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

[F117]

The source from mikroC does not compile in MPLAB, is it normal ?

It was designed entirely with mikroC and the mikroC libraries, so I can assume that some library functions for timing delay, e.g. are not compatible with the MPLab libraries.

I wonder if you can not just attach the vibrator to the audio output. It wil click once every 2 s, but the 1kHz detection tone should result in a continous vibration signal ?

Does MPLAB needs to be installed to make PICKIT working ?

I only used MPLAB as an interface to PICKit3 to program the hex file generated from mikroC.
You probably found this info on how to use PICKit3 directly with mikcroC ?
http://forum.mikroe.com/viewtopic.php?p=110422
I never tried it, I just used MPLab because it worked for me like that for many years already...

[reptooyep]

On the DIP version (not modifyed), i cable the vibrator directly
In place of the audio but in my SMD version, i power the vib. with a bc848 transistor.
To test my SMD version, i'm waiting for the PIC and the IRF740 on the way from China.

To flash with MPLAB, you have to open the HEX from the IDE and configure the project
With the pickit3 and just burn the PIC. MicroC can do the same but i did not install the programmer.
In fact, with this configuration, you have 3 ways to burn the PIC.

[F117]

@reptooyep: Please keep us informed about the build - DIP version first ?
Post some pictures of the PCB's and the end result + performance.

@all: Happy New Year - all the best in 2017, including more coins than pulltabs

Thanks,
- Bernard

[reptooyep]

Hi Bernard,
First, let me thank you so much to share your source code for this detector. It's not usual and it's very kind.
As you know, i want to drive a vibrator with a NPN transistor and the output of the PIC.
I'm quite amazed by your code but i have troubles to identify the detection sequence that drive the PWM output cause i have no experience on timers and PWM.
I want to replace the PWM audio command by :
Disable Interrupt
500ms output high on PIN 11
output low on PIN 11
Enable Interrupt

Can you give me a hint about this ?

I'll continue searching by myself while waiting for your reply.

Thanks again. PIERRE

[F117]

The audio signal is generated from the 500microsecond timer interrupt.
There is a hint in the code to use PWM - this results in a 1kHz audio output that will vary in pulse width and so in volume. I never liked this approach, but you can enable it by adding
#define PWM_AUDIO. This will not help you with the vibrator however...


When you look at the schematic: Pin7 of the PIC is a free output pin.
It can be used to drive a vibrator / LED ON/OFF like this:

Line 22:

// I/O pins
#define DAC_OUT PORTA.RA0 // DAC OUT --> opamp DC offset
#define BEEP LATA.LATA2 // OUT audio -- PWM3 RA2 pin 11
#define PI_TX LATC.LATC5 // OUT MD pulse pin 5
>>> Add: #define DETECTION_PIN LATC.LATC3 // Digital detection output

We can now use pin 7 = pin RC3 as DETECTION_PIN like this:

Line 520:
// Limit pulse diff
if (pulsediff > 64)
{
pulsediff = 64;
}
>> Add: if (pulsediff == 64)
{
DETECTION_PIN = 1;
}
else
{
DETECTION_PIN = 0;
}

This will switch the pin high when something is detected and switch the pin low otherwise.

Succes,
- Bernard

[reptooyep]

Hi, thanks for your reply,
I first saw the definition of PWM_OUT but no use of it, that was puzzling me.
I don't want to use pin7 cause i've already etched one TH and two SMD boards.
Isn't it possible just to out a 500ms signal from pin 11 ?
Thanks

[reptooyep]

like :
Comment the toggle pin 11 line
//BEEP = !BEEP;
and :
// Limit pulse diff
if (pulsediff > 64)
{
pulsediff = 64;
}
if (pulsediff=64){

INTCON.GIE = 0; // Disable interrupts
BEEP=1; //output 11=1
Delay_us(500000);
BEEP=0; //output 11=0
INTCON.GIE = 1; //enable Interrupt

}

[F117]

The code is basically OK, but the interrupts should NEVER be disabled here.
When you stop the interrupts for 500ms, the pulse to the coil - and the flyback pulses will also stop for 500 ms. Normally, they occur every 2 ms.
Because the entire analog front-end is AC coupled, the average DC balance will be disturbed by this and you will have to recover from this each time the interrupts are stopped.

Better is to add a counter and set the BEEP pin active for a number of timer interrupts ( x 500 microseconds).
Personally, I would just do something like:

if (pulsediff=64){
BEEP=1; //output 11=1
} else {
BEEP=0; //output 11=0
}

This will give a continuous signal as long as something is detected ( especially useful when the pinpoint button is active ).
No idea why you want a temporary signal - that could last even longer than the detection of a small target when you swing over it ?

[reptooyep]

You're absolutely right.
500ms is too long and your solution is much more simple.
And most of the time, the simpliest is the best.
Thanks again, i'll share pictures and results

[F117]

Typo correction:

if (pulsediff==64){
BEEP=1; //output 11=1
} else {
BEEP=0; //output 11=0
}

[F117]

Or a little more obfuscated:
BEEP=(pulsediff==64) ? 1:0;

[morganton]

Pics of my completed Pickini

Here are some pics of my completed Pickini 4.











Pics were posted a little smaller than I anticipated. You can use Ctrl - + to enlarge them.

I think I construct my detectors similar to the way most folks do it, but here are some construction details if anyone is interested:

1. All body construction is from PVC plumbing fittings. The lower shaft is stiffened with a dowel that I sand to fit inside the 1/2" PVC shaft. All fasteners are nylon, except for the brass screws that connect the battery case.
2. The coil housing is made of two (2) 10" styrene ceiling fan medallions glued together. I like the center cut-out that makes marking a find very easy.
3. The coil bracket is made from 1/8" acrylic strips laminated together.
4. The arm support and bottom support is a 4" PVC coupler cut in half and shaped.
5. The coil cable is a 3 conductor computer power cable.
6. The hand grip is tennis racket handle wrap.
7. The detector housing is spray painted with Rust-Oleum Ultra Cover for plastic.
8. The arm support material is from a mouse pad.

As soon as it stops raining here in Norcal, i'll do a video of the detector in action.

Thanks again to Bernard and all of the contributors.

[WM6]

Good work morganton.

[F117]

@ morganton:
Thanks for the pictures !
A very nice and professional looking build

[morganton]

Thank you for the kind comments.

Bernard, I hope that this doesn't seem like a silly question (I'm not an engineer), but do you think there is any way to modify your source code to give an indication of probable detection of coins? Say, for example a ding-ding as opposed to a normal ding? I'm not advocating discriminating iron, etc., just maybe a better guess as to coins.