PICKINI V3 - an easy to build PI detector

[BC547]

I updated the hex file for PICKINI V3 here:

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

Improved sensitivity:
- pulse frequency is now exactly 500 Hz. It was slightly lower
- increased pulse width to 120 microseconds (was 100 microseconds)
- increased the pulse slicing level to 3.75 V (was 3.25 V).

Current consumptiion has gone up to 110mA now, which is more in the line of V1/2.

When this new hex file is programmed into an existing V3, you will need to readjust the multiturn R23 for 5V at the microcontroller input pin 3.

Enjoy,
- Bernard

[BC547]

I added a monochrome layout PCB file of the battery charger here:

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

Scroll to the bottom of the page and open the PDF: this is the copper side of the battery charger PCB.

[Arionas]

Pcb please for battery charger board
Thanxs..........

[BC547]

Best regards,
- Bernard

[bernte_one]

here are the latest pictures
pickini v3
lm 386 amp
14500 power source

[BC547]

it is so that my v3 is not sensible/ deep like the v2

The average DC current consumption for V2 in "high frequency" mode (500 Hz) is 135 mA.
The average DC current consumption for V3 - always at 500 Hz is 90 mA.
I assume that the main difference is the different front end: V2 = IRF740 (400VDSS) vs V3 = IRF9640 (200VDSS). RDSon is about the same for both: 0.5 Ohm.

Maybe a more experienced analog designer can enlighten the difference ?

One other thing to try is to sample the flyback pulse closer to the return value => closer to instability.
You can simulate this by adjusting the DC offset of the opamp output a little lower = DC voltage at pin 3 of the PIC. Adjust for e.g. 4.7V instead of 5V.

[bernte_one]

hey bernard

thanks for the info
i use them http://www.ebay.de/itm/5pcs-LM386-Su...item3a90f1d371

but use only minimal setup with gain of 20 instead of 200

it is so that my v3 is not sensible/ deep like the v2
but it is very easy to build (low part count,low cost, low time)

and in relation to simplicity it performs well

powersource are 3 cells 14500 3,7V (trustfire,ultrafire..) with a triple battery holder or external case with switch

http://www.ebay.de/itm/like/370400350159?lpid=106&_configDebug=ViewItemDictionary.ENABLE_PAYMENTS_IN_HLP:true&hlpht=true&ops=true&viphx=1

[BC547]

@Bernte-one : For another application, I needed a compact stereo audio amplifier and bought a couple of these modules ( less than 2€ ) on Miniinthebox:

http://www.miniinthebox.com/nl/high-power-2-kanaals-3w-audio-super-mini-digital-amplifier-board_p905588.html

Delivery within 3 weeks. 5V power supply (available from PICKINI board ).
Sometimes you don't / can't build it yourself when you consider the price / effort ratio ...

Best regards,
- Bernard

[BC547]

Thanks. The main target of this design revision was to keep it simpler and smaller.

Does it suffer from ground effect and are you happy with the sensitivity...?

GEB is adjusted continuously in software - much like the software described in ITMD.
Sensitivity can be adjusted by accumulating more or less samples. Only in a very quiet environment, like the beach or a large field can I crank up the sensitivity.
Overall performance is largely influenced by the type of coil - like with every PI detector.
Time domain sampling can become unstable as you sample close to the flat line at the end of the flyback pulse.
Is equivalent to unstability in the voltage domain when sampling to soon after the pulse, during the vertical line of the flyback.

[Dean Sarelius]

Hi Bernard,
I have also been experimenting to sample the return signal in the time domain. I like your circuit as it is very simple but I am curious to know what you think of the performance overall. Does it suffer from ground effect and are you happy with the sensitivity...?

[bernte_one]

ah ok

thanks bernard

[BC547]

btw: is it normal that it takes around a minute or more that the detector reaches the full performance

It may have to do with the "virtual ground" setup. It takes some time before the circuit around R9/R14/C6 stabilizes --- around 10 seconds.
Final sensitivity should be equal to V2. Largely dependent on the coil used.

[bernte_one]

hello
my v3 works good now
i replaced the standard speaker i used before with a piezo speaker
(v3->LM386 Amp with Gain 20 -> Strong Piezo Speaker up to 12V)

btw: is it normal that it takes around a minute or more that the detector reaches the full performace
or did i take some wrong temperature sensible parts?

the next think i hope is that more will do the new version so that we can compare the results

i will post SL6.lay files tonight or tomorrow
for those who work with SL

regards

[BC547]

I checked the code for V1/V2 again last night and the bug that was mentioned here around the sensitivity potmeter was introduced only later, during the V3 development phase.
Besides, I never noticed it during field tests with V1/V2 and never got complaints about this.
So no worries about (known) SW bugs on the download site.

Enjoy,
- Bernard

[sinclairuser]

Hi Sinclairuser!
I made both versions and I am pleased with the work of the detectors. The only thing I do not like is the sound with a frequency of 1 hertz, which shows that the detector works. I would like the sound to be heard every 10 or 20 seconds when to show that the detector works. To distinguish it from other signals sound can be a double pop of. This is my opinion, which do not require the author in any way.
Of course there is no difficulty from Eagle PCB file to make board by method "iron toner transfer", but I would like to have the updated hex files for ver. 1 and ver. 2.
I wish success to all who make ver. 3. Least I will later do ver. 3 and compare the performance of the detectors.

me too, sorry if my post came across as snotty that was not intended, it is unusuall for any developer either pro or hobbyist to "fix" previous versions, so i take my hat off to f117, those posters who asked about porting this version to the last chip can now rest easy, if ver 1+2 is going to be fixed, the there are plenty of options for builders who can only acquire previous chips, so a winner all round.