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

[BC547]

Here are my project settings:

[eclipse]

Errr... I'm pretty sure cracking the software won't make a difference.
It's probably due to different project settings.

[Arionas]

[eclipse]

While waiting for the parts to arrive I tried to play with the firmware a bit.
I used MikroC 6.6.3 (latest) to compile the source. It compiles without errors however the hex file is not the same as the original.
Perhaps some setting in Project/Edit Project Menu is different?

[Tim]

So the faster the coil, the faster the natural delay will be? And higher the sensitivity will be able to be adjusted??
I am very eager for my pickit to arrive. This sounds like a great project.
Yes effect of comparator level will be interesting. How about RC3/pin7? Can we utilise this for another sample input?
More food for thought Bernard.

[BC547]

@Tim & Ivconic: thank you for all the extra inspirational thoughts !

Variable offset on op-amp.
Comparator reference adjustment.

These 2 are now implemented in the Pickini V4 version.
The comparator slicing level ( reference ) could be a potmeter 0..5V if you wan to have full manual control and like fine tuning.
No idea if this will be useful in a highly mineralized environment.
Also: is there any difference when measuring the pulse width between a slow coil and a very fast one ? ( eg. 10 microseconds vs 100 microseconds )
The temporal change should be far less with a fast coil, but still detectable with a high resolution timer like this ?

I was under the impression that temporal sampling would be less coil dependent than delayed voltage sampling since everything is relative, independent of the absoulute pulse width.
The closer the pulse width can be measured to the steady state value, the more sensitive the detector.

Best regards,
- Bernard

[Tim]

Yes, thanks Ivconic. I have read this and much other theory and testing...So... Looking at 500-1000hZ, 100-160 pulse width, 6-10uS Delay-for a single-sample. I have read the on-time etc. for double samples, and checked the scopics of the SD2000.(Although I don't know these off by heart).
But which of, and how, can these specifications be related or utilised in a temporal domain detector?
I have seen cut-and-shut designs here, and realise many of your masterpieces utilise sections from previous designs. Everything old is new again.
Basically I would like to see enough interest here to stimulate F117's request to optimise the front end.

Some ideas that may or may not float.
Op-amp change.
Variable offset on op-amp.
Comparator reference adjustment.
Maybe the two 74HC221 circuits from the Hammerhead can be adapted as secondary and tertiary delay? One through RC3/pin7?

Please don't laugh at me.

[ivconic]

These are good guidelines, post #63:

http://www.geotech1.com/forums/showthread.php?23417-PI-TX-Settling/page3

[Tim]

Perhaps George or Ivconic or other knowledgeable persons here could make some op-amp/front-end design suggestions? I would dearly love to see more here at Geotech regarding small gold, and brutal mineralisation, as opposed to relics, coins, and deep stuff in "easy" ground.......??
The only things I have found here truly relevant to gold, is the ancient Goldmaster schem, and Lobo. As far as mineralisation, 2 samplers like HH, Goldscan, Pulse1&2, limited stuff on the Mirage. The SD2000 is a pipe dream for all but the most dedicated and learned.
Perhaps there is a way to incorporate temporal sampling, and a multiple ground samples (a-la Hammerhead) to "punch through" the hard ground?
Food for thought?

Thanks for this F117. I have my pics. Now awaiting the pickit on the slow boat from China.

[ripsdevala]

houra!!!!!!!! it is alive!!!!!!!!
it was just one connection between pic pin 8 and 9
whouaiiiiii super !!!!
ready for tomorow beach hunting but this time in wet sand (euro ace can t!!!)
thanks a super lot!!!
just have to fine finish my coil put all that on my shaft and GO! GO! GO!
treasure and gold beware i m coming !!!
send you a foto of finish product
thanks again

Congrats!
If you have a video test that would be great, its been a week but my chip is still stuck on china

[BC547]

@Eclipse

Perhaps instead of C9 (10uF) we can use 220nF-2.2uF polypropylene type capacitor (for lower ESR, lower noise)?

I am not an analog electronics specialist. The only thing I have observed on the scope is that with a pulse duration of 100 microseconds and a pulse cycle of 2 ms, the coupling capacitor has to be large enough to maintain the DC level over the entire period.
When the capacitor value is decreased, it starts acting as a differentiator, resulting in a negative going pulse = flyback pulse, then returning to DC level and gradually dropping again over the rest of the 2 ms.
I am sure that this effect can be overcome in software by just enabling the pulse comparator during the flyback pulse and discarding any level drop later on in the cycle.
This could be an improvement, but so far all the tests I have done were with this 10 microfarad value.

To elaborate even further on this: I am sure that the entire analog front end can be optimized in combination with the PICs capabilities of automatically calibrating/ adjusting some DC levels.
As I mentioned before: I am not an analog electronics guru. But I would like to team up with one to optimize software and hardware capabilities to come to an even better and maybe final solution for a temporal sampling detector ( Pickini V5 ?). Any ideas/inputs in this direction are more than welcome. Probably starting with an entire DC coupled simple front end and a negative supply voltage for the opamp ( ? ) . Or even considering the usage of a PIC with a built-in opamp - (bandwidth ??? ) ....

Best regards,
- Bernard

[BC547]

@Arionas: PCB mask is OK now. Nice work !

[Arionas]

@ Nick Holas: glad to hear it works now. Happy hunting !
@ Arionas: the layout you are showing is OK.
The only thing that I see wrong is the type of PIC controller: it is not 12F1824 but 16F1824 on the PCB mask. A minor detail, but people might purchase the wrong part.

Have fun,
- Bernard

Good Morning Mr F117
THE MISTAKE WAS MY OWN ...... WHEN WE BUILD SOMETHING ALWAYS follows the pattern

[eclipse]

Perhaps instead of C9 (10uF) we can use 220nF-2.2uF polypropylene type capacitor (for lower ESR, lower noise)?

[BC547]

@ Nick Holas: glad to hear it works now. Happy hunting !
@ Arionas: the layout you are showing is OK.
The only thing that I see wrong is the type of PIC controller: it is not 12F1824 but 16F1824 on the PCB mask. A minor detail, but people might purchase the wrong part.

Have fun,
- Bernard