New Homebrew Metal Detector

[zisme]

I just posted the plans for my new P.I. detector at: http://oldradiobuilder.com/MDET.html Let me know what you think of it. Thanks, Mike

[Carlos]

PCB

Hello, you can put more resolution for the PCB because I can not view whit details the lines and the holes of your circuit-
thanks

Carlos

[zisme]

Hello, you can put more resolution for the PCB because I can not view whit details the lines and the holes of your circuit-
thanks

Carlos

Carlos,

Try right clicking, saving the picture then printing it at a smaller size. I will try to improve the image when I get a chance.

Thanks for the feedback,

Mike

[okantex]

hi
did you have field or lab tests for it.
1 euro coin from ? cm
and do you have big coil

[zisme]

hi
did you have field or lab tests for it.
1 euro coin from ? cm
and do you have big coil

I only have my own air tests which are as follows:

U.S. penny: 12 to 15 inches

U.S. nickel: 9.5 to 10 inches

automobile: 4 to 5 feet

These are the absolute minimum detection distances using both sound and meter.

The 9.5 inch coil is the only size I have tried on this detector.

Mike

[Unregistered]

Getting warm

Hi,
I suspect the reason your FET is getting hot is because you don't have a proper drive circuit for it. I had this problem until I used a good gate drive circuit. FETs in this mode want to be turned on or turned off and nothing in between so they must be swithed on and off very fast (ns) and hard. I found this was my problem after putting in larger FETS, heatsinks, adjusting pulse rates, etc and then I finally did some reading and analyzing of what was going on and realized the problem. I thought it was the high current I was pulsing but that wasn't it.
Good luck

PS-now design one using a single 9V battery.

[zisme]

Hi,
I suspect the reason your FET is getting hot is because you don't have a proper drive circuit for it. I had this problem until I used a good gate drive circuit. FETs in this mode want to be turned on or turned off and nothing in between so they must be swithed on and off very fast (ns) and hard. I found this was my problem after putting in larger FETS, heatsinks, adjusting pulse rates, etc and then I finally did some reading and analyzing of what was going on and realized the problem. I thought it was the high current I was pulsing but that wasn't it.
Good luck

PS-now design one using a single 9V battery.

You are absolutely correct about switching FET's on and off. The faster the better. I have been designing power MOSFET circuits for years at the last place I worked. Here, I was driving automobile headlights and other high current lighting. I was using high and low side driver/conditioner IC's which work very well. Some of them used charge pumps to create the driving voltage for N channel devices that could supply upwards of 100 amps to the load. Driving high current FET's was hard because of the high gate capacity these devices have. You need fast switching times, but you also need high current drive to overcome this internal capacity.

However, with the P.I. detector drive, there is a current limiting resistor in series with the gate (R27) that does slow down the switching times somewhat. The advantages of using R27 is to reduce the noise on the circuit board. So, there is a little heat generated at the trade-off of less noise.

With the pulse width and frequency combination I am using, the FET does not get hot or even warm. I allowed space for heatsinks on the circuit board so you could experiment with the timing without worrying about over-heating the FET's. Notice the board that's mounted in the box has no heatsinks at all. As you know, increasing the TX pulse width and/or increasing the frequency will eventually reach a point that will warm up the driving FET no matter how fast they are switching on and off. You could replace R27 with a short, it won't hurt anything, but you will notice an increase in noise.

I have played around with lower voltage circuits, but I didn't get the depth I get with 12 volts. If I could, I would be using 24 or maybe even 36 volts! But you would need a pack mule to carry it around.

Thanks for the comments,

Mike

[Power]

air tests

25x25cm copper plate at 90cm.Is this enough or can i make it better?
Also the current is 20 to 40 mA without vco, when i have sound the total current is about 100 to 120 mA.Is anything wrong?

[Geo]

25x25cm copper plate at 90cm.Is this enough or can i make it better?
Also the current is 20 to 40 mA without vco, when i have sound the total current is about 100 to 120 mA.Is anything wrong?

The detection distance is very good for a coil 8 to 11 in but no good for biger coil. Now about current if i will see the diagram of detector i can tell you but i believe that 100ma is a normal current for a PI detector.

[zisme]

25x25cm copper plate at 90cm.Is this enough or can i make it better?
Also the current is 20 to 40 mA without vco, when i have sound the total current is about 100 to 120 mA.Is anything wrong?

That sounds like a good distance if you are using the 9 1/2 inch coil. I would have to find a 25 cm copper plate to know for sure. I used the following test:

U.S. penny: 12 to 15 inches

U.S. nickel: 9.5 to 10 inches

I would think you might get even better results with a better coil than what I am using. Others have suggested that using Teflon insulated wire or Litz wire will give better results. Also, I glued my coil to a piece of composite fiber board and mounted it inside a case made of the same stuff, which lowers the Q somewhat. The best coil would probably be one that has no surrounding case, but the windings are just held together with epoxy or some other glue. The addition of a shield was suggested, but it seemed to reduce the range a little without improving anything. I am in the process of making a new light-weight coil. I will post the results ASAP.

You can also experiment with the timing to try and improve your results. The charts I supply, are for reference only and are optmized for my coil. Different coils will mean different timing setups, so even if you have a coil made in the same dimensions as mine, some deviation in adjustments will probably improve it.

The current drain on yours is actually a little better than mine. If you are using a speaker for the transducer, make sure that R26 is not installed. Also, raising the value of R25 will lower the current with the audio on. This lowers the volume too, but I think it's too loud anyway. For best current drain, use a piezo transducer.

As I say, this project is a basis for experimentation and improvement, so go at it and let me know your results.

Thanks for the questions and thanks for building it,

Mike

[Esteban]

Search for keywords bbsailor and read all instructive hints about PI coils.

[ivconic]

Hi

Hi Zisme ,
this looks interesting to experiment with!
My handmade pcb is ready, now i am waiting components from provider....
Can you tell more about used coil here?

"If your MOSFET is getting too warm to touch at the TX frequency you choose,
use thinner wire in your coil or add a series resistor of about 2 to 4 ohms and
most importantly, add a heat sink!
....If the MOSFET is getting warm that means that you have exceeded the voltage
rating of that particular MOSFET......(bbsailor)"

A higher peak pulse voltage makes fet hot. (IRF 640-200v, IRF 740-400v....)etc.


regards

[ivconic]

pcb

here is a handmade pcb....

[zisme]

Hi Zisme ,
this looks interesting to experiment with!
My handmade pcb is ready, now i am waiting components from provider....
Can you tell more about used coil here?

"If your MOSFET is getting too warm to touch at the TX frequency you choose,
use thinner wire in your coil or add a series resistor of about 2 to 4 ohms and
most importantly, add a heat sink!
....If the MOSFET is getting warm that means that you have exceeded the voltage
rating of that particular MOSFET......(bbsailor)"

A higher peak pulse voltage makes fet hot. (IRF 640-200v, IRF 740-400v....)etc.


regards

I just wanted to say that I am not having any overheating problems with the timing values I am using. I have added the following to the website to clear up the confusion:

(Added on 6-10-06: I have been receiving e-mails with solutions for reducing the heat in Q1. I wanted to clarify this. Q1 does not get hot with the timing settings I am using. I only included heat sinks on Q1 and the voltage regulator to allow room for experimenting without worrying about overheating these devices. The pot values indicated can allow enough room for adjustments to reach a point where Q1 and VR1 will warm up, but when properly adjusted for best results, you will not have any overheating problems. In fact, the circuit board I used in the completed detector does not have them. Mike)


I couldn't tell you much more about the coil except I did make a coil winding machine in order to wind it. It still didn't turn out much better than just winding it by hand.

Great to hear that another person is building it. Please let me know the results.

Thanks,

Mike

[zisme]

here is a handmade pcb....

Wow! I can't believe you made this by hand! Great job!

Don't forget to install the three jumpers first. lay them flat.

If you need any help in getting parts, or adjusting it, let me know.

Again, thanks,

Mike