understand the differences but would like to know details)

[ivconic]

Another thing;
am i wrong or i see TX amplitude there barely over 10vpp ?
That could be the cause of poor depths.
Usually amplitude is 15-16vpp at TGS/TGSL...

[ivconic]

"...Keep RX coil resonance within +1.8 - +2kHz than TX coil...."

I just wanted to check this on datas from original Tesoro coils that are posted somewhere on Geotech forums in the past:

Brown 8" = 5.59mH(22nF) - 6.23mH(15nF) .....i got 2.11kHz off resonance,
White 8" = 5.68mH(22nF) - 6.36mH(15nF) .....i got 1.89kHz off resonance,
White 4" = 5.62mH(22nF) - 6.13mH(15nF) .....i got 2.29kHz off resonance...

So...i was pretty close! From 1.89 to 2.29 !

[Ultimate_Haze]

Hello,
I have almost same troubles with my last TGSL1265.
Still cant find whats wrong.I have over 12 different coils.Various sizes some are 22 /best I can reach is 20sm for 1e coin/some 27.Different L,including 6 and 6,5.But at this specific device no one from those give me 30-31sm on one euro coin.
Strangest thing is that I need 5,50mH for TX so I can read 14,6kHz.With 6mH its around 14kHz.Have check all caps several times.All ok.Also change different types capacitators at vital points.Have play with C9 from 100p to 68pf,40pf.
The coil cable is 4 cores with common shield like in picture.
Have add tresh. pot-no change.So I restore like it was.
So now no idea what else can I try.
I guess that will have to make another device.But because I have spend so much time on this one,curies what I did wrong?
Oh one more thing.On all coils its impossible to null to 0.07mV.All have too high residual voltages.Around 1mV.And no matter what I try the just refuse to tune better.Maybe too much noise inside?Wireless ,satTV ,etc.?

[ivconic]

Hello,
I have almost same troubles with my last TGSL1265.
Still cant find whats wrong.I have over 12 different coils.Various sizes some are 22 /best I can reach is 20sm for 1e coin/some 27.Different L,including 6 and 6,5.But at this specific device no one from those give me 30-31sm on one euro coin.
Strangest thing is that I need 5,50mH for TX so I can read 14,6kHz.With 6mH its around 14kHz.Have check all caps several times.All ok.Also change different types capacitators at vital points.Have play with C9 from 100p to 68pf,40pf.
The coil cable is 4 cores with common shield like in picture.
Have add tresh. pot-no change.So I restore like it was.
So now no idea what else can I try.
I guess that will have to make another device.But because I have spend so much time on this one,curies what I did wrong?
Oh one more thing.On all coils its impossible to null to 0.07mV.All have too high residual voltages.Around 1mV.And no matter what I try the just refuse to tune better.Maybe too much noise inside?Wireless ,satTV ,etc.?

Ehhh! It was in the past when i had some strange multimeter and i got such funny measurements! I think i already corrected those and posted somewhere that residual voltage at good nulled coils is in range 4mV-7mV...athough it can work very good at up to 15mV too. (i had similar coils)
....
About frequency...you do not need to "chase" exact frequency at all. Keep it between 13.5 and 14.5 kHz and don't worry.
.....
What's than wrong there? Some bad component...or some bad spot in signal paths, nothing else.
Read my todays posts about resonance. Check at your coils (with presented caps) that.

[Qiaozhi]

Qiaozhi and others;

i was thinking to experiment with serial LC at RX input? To make it more "wide band". What do you think about that?
BTW i always asked myself why at some designes we can see serial and at some parallel tuned coils? What are differences? (generally i think i understand the differences but would like to know details)

For a series resonant circuit the reactances (L and C) will create a very low impedance, whereas for a parallel circuit the reactances result in a very high impedance.

The TGSL uses an opamp for the RX pre-amp with a high impedance input and is configured as a voltage amplifier. Therefore I would not expect a series resonant circuit to function correctly in this configuration.

Give it a try and let us know the result.