TESORO GOLDEN SABRE

[Qiaozhi]

"...Disconnect pin 1 of U106a, which will effectively disable the DISC channel,
and let us know how this affects the ground balance..."
I did that. When switched ON it producing "beep,beep,beep,beep..." non-stop and
now way to calm it down, whatever i did.
"..I do not have a 20-turn pot on my detector, and ground balance is quite easy to achieve.."
So i replaced ordinary with multui-turn pot at GEB. Fine adjusted with it to produce
0.000vV. Zero - crossing you've explained above...It was easy (as you said) to adjust to
0 volts. Of course without any metal arround coil. When metal is presented near the coil
this voltage is not 0V any more but respecting signal strength it raises...
I made some measurements at analogue switches.
Rotating GEB trimm. gave me range from -7.66mV to +180mV at switch output (TR5).
Rotating Disc. pot gave me range from -130mV to -190mV at Fet output (TR4)...BUT!
Rotating Disc from one side to other, as approaching to middle point, voltage reversed
and somewhere at the midddle of pot it reads +330mV, further rotating lowered it to -190mV..
I checked everything again and again...still no solution!
Only way this TGSL to work very good without beep's is to Set Disc. at "3" and higher and
Sens to "6-7" in Disc mode. GEB trimmer is adjusted as described above. Adjusted like
this, TGSL is very deep,good discrimination...on the field it can produce very rare...but
very,very rare "beep's"...usually when coil hit some stone or bulky grass, harder...
In All metal mode i have to lower Sens at minimum to make it at least a bit useable..?
I do beleive that source of problem is somewhere after front end. Why?
OK, when switched ON it is silent. Than i provoke it to "sing" with some coin or simillar.
It wont stop as i move away that coin. So, i noticed if touch 3 lead cable (5cm long) from
pcb to Disc. pot...it stops with "beep's" at once??? When cable released it still remains
calm - no sound. Again when coin is moved over coil, it "beep's" without stoping?
If touch randomly other parts on pcb, sometimes i can stop it. But most significant is
that cable and Disc. pot. It stop with "beeping" always, when those are touched with finger?
Also rotating Disc pot has much more influence on detector behavior than anything else.
Rotating Sense pot doesnt affect anything noticeable. Also rotating GEB trimmer wont stop
once provoked "beeping" at all...??? Damn!
I have feeling that solution to this problem is right infront of my eyes...only i cant see it!

Thanks for checking that.
I plan to make a small modification to my own design for the all-metal mode, but this may not be until later today or tomorrow. Will let you know the result.

All that continuous beeping after making it "sing" with a coin ... Hmmm!
I would put money on it being caused by instability in the comparators. The only way you can get this continuous audio signal is if the wire-ored outputs of the two comparators are going high together. I did notice that there is no pull-up resistor at the output of the comparators, and these devices have an open collector output. Try putting a 1M resistor between the wire-ored outputs and the power supply rail. i.e. at pin1 U106a.

[ivconic]

You missed it! Already there, see R54..

BTW i found a huge source of noise here! Damn! It was damn Disc. pot 100k...I replaced it with brand new one and all blleps stopped in Disc mode!
But it was new also!? Damn!
Now i have some bleeps only in All Metal mode...
Damn!

[Max]

"...Disconnect pin 1 of U106a, which will effectively disable the DISC channel,
and let us know how this affects the ground balance..."
I did that. When switched ON it producing "beep,beep,beep,beep..." non-stop and
now way to calm it down, whatever i did.
"..I do not have a 20-turn pot on my detector, and ground balance is quite easy to achieve.."
So i replaced ordinary with multui-turn pot at GEB. Fine adjusted with it to produce
0.000vV. Zero - crossing you've explained above...It was easy (as you said) to adjust to
0 volts. Of course without any metal arround coil. When metal is presented near the coil
this voltage is not 0V any more but respecting signal strength it raises...
I made some measurements at analogue switches.
Rotating GEB trimm. gave me range from -7.66mV to +180mV at switch output (TR5).
Rotating Disc. pot gave me range from -130mV to -190mV at Fet output (TR4)...BUT!
Rotating Disc from one side to other, as approaching to middle point, voltage reversed
and somewhere at the midddle of pot it reads +330mV, further rotating lowered it to -190mV..
I checked everything again and again...still no solution!
Only way this TGSL to work very good without beep's is to Set Disc. at "3" and higher and
Sens to "6-7" in Disc mode. GEB trimmer is adjusted as described above. Adjusted like
this, TGSL is very deep,good discrimination...on the field it can produce very rare...but
very,very rare "beep's"...usually when coil hit some stone or bulky grass, harder...
In All metal mode i have to lower Sens at minimum to make it at least a bit useable..?
I do beleive that source of problem is somewhere after front end. Why?
OK, when switched ON it is silent. Than i provoke it to "sing" with some coin or simillar.
It wont stop as i move away that coin. So, i noticed if touch 3 lead cable (5cm long) from
pcb to Disc. pot...it stops with "beep's" at once??? When cable released it still remains
calm - no sound. Again when coin is moved over coil, it "beep's" without stoping?
If touch randomly other parts on pcb, sometimes i can stop it. But most significant is
that cable and Disc. pot. It stop with "beeping" always, when those are touched with finger?
Also rotating Disc pot has much more influence on detector behavior than anything else.
Rotating Sense pot doesnt affect anything noticeable. Also rotating GEB trimmer wont stop
once provoked "beeping" at all...??? Damn!
I have feeling that solution to this problem is right infront of my eyes...only i cant see it!

Hi,
when you put finger on cable... you put there static charges... and that maybe out tune the self oscillating part.

Today I've made another 255x137 using multimeter and scope to compare:

- used multimeter in ac at 200mV range (good quality one)
- used a scope that has an input of 1Mohm and 25pF capacity with a 100MHz rated probe (all made in japan stuff) at x1:1 attenuation

Frequency of TX with new TX coil (100 turns 30AWG) = 14.59KHz without shield

14.57 KHz with shield.
Ok here the results of four tests:

1. RX with nothing else MULTIMETER READ 0.0 mV ; SCOPE read a bad signal sinusoidal with intermodulation of fundamental resonance freq. for coil itself + other sources... results is a sinusoidal multi-modulated signal, that's cause multimeter read exactly zero at a point... and at same point instead scope read maximum amplitude of 40mVpp and min of 20mVpp

2. RX with a 15nF poly cap in parallel :
- multimeter read 11mV at min
- at same point scope read a perfect sinus with amplitude of 70mVpp very different from reading of multimeter... maybe due to frequency jitters

3. RX+cap15nF parallel+1Kohm in series:
- multimeter read 10.5mV at min
- same point scope read 63mVpp - perfect sinus

4. RX+cap15nF+1Kohm all parallel:
- multimeter read exactly 0.0mV at zero point
- scope read 5.5mVpp at same point

5. same of 4. but to read 0.1mV on multimeter
- mult. 0.1mV stable
- same point scope read instead 20mVpp

Comparing 4 and 5 you see that's the best way to null the coil... you get 0.0 on multimeter when also on scope voltage is at minimum for real. But then you have to do it with much care... cause 0.1mV reading on multimeter means actually about 15mVpp on the scope.

This last results make me thinking about Quiaozhi assertion of that Tesoro null coils at 15mVpp... thus I think they maybe use a multimeter like mine !

Who knows ?

Conclusion: just the method used in 4 and 5 is right for actually nulling with digital meter at 200mV range: RX+capacitor parallel+1Kohm resistor parallel.

Variation could be adding a series resistor to compensate some probe effects... but I feel that with multimeter is absolutely not required.

Best regards,
Max

[Max]

You missed it! Already there, see R54..

BTW i found a huge source of noise here! Damn! It was damn Disc. pot 100k...I replaced it with brand new one and all blleps stopped in Disc mode!
But it was new also!? Damn!
Now i have some bleeps only in All Metal mode...
Damn!

Hi,
yes sometimes graphite layer is so thin that produce noise also in new parts.

Best you can find is wire wound (like multiturn) pots... but too expensive for more than one detector having that.... then they are also heavy 3 or 4 times a small pastic one or aluminum boxed type.

Instead try an LM393AN at last comparator to gain better offset level... could reduce noise.

Kind regards,
Max

[ivconic]

Hi,

when you put finger on cable... you put there static charges... and that maybe out tune the self oscillating part.

Today I've made another 255x137 using multimeter and scope to compare:

- used multimeter in ac at 200mV range (good quality one)
- used a scope that has an input of 1Mohm and 25pF capacity with a 100MHz rated probe (all made in japan stuff) at x1:1 attenuation

Frequency of TX with new TX coil (100 turns 30AWG) = 14.59KHz without shield

14.57 KHz with shield.
Ok here the results of four tests:

1. RX with nothing else MULTIMETER READ 0.0 mV ; SCOPE read a bad signal sinusoidal with intermodulation of fundamental resonance freq. for coil itself + other sources... results is a sinusoidal multi-modulated signal, that's cause multimeter read exactly zero at a point... and at same point instead scope read maximum amplitude of 40mVpp and min of 20mVpp

2. RX with a 15nF poly cap in parallel :
- multimeter read 11mV at min
- at same point scope read a perfect sinus with amplitude of 70mVpp very different from reading of multimeter... maybe due to frequency jitters

3. RX+cap15nF parallel+1Kohm in series:
- multimeter read 10.5mV at min
- same point scope read 63mVpp - perfect sinus

4. RX+cap15nF+1Kohm all parallel:
- multimeter read exactly 0.0mV at zero point
- scope read 5.5mVpp at same point

5. same of 4. but to read 0.1mV on multimeter
- mult. 0.1mV stable
- same point scope read instead 20mVpp

Comparing 4 and 5 you see that's the best way to null the coil... you get 0.0 on multimeter when also on scope voltage is at minimum for real. But then you have to do it with much care... cause 0.1mV reading on multimeter means actually about 15mVpp on the scope.

This last results make me thinking about Quiaozhi assertion of that Tesoro null coils at 15mVpp... thus I think they maybe use a multimeter like mine !

Who knows ?

Conclusion: just the method used in 4 and 5 is right for actually nulling with digital meter at 200mV range: RX+capacitor parallel+1Kohm resistor parallel.

Variation could be adding a series resistor to compensate some probe effects... but I feel that with multimeter is absolutely not required.

Best regards,
Max

Very good job Max! Thank you for posting these detailed here.
Yes i agree with you. But it seems that we do not understand each other?
This is due my lacks in English. I can talk ordinary, everyday english, but here we are using more technical language and i have so many lacks...dont have enough words fund to explain what i really mean to say...Shame on me!
I absolutelly agree scope is unreplaceable in measurements! But in case we dont have it (my case) we are stuck to tools we have, in this case it is multimeter with 200mV range...
So in previous posts i elaborated techniques how to do most accurate measurements with only multimeter..
Let's accept that coil is not nulled and producing some 30mV at RX...OK. Now, if we have multimeter with lacks it will read ...let's say some 25mV...So we gonna have 5mV error, right?
If we put 15nF capacitor parallel to RX coil, voltage will raise up to 27-28mV.
If we add parallel resistor, voltage will fall to 10-15mV...Huge error! That's what i ment! So now, if we null coil with that setup and achieve null, actually we not gonna null it to null but to 30-10=20mV or 30-15=15mV...At the end we wil have "nulled" coil with 15mV residual voltage..which is in a way to much...If we omitted resistor, and than null coil, we'll get 30-27=3mV, meaning "nulled" coil with 3mV residual voltage, which should be acceptable more than previous case...See what i mean?
Quiaozhi once stated that higher residual voltage is not a big deal, only can make minor problems at discrimination....I say no it is BIG problem here at Tesoro Golden Sabre. Not sure about other detectors but here at TGS, disc and geb are practically two parts of same stage and do have huge influence on each other. Actually, here at TGS, geb is a sort of extended discrimination applied to ground signals...Isnt that true?
That's why i insist on absolute nulling here, specially to TGS design...again i am not sure about other designs...
An it showed to me in praxis that i am very right. Actually i established my claims exact from praxis...I made various coils with different residual voltages. Best working coil is with lowest res.voltage..disc and geb works excellent on those...
That's what i ment from very begining....I hope now explained a bit more understandable..?
Besides, all other things you and Quiaozhi stated here are right and i can only agree fully!
Regards!

P.S.
Yes, it was damn Disc. pot from the begining. I have 8 finished pcb's here at my workshop. But choose only one as testing pcb, for nulling and testing coils. By damn Marphy's law that one pcb was only one with bad potentiometer...! Damn! that's why i had so many problems last few days....
It was infront of my eyes and i couldnt see it!!! Damn!

[Qiaozhi]

You missed it! Already there, see R54..

BTW i found a huge source of noise here! Damn! It was damn Disc. pot 100k...I replaced it with brand new one and all blleps stopped in Disc mode!
But it was new also!? Damn!
Now i have some bleeps only in All Metal mode...
Damn!

If you look at the schematic posted earlier, where I asked you to disconnect pin 1 of U106A, there is no 1M resistor. BUT - if I look again at the original schematic, then there is a 1M resistor, but it is connected by a dotted line with some words that say "To check this!". Must be an old schematic.

Glad you have fixed the problem.

This last results make me thinking about Quiaozhi assertion of that Tesoro null coils at 15mVpp... thus I think they maybe use a multimeter like mine !

Looks like you really had 15mV after all ...

Quiaozhi once stated that higher residual voltage is not a big deal, only can make minor problems at discrimination....I say no it is BIG problem here at Tesoro Golden Sabre. Not sure about other detectors but here at TGS, disc and geb are practically two parts of same stage and do have huge influence on each other. Actually, here at TGS, geb is a sort of extended discrimination applied to ground signals...Isnt that true?

I plan to do some experiments soon to look more closely at this question. At this time, however, I still think that nulling the residual voltage to the nth degree is not necessary.

[ivconic]

Looking forward to see your experiences...
To tell you the truth i learned more from your and Max's post in last couple weeks than from many other publications in last year...First few years i spent building detectors pretty much automatically - without much of understanding. So i missed up much elementar knowledge which must have to understand md principles of working. Now i am trying to compensate that huge lack by asking you to talk here...
But, most easiest way to learn something is when you have possibillity to experiment in "vivo" leaded by someone who already know things...
Max already made TGS...it is so easy to talk with him and exchange experiences...also to obey hints and apply mods...It will be good if you were made TGS also so we can talk here more practically...anyway i had feeling that you can explain things here more. It's up to you and your good will. But let's go step by step...Do what you planed to do and later post here your experiences...


I have some "misty" things unsolved yet, some question unanswered;
i'll put those as simple as i can;

Gaining more depth with TGS is not subject to me, right now i am very satisfied with it's depth, i do want to solve stabillity issue here...Yes it is very stabile but not as i wanted...
I want to make it stabile as a rock.
Discrimination is not issue also, TGS has excellent discriminaton,
I am wandering is there simple way (without changing pcb) to mod it's oscillator and stabilize it better,
also i want to mod it's audio..audio is useable, but what i have on my mind is Fisher CZ5 look-like audio or simillar...
and for the end if it is possible to mod oscillator to radiate more power in TX coil, but to control it and avoid "Anker apparition" - other words not to radiate to much power and cause "choking" of RX part, further having huge instabillity on the ground....
All of those i am planing to achieve this winter so when spring comes i will have complete detector design.
I made so many different detectors so far, kind a tired of mixing things. So trully decided to stop and take TGS as "platform" and improve it more and more.

regards!

[Geo]

Maybe you want a knew detector
The 1st problem is the pcb. So you can use bigger voltage only for the oscillator and with a LM317 to regulate the voltage and this way and the power. If the voltage is in a normal place then you have power change without phase shift. For more stability you must use bigger transistor so the heating to be less and to use 2 or 3 capacitors parallel instead ofl 1 capacitor so the current inside the capacitor to be less............ less heating ......... more stability.
More tomorrow.
Oh i forgot......... today i build a TGS, a little parts yet and it is ready. Maybe after 2 days
Regards

[ivconic]

Maybe you want a knew detector
The 1st problem is the pcb. So you can use bigger voltage only for the oscillator and with a LM317 to regulate the voltage and this way and the power. If the voltage is in a normal place then you have power change without phase shift. For more stability you must use bigger transistor so the heating to be less and to use 2 or 3 capacitors parallel instead ofl 1 capacitor so the current inside the capacitor to be less............ less heating ......... more stability.
More tomorrow.
Oh i forgot......... today i build a TGS, a little parts yet and it is ready. Maybe after 2 days
Regards

I'll leave power mod as last to do here.
First to do first! Stabillity issue...
But you mentioned replacing one capacitor with few parallel... So i am thinking; we have 220nF at TGS oscillator. We can easuly replace it with 3 parallel of 68nF and forth of 22nF. It will make 226nF..Cool! But what benefits we do have from those, except less current as you said?
Let's explore this more; when oscillator runs better? With one compact or few parallel? This must be scoped to be sure. Second; which type are best suited to put in oscillator? I am using P.P. capacitors mostly..
Also which transistor? I am using 2N2907 although BC557 tested and havent noticed significant differences...
This subject already explored in the past, but no harm if we elaborate it further..
Imoprtant question; how to replace stabilisation with fet in oscillator. Using termistors is bad idea, more obsolete than fet setup. Also i havent noticed jitters in TGS oscillator but that doesnt mean it is perfect. To be honest i havent payed to much attention on its stabillity at all. First time i made TGS just checked it for 2 hours on eventual jitters, and when i saw there are no significant jitters, i left it as it is.
One thing is not clear to me is zener diode. 4V3 is suggested, i put 4V3 and later 4V7 - havent saw any differences too?! So what's point with zener here after all? I think to put 3V9 or 5V1...or even 6V8 to see if any changes?
Should i?
And there are so many other question to be answered....but i am not in a hurry...winter is coming, we have plenty of time...
Regards!

[ivconic]

Quiaozhi:

"...Yes - I know it says "... non motion type of detector with fast auto tune ..." but it still requires the coil to be in motion. You cannot hover over the target, otherwise the signal disappears. The auto tune is "fast", as it says on the..."

Quiaozhi! This sounds like audio i was thinking on, all this time!? Is there any chance to mod TGS to work like you described here?
Of course i do not suggest you to offer here some secrets from that detector you talked about! Interested only in principles and eventual mods...

[Qiaozhi]

Quiaozhi:

"...Yes - I know it says "... non motion type of detector with fast auto tune ..." but it still requires the coil to be in motion. You cannot hover over the target, otherwise the signal disappears. The auto tune is "fast", as it says on the..."

Quiaozhi! This sounds like audio i was thinking on, all this time!? Is there any chance to mod TGS to work like you described here?
Of course i do not suggest you to offer here some secrets from that detector you talked about! Interested only in principles and eventual mods...

This probably not what you're thinking. It has nothing to do with the audio circuit.
All I can say is that it's not the same type of differentiating circuitry as the TGSL. Have you looked at the Silver Sabre? It uses an auto tune circuit followed by a differentiator.

[Qiaozhi]

I have a little circuit for a speech processor somewhere that I can scan in and post tomorrow. This might be what you're looking for.

[ivconic]

I have a little circuit for a speech processor somewhere that I can scan in and post tomorrow. This might be what you're looking for.

Ehm! No...i am looking for audio simillar to CZ5; a bit longer lasting,changing pitch,decayed and as you explained earlier in that post i quoted above..
Should not been a huge problem...?
Regards

[Geo]

I attach here the pcb with errors that Max sent here. I have checkthe 10pF... the 1M and one diode that there is at the schematic but no at the pcb. Tell me if are correct.
I have 2 LM308 with metallic case. Where is the ledge??? Ivconic said that it is on pin1......... i looked at datasheet and shows it at pin8. Who is the right.... the shop had only 2 pieces so i dont like to burn it.
Regards

[ivconic]

No Geo! Beware! I sad that made MISTAKE considering pin 1. under notch...
It is pin 8. under notch not pin1. !
Yes there are 4k7 resistors in chopper, not 47k as showed in schematic.
There is also 2k2 instead 33k....
about 470k in battery test....i put 470k,it is working...
1M resistor goes parallel to diode...only if you want to boost audio. For now do not put it.Leave as it is...