Arduino Nano PI Main Discussion

[godigit1]

I am one that has experimented with digitizing the decay curve. Used the front end of my HammerHead2 and a PIC32 processor. I did an exponential curve fit to obtain a TC. This worked fairly well for well defined non-ferrous targets like coins but odd shaped targets and iron gave odd results, TC could be anything so really did not discriminate ferrous/nonferrous.
I never wrote this up and posted on Geotech but did write up some of the PIC32 code methods used to generate signals.
Once I get some time I would like to revisit this with the VooDoo circuit that has ferrous discrimination.

Hi Waltr

Im thinking this is still very useful even if were not getting fully accurate disc or even iron rejection. (even a VLF does Not) I think there could be data value there translated into the field.

With any detector Ive used after about a 1000 targets or so you start to really notice the nuances of the detector.

With Pi tied to a numeric VDI scaled off the TC it would take several weeks in the field but patterns may start to emerge and you could disc targets by TC (for small nuggets on shallow bedrock you could only look at 1 to 3 Tc targets and ignore the rest if you knew there to by no big nuggs in the area.

As George has stated these projects seem to go by the way side.

Maybe we are expecting too much as well, when all the time we have a new and innovative tool at our disposal.

I wish I had the mad skills that some of you all have I could temp this on my own (maybe someday) until then I have to wait to get that tool in my hands.

Thanks for all you contribute to all of us .
Be well

[waltr]

Maybe. I haven't investigated this closely yet.

Several people have tried digitizing the decay curve in the vain hope of using the data for target discrimination. Usually they start off very confidently, then get into an argument with someone in the forum and disappear into oblivion to consider patenting the idea, and making vast amounts of money from commercialization. Spoiler ... of course, none of this ever happens.

I am one that has experimented with digitizing the decay curve. Used the front end of my HammerHead2 and a PIC32 processor. I did an exponential curve fit to obtain a TC. This worked fairly well for well defined non-ferrous targets like coins but odd shaped targets and iron gave odd results, TC could be anything so really did not discriminate ferrous/nonferrous.
I never wrote this up and posted on Geotech but did write up some of the PIC32 code methods used to generate signals.
Once I get some time I would like to revisit this with the VooDoo circuit that has ferrous discrimination.

[Qiaozhi]

True, but not sure if thats at the flat-out speed though ? Will look further.
Do you know if anyone has ever tried digitising and then analysing the curve with a DSP ?

Maybe. I haven't investigated this closely yet.

Several people have tried digitizing the decay curve in the vain hope of using the data for target discrimination. Usually they start off very confidently, then get into an argument with someone in the forum and disappear into oblivion to consider patenting the idea, and making vast amounts of money from commercialization. Spoiler ... of course, none of this ever happens.

[gbrowne]

True, but not sure if thats at the flat-out speed though ? Will look further.
Do you know if anyone has ever tried digitising and then analysing the curve with a DSP ?

[Qiaozhi]

Hi George, my book arrived earlier from the USA than expected, good job on the book nice clear schematics etc.

Thanks.

Just a question about this pico thing, are you planning to use it to sample just one point on the delay curve out of the preamp ?

I haven't done anything else with the RPi Pico yet, as I've been side-tracked onto something else. The Pico is immensely more complicated than the Nano, and may not actually be a good candidate for a Geotech project, but I'll get back to it at a later date.

[6666]

If we supply the 4066 with a single 5 volt supply
then it opens and closes normally,
but the analog signal that is managed and passes through
it, this must also be from 0 to 5v. thats why i reversed the
input signal.

Understood, thanks.

[6666]

The RPi Pico's ADC has 12-bit resolution. At least that's what it says in the datasheet.

Just a question about this pico thing, are you planning to use it to sample just one point on the delay curve out of the preamp ?

[6666]

If anyone is having problems ordering a copy of the book from Amazon in their particular country, the Book Depository can supply copies with free delivery worldwide.

https://www.bookdepository.com/Arduino-Nano-Pulse-Induction-Metal-Detector-Project-George-Overton-Overton/9798719427911

I will also add this information to post #1.

Hi George, my book arrived earlier from the USA than expected, good job on the book nice clear schematics etc.

[Qiaozhi]

If anyone is having problems ordering a copy of the book from Amazon in their particular country, the Book Depository can supply copies with free delivery worldwide.

https://www.bookdepository.com/Arduino-Nano-Pulse-Induction-Metal-Detector-Project-George-Overton-Overton/9798719427911

I will also add this information to post #1.

[Qiaozhi]

Interesting point George. As you say, the pi pico is nearly 10x faster and has ADC that can reputedly return 10-bit ADC at up to 500k/s at a push. Here's an example of a pi pico doing FFT...

https://www.hackster.io/AlexWulff/adc-sampling-and-fft-on-raspberry-pi-pico-f883dd

I'm much stronger on software/firmware than I am on hardware, so I'd be interested to see if I can digitise the interesting bits of the flyback signal and then process it in software. Strongly suspect that this sort of processing might be better done with the help of a dedicated DSP chip ? Also, it's probably unlikely that a pi pico has the grunt to keep up.

The RPi Pico's ADC has 12-bit resolution. At least that's what it says in the datasheet.

[gbrowne]

Yes, I wrote the Arduino PI sketch.

Obviously the Pico is a more powerful device than the Nano, with a lot more memory, so I might change my mind later as I get more familiar with it.

Interesting point George. As you say, the pi pico is nearly 10x faster and has ADC that can reputedly return 10-bit ADC at up to 500k/s at a push. Here's an example of a pi pico doing FFT...

https://www.hackster.io/AlexWulff/adc-sampling-and-fft-on-raspberry-pi-pico-f883dd

I'm much stronger on software/firmware than I am on hardware, so I'd be interested to see if I can digitise the interesting bits of the flyback signal and then process it in software. Strongly suspect that this sort of processing might be better done with the help of a dedicated DSP chip ? Also, it's probably unlikely that a pi pico has the grunt to keep up.

[Xtrem]

Why does the sound go away when the object gets close to the coil

the sound dont go away, but we can not hear it because the high limit frequency of the VCO is
too high.

[Xtrem]

can you draw it ?, do the 4066's turn on completely ?

If we supply the 4066 with a single 5 volt supply
then it opens and closes normally,
but the analog signal that is managed and passes through
it, this must also be from 0 to 5v. thats why i reversed the
input signal.

[Matt_Rowe]

Yes

https://www.youtube.com/watch?v=7M5RpMLhRrs

plus changes on the sound.

Why does the sound go away when the object gets close to the coil

[6666]

The drivers for the 4066 are not necessary, if you handle only the positive part of the signal (0 to +5V)

can you draw it ?, do the 4066's turn on completely ?