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

[PIERE]

OK for everything Bernard!
I have found the hex file for Pickini v4.0- v4.1-v4.2 on the page I indicated.-
As soon as possible I try all the versions to find the best one!!
NO, the electronics box is not too close to the coil!!.-
Now I send you the photos of my construction.-
I think that's fine.-
Thank you very much Bernard, I will let you know!!.-
Regards

[BC547]

PIRE,
you can find the hex file for Pickini v4.0 on the page I indicated: http://users.telenet.be/willaert/MD/...Pickini4.0.hex
Together with v4.1 and v4.2.

Nice construction !
Just make sure that the electronics box is not too close to the coil.
When you also store the batteries in this box, you may have 2 adverse effects - when close to the coil:
- the magnetic field may be "disturbed". Not sure if you will notice anything though...- as long as the box is not moving with respect to the coil.
- the balance of the detector will be wrong. In order to save your elbow/arm it is better to move the weight backwards, behind the handle, e.g. under the armrest, to have the detector in balance once the coil is mounted.
From my personal experience, I know that swinging a detector that is not properly balanced can be painful afterwards ( for a long time ).

Success,
- Bernard

[6666]

other photos of my Pickini 4
[ATTACH]47843[/ATTACH][ATTACH]47843[/ATTACH][ATTACH]47843[/ATTACH][ATTACH]47844[/ATTACH][ATTACH]47845[/ATTACH]

Very inventive, all DIY, very good

[PIERE]

other photos of my Pickini 4

[PIERE]

OK, OK!!, Dear Bernard
I thank you for your prompt reply, and for your explanation,yes I want to try version 4.2.hex as soon as another PIC 16f1824 arrives where I can load the new firmware.-then I'll let you know how it works !!
sorry but to your internet address I seem to find only version 4.1 and not 4.2, maybe I'm wrong?Thanks again Bernard for your suggestions.Regards

[BC547]

Pickini V4 software versions

The difference between PickiniV4.0 and 4.1 is that in the old version, the value for the comparator offset level, = DAC output from the microcontroller, always starts from the middle value and tries to achieve 4V for the analog output DC level = microcontroller comparator input for the analog flyback pulse.
In Pickini4.1.hex, this value is stored in EEPROM and is the starting value for the successive approximation, instead of this middle value. This leads to a shorter offset calibration time.
It has been reported by a couple of people that this method sometimes leads to instable behaviour. Personally I have never experienced this and don't have a root cause for it. It may have to do with EMI inside the building...
Anyway, here: http://users.telenet.be/willaert/MD/.../software.html
are 3 versions available:

• Pickini4.0.hex = offset calibration starting from the middle and then successive approximation = longer calibration time
• Pickini4.1.hex = offset starting value taken from EEPROM = previous calibration value = shorter calibration time
• Pickini4.2.hex = senisitivity setting influences the comparator slicing level: it starts from an initial default setting and adjusting the sensitivity will decrease this level = less sensitive or increase the slicing level = more and more sensitive as it approaches the flyback steady state level. Up to a point where it will become unstable and if cranked up even higher, the "coil error" low tone indicates that the slicing level is higher than the flyback puls level, leading to absence of digital flyback output at the comparator ( identical to a coil short or open ).

I recently started field testing with Pickini4.2.hex = the newest version and if I were you, I would give this a try since it offers the most degree of freedom to the sensitivity adjustment.
Altering the slicing level in temporal sampling can be compared to altering the sample delay in voltage sampling: the closer to the edge, the more instability - but more sensitivity.
With the latest version, you are no longer dependent on the automatic generated values for offset and slice level, instead you have manual control of the slicing level.
Sensitivity setting now controls the slicing DC level instead of altering the number of samples accumulated in v4.0 and v4.1.

regards,
- Bernard

[PIERE]

Hi Bernard,
GOOD Project!!

I built the pickini 4 and it is a nice project, but in my case the defects already highlighted in post 294, have occurred
“It seems that with my build, the detector is not as stable with the new firmware. It randomly and often beeps after the initialization, although the detection depth is about the same as with the original firmware”.

Can you post the hex file of the first version of pickini4?

You said that The difference in calibration is that the starting point for the DC offset is now kept in EEPROM and started from there.
Previously, it was a successive approximation, starting from the middle value.
I want to try the first version to see if it works better!!
Thank you very much Bernard and excuse my bad english!!

[ripsdevala]

Thanks for the info

[BC547]

As long as this small piece of metal is not moving with respect to the coil, that is ok. But try to limit the metal parts to a minimum. Cat 5 cable is also ok.

One tip yet: do not test this coil indoors. It will pick up all possible noise signals in the building. Go to a place away from EMI. And do not wear safety shoes with metal tips like the first time i tested that cart. Signaled every step...

[ripsdevala]

What about cat5 cable, i can't find any stranded cat4 lan cable.
Theres a metal rod on the cart wheels, is that okay?

[6666]

That cart is a great idea.

[ripsdevala]

No need to modifying the circuit and firmware right, i read on this thread to increase the pulse timing for better results. I will create the 1x1m sensor first. Thanks

[BC547]

I summarized some popular coil dimensions here:

http://users.telenet.be/willaert/MD/pickini/pickini_en/V3/coil.html

For a 1m coil I usually take 5 twisted pairs from a network cable - 4 m long.
In a standard CAT4 cable, there are 4 twisted pairs, so you will need another length of 4 m for the 5th pair.


Bundlle them together in a new cable by wrapping it with electrical insulation tape:



Then a layer of shielding ( alu foil eg ) and then another layer of electrical tape.

You can then place the cable in a PVC drain pipe frame and connect the ends of the wires to have a bundled coil of around 350 microHenry and 2-5 ohm:



Add a control box with batteries:



Drag it behind you, carry it or mount it on plastic wheels:

[ripsdevala]

Hello benard,
One of my collage ask me to create 1x1m sensor for 1m+ deep for big target, lets says 10x10cm plate, beach area.
What should i need to modify, and what is the parameters of the coil?

[sinclairuser]

thanks bernard, i only use ceramics in really small values, i like box caps in amps and detectors just a personal preference, i once had some bad ceramics that caused oscillations, as for the different types i would have used red wima's in all positions but couldn't find my box of them, by the time i found them i had already built the power supply.
as for the heatsink i bought some for a repair i did that used 26 of them and the cheapest way to get a quantity was to buy 100 of them so every mosfet or regulator in that package gets a heatsink just to use them up.
aly.