Can pinpointer detecting thin chain?

[Nupi]

Hello nupi,

I have changed R3 to 1K and set the proper voltages with R2 100K pot with L2 (lead nº1 from coil out).... then resolded it
Adjusting sensetivity with R1 still motor
Also no movement from motor, with or without metal over it

Also tried to replace C2 with: only one 10n cap / 2 paralel 10n cap and also 3 10n capacitor ... but no joy

I only get 1 second of movement from motor, connecting or disconnecting L2 (lead nº1) from the powered ON circuit

Tomorrow will try diferent settings

Best regards

I want to solve through the posts because other have ProveIt of
but if that does not succeed, because writing is very dificult for me,
I will call you. Is that an option? Phone number you can give if you want me. Using pp.

[Nupi]

That the motor get 1 second of movement by connecting or disconnecting L2 (lead nº1)
from the powered ON circuit is a good signal.
[He's waking up.(joke)]


Okey step by step now.
Forget detecting with a object and adjusting with R1 that is not necessary now, that will come later.


If all goes well you should get, when adjusting R2, (without metal by the coil) a detection point.
You have to rotate R2 very slowly.
A detection point, which increases in frequency, clockwise and counterclockwise. This is important.


---Connect the coil.
---Place the 3x 10nF Capacitors.
---place R3 a 2,2k
---Remove the motor and place a 1k resistor in place.
---Measuring the DC voltage at the emitter of the transistor.
If the pinpointer not detected should be close to the power supply voltage.
---Now you have to R2 first set at the left or right.That does not matter.
---Now you need to run very slowly to R2.
---THE voltage must come running to a point that the voltage drops to about 5 volts.
This means that the pointer is working and that you can not use that motor here.

It would be easier that you replace the motor with a resistance and a LED to try.


Regards

[Davor]

Hello Davor, b.t.w, how do you progress with the Colpitts oscillator and how about the push button stabilisation?

I'm not in my electronics mode at this point. Soon.

[Nupi]

I'm not in my electronics mode at this point. Soon.

Hello Davor,
Apology, you're right in order to increase the base resistance.
I had already tried it yesterday noon but to no avail.
This morning I was again working on and saw that I had (to try it out) not raised the base resistance with a variable resistor
from 470k but but decreases with a 47k.
That's what happens when you blindly something grabs a compartment and not checked.
I've now tried with a 470k and work. Sorry

[Nupi]

Hello , Nezocas.
You can simply run the tests but replace R2, preferably with a variable resistor
of about 470k if you have one, otherwise leave R2 100k and and replace R1 with 100k increase with
each time with 100k to about 400k in total, and in the meantime check.

[Nezocas]

Hi,

I have replaced R2 with 470K pot
Also replaced R3 with original value 2K2
3x 10nF cap are still in the board
Now, if I turn very slowlly R2, I get to a point where the motor starts moving 1 second, and stops another second, and them moves again (1 sec sotps / 1 sec stops /1 sec stops)
At this point if I move metal over coil, no change at all

I have tried several positions of R2 and R1, but no detection at all

OFFtopic
Nupi, I have no problem to give you my phone, but there will be the language barrier, most of my english is with google translator
I have no problem also to send the pointer to you (if you have the time of course) and pay shipping it back, and your labor

Regards

[Nupi]

Hi,

I have replaced R2 with 470K pot
Also replaced R3 with original value 2K2
3x 10nF cap are still in the board
Now, if I turn very slowlly R2, I get to a point where the motor starts moving 1 second, and stops another second, and them moves again (1 sec sotps / 1 sec stops /1 sec stops)
At this point if I move metal over coil, no change at all

I have tried several positions of R2 and R1, but no detection at all

OFFtopic
Nupi, I have no problem to give you my phone, but there will be the language barrier, most of my english is with google translator
I have no problem also to send the pointer to you (if you have the time of course) and pay shipping it back, and your labor

Regards

Hello Nezocas,
It's fine. I will ensure that the pointer does.
I'll give you my address in a PM later.

Regards

[Davor]

I tried a few approaches to automatic stabilisation and ... they actually work. Simple RC filtering seem to be a better solution than integration, but in that case I have a bit less control over it's behaviour, the fast and slow thing, and it must be always on. Maybe a nested solution will work the best?
I'll play with it some other time. These simulations take time - which I don't have.

[Nupi]

I tried a few approaches to automatic stabilisation and ... they actually work. Simple RC filtering seem to be a better solution than integration, but in that case I have a bit less control over it's behaviour, the fast and slow thing, and it must be always on. Maybe a nested solution will work the best?
I'll play with it some other time. These simulations take time - which I don't have.

I have tried several times in different ways but with no real success.
I mean I do have fairly stable, but the sensitivity is 30% less and at the starting point of detection,
he is very restless.The best result (which is satisfactory for me), I have with the circuit and the coil in epoxy.


I have no understanding of spice so I do not attempt it.
I try my idea first on a breadboard and refine it further.


Few days ago I have nupi pointer and Garrett pro pointer taken to the field to test it.
My M.D. is very sensitive even to very small particles I had a signal with the MD but could not find it
with Garrette. With nupi pointer it was no problem (it was a very small piece of aluminum foil).
I took a can of soda from 25cl that I had found and buried. At 21 cm depth was a clear signal.


Thanks for your effort.

[Funfinder]

http://www.geotech1.com/forums/showt...proved-version

[Funfinder]

3 small chain testmovies!

http://www.geotech1.com/forums/showt...050#post190050

I hope you can beat that sensitivity with your pinpointers...

Good Luck!

[Nupi]

http://www.geotech1.com/forums/showt...050#post190050

I hope you can beat that sensitivity with your pinpointers...

Good Luck!

Piece of cake. Nupi pinpointer exceeds yours already.
see post #27 - #251

[Davor]

I think I cracked it, but it will not be this simple. It will require a dual FET-input op amp and a CMOS switch.
The idea is as follows...
There are 2 regimes:
- self-stabilisation
- active

The self-stabilisation "catches" the tune level that corresponds to a very nervous behaviour in the air, and to avoid annoyance this can be muted (or not). Either by expiration of time or by manual manipulation the self stabilisation switches off, the integrator becomes SAH and the tune level is translated by a predefined step voltage to pull the device from nervous behaviour, but only just. That is the active regime.
This would make it work as a pro pinpointer, but with Nupi sensitivity.
Simulations show that a good average of the tuning voltage can be achieved in about 1 second, make it 2 to be sure.

[Nupi]

I think I cracked it, but it will not be this simple. It will require a dual FET-input op amp and a CMOS switch.
The idea is as follows...
There are 2 regimes:
- self-stabilisation
- active

The self-stabilisation "catches" the tune level that corresponds to a very nervous behaviour in the air, and to avoid annoyance this can be muted (or not). Either by expiration of time or by manual manipulation the self stabilisation switches off, the integrator becomes SAH and the tune level is translated by a predefined step voltage to pull the device from nervous behaviour, but only just. That is the active regime.
This would make it work as a pro pinpointer, but with Nupi sensitivity.
Simulations show that a good average of the tuning voltage can be achieved in about 1 second, make it 2 to be sure.

It sounds hopeful, though I must admit that I did not get everything.
I think the language is a small problem now.
When I was working on with a LM 062, I do not know the exact values more
(I've searched but can not find them anymore) but I had good results to a capacitor
in parallel to R3 approximately 390pF (st.b. filter) and 15nF parallel to L1.
Is it a lot of work to see what happened in Spice?
In nupi pinpointer is a minor feedback loop from the supply is present.

[Davor]

I can tell you off-the-bat that these changes will not affect functioning of this device, but there is no obstacle for you to install LTspice and load the project I attached here. Even if it does not work well, you'll be able to reach the solution - the device I sent works in spice very well. Don't be confused by the 2 inverters at the 4024 input - these are simulated with MOSFETs instead of digital primitives to better mimic a real circuit. Otherwise two inverters in series only generate some negligible delay. Real 4024 has a same topology in it's input, so there you have it.

Your biggest achievement with this pinpointer is the 4024 detector. It outperforms all the envelope detectors by far and wide, and it provides a straightforward beeping tone. That said, every improvement should start from the 4024 circuit.

What 4024 circuit does is produce a series of beeps when oscillation envelope approaches Vcc/2, and it does so in a narrow span. This makes it so sensitive, but also touchy to tuning.

Now, my approach was as follows... I integrate voltage on pin 5 which is at 5V between beeps, and oscillates during beeps. It makes average voltage during beeps ~2.5V. Integrator reference is set at about 4V so that it promotes integration during beeps. Initially I tried applying this correction voltage to the transistor base and establish some kind of permanent and motion compensated correction, but the very nature of integrator acted so that I lost a great deal of periodicity as related to target proximity. It would beep in the air, and it would beep almost the same with a target, so in effect it is annoying and not sensitive. The only way of regaining sensitivity would be to stop any kind of regulation while detecting. Then it dawned on me - I can freeze this level SAH style and translate level to just below the constant beeping, and that's it.

I'm still considering the best way of making this process as automatic as possible.