PICKINI V4 - an easy to build, self adjusting PI detector

[sinclairuser]

hi all i'm currently tinkering around with the pickini, in my cap stash i have 3300uf 50v elcos, 3300uf 25v and 2200uf 16v,
but not 2200uf 25v, so what do you all think should i use 3300 25v or2200 16v?.
normally i would use the 3300uf at 25v, instead of 2200uf 16v better to use a slightly bigger cap than a lower voltage version.
but reading through the thread turn on inrush current can be a problem here, will the extra capacitance be a problem, or will a 16v 2200uf cap be ok?.
thanks,
aly.

[waltr]

If pi pulse width is 100 ms. , a little golden thing, sees your device 8-10cm. then with a width of 250ms. will see 10-15cm.

Do you mean us (micro second) not milli-second?

oh filos, if only it was that simple.

Correct. Also for low TC targets increasing the TX pulse time does NOT increase the target response. Small gold object have low (short) TCs and a 50us pulse will fully excite them.
On the other hand, long TC objects will give a larger response with longer TX pulse. Silver coins, for example, respond best with a >200us pulse.

[sinclairuser]

If pi pulse width is 100 ms. , a little golden thing, sees your device 8-10cm. then with a width of 250ms. will see 10-15cm.

oh filos, if only it was that simple.

[filos]

If pi pulse width is 100 ms. , a little golden thing, sees your device 8-10cm. then with a width of 250ms. will see 10-15cm.

[BC547]

@ filos

To increase "the power" you probably mean: apply a longer pulse to the FET ?
At the moment this is fixed to 100 microseconds.
If you can build the source code for the PIC yourself, it should be obvious to change this setting:

#define TX_PULSE_WIDTH 100 // TX pulse width in microseconds

The only time I increased this to 200 microseconds was for a 1m x 1m square coil to detect WW1 unexploded ordnance here in Flanders Fields.
These grenades are typically 30-50 cm long and 10-20 cm diameter iron shell that can easily be detected well over 1 m in depth, resulting in a lot of digging. Well... I detect and mark suspicious targets and other people do the digging. There are still thousands - nobody knows exactly how many - of these things buried here.
Increasing the pulse width to 200 microseconds instead of 100 also doubles the current consumption.

Here is the square coil setup - on plastic wheels from a small child bicycle - and some "treasures" retrieved with it ...



Control box - hip mounted with 12V Li Ion battery mounted under the PCB:
The pinpoint pushbutton was replaced by a switch because I needed both hands to pull the cart.



Some finds - a large grenade + unexploded detonator:



A nice collection of finds from a piece of farmland that was going to be plowed for the very first time in 100 years.
It used to be a pasture for cows. As a farmer, you don't want the plow to run over the detonator of one of these :



In my experience: increasing the power results in a lot of deep digging - and a lot of back ache in the end.
A soda can burried on the beach gives a clear signal on 1m depth with the standard settings ( depending on your coil of course ).
So, - no thanks - I don't need more power...

[filos]

Is it possible to increase the power of your device ..

[BC547]

@moorejl57:
Earth Field Effect in voltage sampling is countermeasured by taking 2 samples - one much later than the other and subtracting the 2 samples.
With pulse width measurement, there is only 1 sample.
In this design, the samples are accumulated and go through a low pass filter = running average.
Both the number of accumulated samples and the running average array size are influenced by the sensitivity setting.
The source code is available here if you are interested or have any tips to improve it : http://users.telenet.be/willaert/MD/.../v4/Pickini4.c

[6666]

@666:
library: con-wago-508
device ( 3 pins) : W237-3E
device ( 2 pins) : W237-02P

Thanks

[BC547]

@666:
library: con-wago-508
device ( 3 pins) : W237-3E
device ( 2 pins) : W237-02P

[6666]

Hi Bernard, in eagle what are the 2 and 3 terminal blocks parts called ? thanks

[moorejl57]

Fantastic project Bernard, along with your patience in helping folks out with the build. I read through much of the thread and the info on the project site and have a quick question. How does the temporal method handle EFE at the end of the swing?

[BC547]

Guzzi,

you can find all construction details here : http://users.telenet.be/willaert/MD/.../V4/intro.html

The PCB and schematic page: http://users.telenet.be/willaert/MD/...ectronics.html
.
It is a single sided PCB with a ground plane and the FET is IRF740.

You can make it easy for yourself by buying the semi-kit with the PCB and all semi conductors including the programmed PIC controller:


Regards,
- Bernard

[Guzzi]

Is this a double sided PCB with ground plane? What type switching fet is used ?

[6666]

The chip seems to test ok thanks.

[6666]

The sensitivity potmeter between 5v and Gnd can be 4k7 or 10K.
It is not listed in the parts list.
The parts list is just a copy of the BOM generated by Eagle - and since the potmeter is not on the PCB, it's not on the parts list.
Neither is the audio jack, on/off switch, pinpoint pushbutton, ... anything that is connected via wires to the PCB.

If you power the microcontroller only - 5V only, it should produce the welcome sound on the audio output.
And after a while, a low error sound because the coil flyback is not detected ( coil not connected ).
You should also have a 100 microsecond pulse every 2 ms on pin 5 of the controller - this goes to the IRF gate after level conversion to 12V.

Thanks reply, will try and test the chip tomorrow