PICKINI V3 - an easy to build PI detector

[6666]

OSH park pcb's in the mail.

[N0QBH]

Pickini 3 testing progress.
The prototype wooden detector has been assembled and I've had maybe an hour total operating so far.


One of the first things I discovered was I have to keep the sensitivity under 2 volts and lower or it will start to false.
Also I have had the best behavior/performance with R23 adjusted to about a bit over 5.4v on my Fluke. That seems to be about the point where the PIC's diodes begin to conduct on mine.
With it set to 5.3v, the pulse was a lot wider on the scope as well. When I had it set at 5.3v the unit would start to false sporadically after a few minutes running.
The battery voltage hadn't changed.
For a test, I buried two newer US pennies, one at 4" and the other at 6". The 4" is a sold hit every time - even found it after a storm erased my mark.
The 6" is hit & miss. If I turn up the sensitivity a bit or keep the head on the dirt I can hit on it, but I'm right on the edge of it falsing.
Mike

[6666]

Invalid attachment .

[N0QBH]

Invalid attachment .

Try this

[N0QBH]

Pickini V3 instability issue.
While testing the Pickini V3 in my buried metal "garden" I've noticed that the unit becomes more sensitive a few minutes after powering up and keeps getting more sensitive until it starts sounding almost continuously after about 10 minutes of operation. Lowering the sensitivity pot seems to have no effect and the voltage to PIC pin #3 (measured with a VOM) will have dropped to 4.0 - 4.5 volts from it's normal 5.3-5.4 volts.
If I set the unit on the bench and leave it powered on, it will eventually go back to normal.
So far I've tried with the battery at full charge (12.5v) and down to 11.9v with the same results.
Changed the 5532 op amp and it made no difference. Adding a .1 uf bypass cap between op amp power pins made little difference.
The only thing that is not exactly to spec are the 10 uF caps. I inadvertently used non polarized electrolytics for C8,9 and 10.
Could this cause the op amp to drift like I'm seeing?
Regards,
Mike

[N0QBH]

Changed C9 to a polarized part to no avail. Going to put it on the bench and see if I can find the culprit.
Am very open to suggestions at this point.
Mike

[6666]

You are useing a regulated 12v supply ?

[6666]

The Pickini V3 pcb's have arrived and they are extremely well made with gold plated solder pads
I have 2 spare pcb's for sale, and will put an advert in the for sale forum.

[N0QBH]

No, I'm using a 3 cell LiPo 2200mAh RC battery.
I had tried a 4 cell with a 12v regulator early on and the problem appeared on the bench.
I switched to the 3 cell unregulated and the problem went away - on the bench, but reappeared while using it after a few minutes.
Last night I was testing after some more adjusting and noticed it didn't act up, so I have some more investigating to do.
Mike

You are useing a regulated 12v supply ?

[N0QBH]

After a lot of trial and error (mostly error) testing, I think I found the cause to the Pickini V3 instability issue I was seeing.
The sensitivity pot was intermittently flakey, putting the input to the PIC pin#6 all over the map.
I haven't taken it out to the field yet, but this is an answer that "fits" the symptoms and was easy to reproduce on the bench.
Stay tuned...
Mike

[N0QBH]

Replacing the sensitivity pot eliminated part of my problem, but the DC voltage level of the pulse at the PIC still sags down to 3.9v from the normal 5.3 volts while operating and the unit goes bananas. I've attached a voltmeter to monitor as I operate in order to get a better feel for what and when this happens. Setting the unit down in a non metal environment seems to slowly bring it back to normal, however when I try to use it again the cycle repeats.
Mike

[FatBob]

I bought a board for this project but have trouble reading the Eagle Schematic.
Why do they make it so hard to read?

Well I redrew it in ExpressSch. Here's a picture;



And the Express File;

Pickini V3.zip

I'm thinking of doing a dual input with separate virtual ground circuit like this;


I also started rewriting the code in Basic though the posted code is not current as it shows
too many pins on the PIC chip. It's a bit of a struggle as there is so many IF statements...

[N0QBH]

I let the smoke out of mine a few days ago while poking around with the VOM.
Got across the 2200uF cap and the circuit was powered with a LiPo battery.
Melted the main 12v trace and burnt the narrowest part of the ground return.
I've got a couple more pcb's to go, so maybe building a new unit will solve my issue of instability.
Else will try a different coil.

[F117]

@FatBob:
No idea why you would rewrite the software in Basic?
Here is the C source code for V3: https://github.com/bwillaert/MD/tree/master/Pickini3

I am not sure what you mean with " so many if statements " ?
The main idle loop just checks for flags that have been set in the interrupt handler and perform the real action IF needed.
I alwayst learnt that you shouldn't be doing anything really in the ISR itself, since this is already a couple of levels deeper on the minimalistic stack of the PIC.
Once you go one level too deep in the stack and you get memory corruption, you won't stop laughing. Believe me, been there, done that. All sort of unwanted behaviour will turn up unexpectedly and there is no way to recover from it, except a hard reboot, just to end up with the same mess some time later, resulting in a lot of hair scratching, and all sorts of other signals indicating that you have no idea what is going on.
So, I just set flags in the ISR and check in the main loop IF they occurred.

All the IF's in the ISR are just to go over every possible interrupt source and translate it into a main loop flag.

Success!,
- Bernard