How can i make a metal detector

		Quote:			Posted by xavier
	Hi Goldigger There is a post by Cornelius he has posted a BFO with discriminator I think that it should make a good project did you see it?  ::)The thing with BFO is the stability they just seem so unstable I have build two of them and had no fun with them at all. Ground penetration was about 10cm on large coins you could get 15cm now with this VLF on the bench test that I did I could detect a 3gr gold ring at 22cm a large coin(5 Swiss Frank)at 28cm now the funny thing is that the coin is far larger than the ring and yet the distance dose not change very much however the window frame?s are of aluminium and I detect them at 110cm and my solder sucker (aluminium) 40cm I wanted to shield the coils but by doing so it did not work at more than 5cm for the large coin. The coil is encapsulated so I will make a new one the thing is that I use old copper wire that I got out of a TV set from the deflector coils I think that?s what they are called. As far as I am concerned the electronics are OK it? just a transmitter and a receiver the rest is up to the coils (that?s the hard part) how ever I may be totally wrong about that. The way that I see it you could use any transmitter with any frequency from 1Khz to 30Khz as long as your receiver is tuned to the same frequency. I would like to try experimenting with LF from 30Khz and see what I get but I think that the problem is like you mentioned with the capacitance and induction I don?t have all the necessary measuring equipment and to top it all the coils are jumbled wound so two coils are never the same and I should use an LC meter to see how many uH?s I got on the coil and go from there any how that?s not the end of the word I may have to make a few coils to get it right that?s all. I have been thinking of making flat coils it should be much better and much easier to remove a turn or two but the wire is very thin 0.3mm that?s the problem .With thick wire making flat coils is much easier as the coils have less tendency of over lapping one another so now you are going to say use thicker wire yes but the coils will be bigger and heavier. As far as parts are concerned we can get mostly everything in SA but I live in a remote area with the closest big town 80km from me and often they have to bring the parts in from Johannesburg and that 1200km from me get my problem? By the way here is Q for you if I have a frequency of 8Khz I will have a wave length of 37.478m now 8.000Khz has a wave length 37.48m if I use a wire 37.48m long and make a coil out of it what will the result be if I use it on my 8Khz receiver????See that?s how I did the coil for the metal detector and it could be that I did it all wrong it did make sense to me at the time but now I am having doubts about it.  Best regards Xavier

OH,Oh..... wavelength = 300,000 / kHz or wavelength = 300 / mHz

I can see where you got the idea of wire length, I have to tell you the wire length, in the coil, is unrelated to wave length, which is 37.5 kilometers, at 8 kHz! Ooops. NOT meters. 37.5 meters is 8 mHz.

Wire size does not have much effect either, basically it is turns, of course fine wire means more turns, while thick insulation means less  distributed capacitance and less proximity loss. Finer wire means less current, either ac OR dc, and there are chart that give wire size, relative to current (without excessive heat.)

Yes, an LC bridge (I am working on an idea, using an op amp and a cheap digital frequency counter) and I think a simple Wien bridge oscillator may work, with a few xtras.

I have not looked at Christians BFO, I already have in mind a couple of circuits I want to try, besides, I do not like discriminate, instead I would  have one sounder but 2 LEDs: one red, for ferrous, and one green, for aluminum, gold, silver, etc, (good stuff,) non-ferrous.

I have been sweating/swearing over a mixer and the SE612 is perfect, and it will have the reference oscillator.

How BFO works is one oscillator is affected by the metal content, in the  ground: its frequency goes down for ferrous content and up for non-ferrous content. Simple,eh? So the search oscillator is mixed with the reference oscillator, to produce a tone..... if they are perfectly the same frequency, there is no tone or it is below human hearing, which begins around 100 Hz... this is why I do not like commercial BFOs, it has to shift at least 100 Hz, before you can hear it. At 850 Hz (comfort zone) you can distinguish 10 Hz variation, also, if you use 4 X the fundamental search frequency (better results with odd harmonics,) you get a gain in shift..... that is, with the 4th harmonic, the shift is 4X, so the 10 Hz becomes 40 Hz. Reversed, the search osc only needs to shift 2,5 Hz to give 10 at the 4th harmonic. It works, I made one years ago.

Better yet, is a phase sensor, because you only need a 1/4 Hz shift, from zero to full detector output! Same detector my VMSK uses. That is one reason I designed it.... one led to the other.

There has to be no interaction between the 2 oscillators, except in the mixer, if they are connected too closely, they can interlock, thus the shift of the search osc, shifts the reference osc, because they are interlocked, this lowers sensitivity. This IC I got recently, has a doubly balanced mixer, so that it is impossible for them to interact/interlock. Shielding and filtering does the rest.

Yes, a planar (flat)  coil should be easier to work with, up to a point: it would be good for a PI coil, and for 2 coils, one pancake on the other and add syrup! For a frequency sensitive, coil, you may not be able to get enough turns to get enough inductance. 

For a PI,  the inductance represents an ac resistance (impedance,) which varies with frequency. If you have a  pulse frequency of 500 Hz and you need 8 ohms, the inductance has to be 8 / (.0005 X 2pi) or 2546.55 uHy.....  that is another reason I do not like the commercial PI machines.... that inductance is too big for a planar coil.   

If you use 15.36 kHz, you can use a standard  30.72 kHz crystal, div. by 2..... use a one shot to get 4 or 5 usec pulses (the 15.36 kHz square wave are 32.55 usec long for a half cycle.) Now the pulse frequency of 15.36 kHz only needs 82.89 uHy for 8 ohms impedance.... easy to make a planar coil of this low inductance and it will carry the same current, as the bigger coil, at lower frequency. (I figured to use 16 ohms = 165.8 uHy and 165 ohm swamping R.)

The driver I want to use will provide about 1.5 amp through the coil, so the power is  1.5 X 1.5 X 8, or 18 watts. I think that may work, I stole that circuit from Honeywell, it was for the magnetometer ICs to give demagnetization of the IC bridge, inside the IC. It is a 5 ohm resistance requiring a very brief (5usec) pulse, at 1.5 amp. Sound familiar? 

When you have an 8 ohm coil, the swamping resistor should start at around 82 ohms and reduce this to find an optimum value, that will still control back EMF. The rough calculation is 10 x impedance, for starters.

The originator,s PDF says this high voltage pulse must be kept low... how every body gets the idea that higher is better, I do not understand, because the response is measured for phase, not voltage. It is the current that creates the field, not the voltage. 

The whole PI subject is quite contentious, so I am going to go with MY design, when I build one, and we  will see who is right.     

goldigger



Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,8129.msg58030.html#msg58030

		Quote:			Posted by xavier
	Hi goldigger How can I print all this stuff?  Ho yes I will re-design the casing of VLF as you are not the first person to tell me that it looks like a car (jeep) now I got a complex .   :'(Did the drawing of the PCB and will try to get all the stuff for it by the end of the month. Is there a way that I could print the post that you have send me as there is no way that I will be able to take all of that in just like that. If you design an inductance meter could I please have a copy of it as the one that I have found on the net looks good but you need to be able to do programming on processors and I know nothing about that. Ok now i'm going to do some work on my coil I want to see if it will behave better. Best regards Xavier

If you have FireFox, you can screengrab, either the whole page or visible parts; if you use Nitro PDF maker, the free version (I  will print the link, and if it is not live, just cut and paste into the browser,) you can (oops, not this forum, it does not work on the forum,) make a pdf and print that. OR, you can use the windows, print-screen facility, just get what you want in the screen, push the PrtSc button, call up Painnt, paste and save. Also, you can highlite everything, copy, then past into notepad. Can not find the address, just Google search  Nitro PDF, download 3.0.0.1, if you have FireFox.

Did I miss anything?

If I send a diagram, it will be a JPG or GIF and you can print to fit page, either letter or landscape. I can only upload JPG  or GIF and TXT. 

Right now, I am looking for a cheap frequency counter to build into a box with the LC meter, I have the data for a Wien bridge oscillator which is a sine wave osc. I also have some other stuf set up for measuring inductance and calculating distributed capacitance in winding your own search coils, then I realized I need a bit more. 

I would like to use a one chip counter like the ICM7216 but they are rare, now and worth about $45US! So I may go to a PIC16F84/6/628?

I have to work that out, yet.... the cheapest unit I have found is $9.99 but the readout was binary, with a row of LEDs, and another version sent the readout in Morse code, neither are really appropriate. All the other Frequency counters are around $35US.

If I can organize this gizmo, some how, I certainly will share.

Just to show you how well the copy and paste works look at my TXT upload of the quote of YOUR post!

goldigger

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,8129.msg58147.html#msg58147

		Quote:			Posted by xavier
	Hi Goldigger Thanks for the info I saw on the net a schematic where a NE555 was used now don't quote me on this but from what I can remember...... clip Best regards Xavier

OK.... It is MUCH simpler to use a resistor bridge, with the inductance or capacitance as one leg of the bridge.... a diode goes from each knee of the bridge, to a filter cap and a voltmeter, one side is  2 equal resistors, say 1k, the other side is a pot 1 to 5 k at the top side, and the bottom resistance on that side is two binding posts to which you connect the unknown value, even a resistor. Apply some ac, of appropriate b frequency, to the top of  the bridge (I will draw  you a picture)and with the unknown resistance/impedance at the BPs, set the pot for a voltage dip or null. disconnect the unknown and the ac and measure the pot resistance..... you can then calculate the unknown,s value. For an unknown resister it is easy... the pot equals the unknown resistor. THIS is an LCR bridge.

I will whomp up a sketch and upload it tomorrow... I may have one already made, I will check, and if so I will upload now.

Reverse the diodes, I have them drawn backwards, as it is more convenient to measure a positive voltage. Another case of anders kant bo. Maybe it is catching?

I will have to give you some formulae to calculate L and C:

for L: pot resistance / (frequency in mHz X 2pi)  of course 2pi is 6.283 and values are in uHy.

for C: (1/pot resistance) / (frequency in mHz X 2pi) the result is in ufd.... multiply by  1,000,0000 to get pf

Simple math, so it should not give problems. Of course for smaller values, you need higher frequency and for large values use lower frequency.

goldigger

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,8129.msg58174.html#msg58174

		Quote:			Posted by goldigger
	I would like to use a one chip counter like the ICM7216 but they are rare, now and worth about $45US! So I may go to a PIC16F84/6/628? goldigger

Hi Goldigger and Xavier,

I've built a LC meter device based on a PIC 16F628. There appear to be many LC meter projects/schematics spread across the web, many similar. The PIC 16F628 has an internal comparator which simplifies hardware. My meter is based directly on the "Digital LC Meter Version 2" which provides all info, firmware and pcb layout to build the device. Just google "Digital LC Meter Version 2" if you're interested.

Cheers
Xero

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,8129.msg58204.html#msg58204

		Quote:			Posted by xavier
	Hi Goldigger,Homefire and Xeronium Thanks for all the info     but the thing is the PIC 16F628 needs to be programmed    and I don?t have the hard ware nor the know how I have not got the daintiest clue as to how to go about it I don?t even know what it looks like. But it dose look simple enough to build so where do I get the holder to program it and how?? Or could I make such a holder see I don?t even know what its called (chip programmer chip holder ?) I have to get more into this  Best regards Xavier

Hi Xavier,

Writing the firmware to a chip is a fairly easy process, provided you have a programming device connected to a serial, parallel or usb port on your PC, and the software that supports the device to transfer the firmware (hex file) via the device to the chip. The programming device is basically used to apply a specific voltage to a specific pin on the chip which places the chip into programming/write mode, and the software does the rest. The software will tell whether the hex file was succesfully transfered.

The type of programmer (serial/parallel/usb) to use is often determined by what ports are available on your PC/laptop, also some PC's don't provide enough voltage over a particular port which may mean trying different devices until a good setup is found. On my PC I first tried a serial port (the simple yet popular JDM style) programmer which never worked for me, I then built a parallel port programmer (PIC-Brenner5) which has always worked great.

If you've got a spare serial port you can have my JDM programmer (just pm an postal address). You will need to download a free program called IC-Prog to transfer the hex via the device.

Best regards,

Xero



Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,8129.msg58224.html#msg58224