Let's made a PC-base metal detector with usb interface !!!

[Aziz]

Hi all,

Pre-Amplifier results:
Inverting Pre-Amp: NJM2068D: -101.5 dB (3 dB better than NE5532P)
Non-Inverting AC Pre-Amp: NJM2068D: -103 dB (4.5 dB better than NE5532P)
(Rin=47, RG=4.7k, RG2=47, G=1+RG/RG2 = 101)

It got slightly better. The op-amp amplifier noise is dominating now. Yep, it makes sense to do the new amplifier design. It also blocks the DC offset error of the op-amp and incorporates a high-pass filter too. It requires two more capacitors and two more resistors now. But they should fit into the new PCB.


Cheers,
Aziz

[Aziz]

Two Channel Pre-Amplifier

Hi all,

BTW, we can combine both amplifier channels to get additionally 3 dB more signal-to-noise ratio (taking the digital average of both channels).

With the NJM 2068, the noise floor level will achieve -106 dB (only 3 dB distance to the LT1028 ). This is a good performance for a real bargain Japanese dual op-amp, if you consider it's current price: 0,57 EUR (at www.reichelt.de, Well, the price will very likely rise now.)

This is one of the reasons, why I insist on a dual channel pre-amplifier.
More flexibility!!!

Aziz

[lonelyWOLF]

hi aziz
good choice NJM2068 YAMAHA EMX5000 powered mxer (2x500w)used it in the mic. input channell

[Aziz]

!!! LM4562 !!!
http://www.national.com/pf/LM/LM4562.html#Overview

2.7nV/ √ ^Hz

Pity, I have to look where to get it. If you can get it, use it!
Aziz

[Aziz]

Another bargain Japan-IC found:
NJM 2122
http://semicon.njr.co.jp/eng/product...e/NJM2122.html

(max. +/- 7V!!! for 12 V systems only).
1.5nV/√Hz typ.@ f=1kHz

Aziz

PS: Have a look here:
http://nwavguy.blogspot.de/2011/08/o...surements.html
Interesting!

[simonbaker]

Another bargain Japan-IC found:
NJM 2122
http://semicon.njr.co.jp/eng/product...e/NJM2122.html

(max. +/- 7V!!! for 12 V systems only).
1.5nV/√Hz typ.@ f=1kHz

Aziz

PS: Have a look here:
http://nwavguy.blogspot.de/2011/08/o...surements.html
Interesting!

Our brain waves will register as EMI!

[Mechanic]

Our brain waves will register as EMI!

Hi Simonbaker,

Thats why you have to clear you mind for detecting Ahhhh but not any more!

Been fiddling again, firstly figured out why I was not getting good results last week, solder bridge in the middle layer of my front end board effectively shorting out my receive coil! Found this out the hard way when I went to hook up a mono coil out bush and she didn't want to work! The tx fets just about de-soldered themselves

Anyway, back in business I have an 18" "mono" coil running with a 18" rx coil. The null is adjusted by adjusting the gain on the preamps then the 2 signals added into one. I have set the -rx coil about 1" above the tx(rx+) and I am happy to say that it seems to detect targets quite well. The preamp gains can be tweaked a little to give one coil more bias otherwise a target can just about be cancelled out. Anyway I can now have my detector running inside the house and it is relatively smooth, especially compared to just a normal 18"mono. I am running with a maximum gain of 91 on the preamp and the next stage has a gain of 3. I can now turn the coil on its side and not get a barrage of warbling. Time to wind up a similar setup for my 25" I think! I know that this is not IB, this is just plain old noise cancelled mono, but at this point in time this is what I need.

Check out the piccys of the output of the last stage of filtering just before the audio section. Both taken with full front end gain, first just a standard mono @1V/div 0.2 seconds/div and the second is the emi cancelled mono, @ 0.2V/div 0.2 seconds/div
No more pesky electric fences for me

Cheers Mick

[Aziz]

..
Anyway, back in business I have an 18" "mono" coil running with a 18" rx coil. The null is adjusted by adjusting the gain on the preamps then the 2 signals added into one. I have set the -rx coil about 1" above the tx(rx+) and I am happy to say that it seems to detect targets quite well.
..

Hi Mick,

1 inch vertical distance of TX(RX+)/RX- isn't much. Take coil diameter distance to get more significant target response.

Aziz

[Mechanic]

Hi Aziz,

I realize that more distance would be better, but, with a deep target a lot of the time you lift the coil 1" and it is just about gone. If more receive bias is put on the closest coil to the target this difference will be a bit greater again whilst still keeping emi levels way down compared to the conventional method. But yes I will try adding more distance to see the full effect. Test pit here I come.....

Also emi levels out bush are never as high(no where near it) as what they are here at home. A really bad day may be comparable to the noise canceled waveform at my house, so I can afford to not have the 2 coils %100 cancelled.

Cheer Mick

[simonbaker]

Hi Aziz,

I realize that more distance would be better, but, with a deep target a lot of the time you lift the coil 1" and it is just about gone. If more receive bias is put on the closest coil to the target this difference will be a bit greater again whilst still keeping emi levels way down compared to the conventional method. But yes I will try adding more distance to see the full effect. Test pit here I come.....

Also emi levels out bush are never as high(no where near it) as what they are here at home. A really bad day may be comparable to the noise canceled waveform at my house, so I can afford to not have the 2 coils %100 cancelled.

Cheer Mick

I think you need to really separate the coils for this concept to work at all. The closer the rx coils to each other, the weaker target reception will be (assuming you use 100% EMI cancellation).

The RX coil that is coplaner with the TX coil is the real receive coil -- the other coil is just for EMI cancellation and should be far away so it doesn't pick up much target at all.

At least that's how I think it works, if it is the same idea I have been playing with for some time.

Regards,

-SB

[Aziz]

1 dB more SNR!

Hi all,

If I lower the input resistance to 10 Ohm, RG=1k, RG2=10 (G=RG/RG2=101), I get 1 dB more SNR with NJM 2068. So the input shorted AC amplifier with G=101 will arrive the -104 dB noise floor level. With two combined amplifiers, -107 dB is achiveable. Not bad for a cent priced (0,57 EUR) Japanese dual op-amp. (Bye bye LT1028, AD797, ..)


The NJM 2068 offers enough unity gain bandwidth product, so we can go for high gain easily. It also offers enough drive capability (low load resistance), so we can reduce the RG resistor down to 1 kOhm (no problem with non-linearities).

The only issue is the high-pass cutoff frequency and the required high capacitance DC blocking capacitor in the AC amplifier design (requires 10 µF .. 100µF). But we can use special bipolar electrolytic capacitors or even normal AL electrolytic capacitors (they withstand 1.5 V reverse polarity voltage and we have the clamping diodes on the input to +/-0.7V).

Let's look, how the simple AC amplifier performs in a rigid environment.

Cheers,
Aziz

[Aziz]

Howto..

Hi all,

during gluing the AL-strip TX coils, I have found, that the hot glue is dissolved by acetone.
That gives the possibility of cold glue painting via brushing. The fluid plastic cold glue.
*LOL*

Aziz

[wallker]

Hi all,

during gluing the AL-strip TX coils, I have found, that the hot glue is dissolved by acetone.
That gives the possibility of cold glue painting via brushing. The fluid plastic cold glue.
*LOL*

Aziz

Hi Aziz, for gluing AL-strips on coils i make my own glue, it's simply mixture of nitro thinner and styropor.
Try it, it aint bad
Regards Marko.

[Aziz]

Hi Aziz, for gluing AL-strips on coils i make my own glue, it's simply mixture of nitro thinner and styropor.
Try it, it aint bad
Regards Marko.

Hi Marko,

yep, that's for sure a good styrofoam solvent.
It's only a question of drying time.

Best solvents for styrofoam:
- balsam turpentine
- nitro thinner

Note, some substitues for turpentine are bad solvents. Even acetone doesn't solve styrofoam good (for painting useless, until you wait for a long time for dissolving duration). Unfortunately, Europe has banned the turpentine from the market.

Cheers,
Aziz

[Aziz]

Hi all,

ok, the 10 Ohm input resistor to the high voltage clamping 1N4148 diodes seem to work without damaging the receiver amplifier op-amp (yet). I have put the RX coil to the maximum induction coupling and there is no problem at all. Note, that the clamping diodes must withstand the external EMI (other detectors, etc.). In induction balanced coil configuration, the own generated high TX voltage won't be seen.

To be save for operation, just increase the input resistor to a higher value (e.g. 47 - 100 Ohm, which causes more noise of course).

What's next?
Input compensation, new circuit and PCB with incorporated non-inverting AC pre-amplifier.

Cheers,
Aziz