New Idea

I am a new member on this site, but I like it already. This is to all of you that have designed a PI Detector, or built one. I have been experimenting with a all kinds of detectors for years, but I have gravitated to PI. I have noticed that most of the PI designs seem to be all the same. Explanations of how the circuits work vary widely. All seem to have chosen a power mosfet as the coil driver and I agree. After that I can see design flaws every where.

#1. Pulse generation for the driver mosfet. It must be stable, of the correct pulse width and rate and duty cycle. Using a 555 is out. Use a microprocessor or a clock module with a divider chain. Pulse rates should be between 120 and 400 Hz. Pulse widths should be between 120 and 250 microseconds. Duty cycle between 3% and 4%.

#2. Use a single power source. 12 to 20 volts. Do not use derived power supplies.

#3. Do not try to detect the signal directly off the Transmit coil. Use a secondary coil with the same number of windings, and reference it to ground. Eliminate the derived negative portion of the pulse with an in series signal diode of the proper voltage (200 volts).  Skip the high gain amplifier, and diodes across the coil. Reduce the resulting CEMF pulse down to about 10 volts with a resistive divider and not a zener.

#4. This pulse is what you need to look at and not the delay time after it so ditch all the fancy pulses, electronic switches, etc. They are not needed. Pay attention to what the pulse is doing when you place ferrous and no-ferrous metals near the coil. Ferrous metals widen the base and drop the amplitude slightly. Non-ferrous metals narrow the pulse width and drop the amplitude a lot more. Detect the width of the pulse half way up the pulse and you will have a way to discriminate metals.

#5. Eliminate as many components that are naturally drifty.( Capacitors, diodes, transistors). Thermal components can make a detector very hard to stabilize. Eliminate as many controls as possible. Who wants to spend all day metal detecting twiddling with knobs anyway. The simpler and more stable that a detector is the more finds you will be digging.

I hope I did not come across as a "know-it-all", because I am  not. This advice has been derived from years of experiment, building and testing, and research. How many of you have tried any of these ideas or are working on it now? Please reply. I want to hear your experience and maybe you can teach me something too.

Linkback:

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

http://www.thunting.com/smf/index.php/topic,56782.msg275387.html#msg275387

A bad day at gold prospecting or metal detecting is better than a good day at work.

Been designing Pi defectors for some time now so particularly interested in this thread.  I have been using PIC's for the past three or so years in my Pi designs and yes dual coils are an advantage when doing DSP but for the regular PI single coils work well. One may say that the design is not good on paper but in the field they work well,  theory and practice are two of a kind.
.
Regards

Linkback:

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

http://www.thunting.com/smf/index.php/topic,56782.msg275390.html#msg275390

Xavier

I also use i simple pi mono coil like i learned some years before on old pulsdetektor forum.
I improved my way of winding and shield.
And with the right electronics i can identify my targets very good (iron/non-iron)

But you are also right in the points you call.
Simpler circuits are better, because less can go wrong.

regards

Linkback:

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

http://www.thunting.com/smf/index.php/topic,56782.msg275441.html#msg275441

my machines at work http://mueller-meteoriten.de

Sorry mate but I can't agree with you on this one

Linkback:

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

http://www.thunting.com/smf/index.php/topic,56782.msg275468.html#msg275468

Xavier

Just think of it this way, to get all your timings right with an analogue system you will need four ne555 and all the components that go around them and should you want to change the timing you have to start replacing resistors or capacitors, with a digital system you don't have all these components and you can change your timing in a matter of seconds,  it's less messy and it can be done in the field.  I honestly think that you should give digital detectors more consideration.

Linkback:

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

http://www.thunting.com/smf/index.php/topic,56782.msg275471.html#msg275471

Xavier