Open source Microcontrolled PI

Thanks for the ideas I am going to use a stm32f4 as I want to get a bit fancy with the gui also it is where all my experience is.
The board size should be attainable as my current design is moving towords just using the stm32f4 discovery and have the md piggy back like a shield.


Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,60153.msg295069.html#msg295069

I recommend you start out with a 22khz signal pulsed at 20ms and work the frequency up from there. Don't go above 300khz as years of experiments by engineers smarter than us have determined that eddy currents on the surface of metal don't linger when you go above.

Don't worry about using the microcontroller yet to decode your response. Instead, develop a Smith chart to see what happens to your Tx field when you put metal in its field. Once you see that, add a receiver coil and do it all over again. If you use the same scale for both Tx and Rx fields, you can resize and overlap them in your computer to see what you need to do for eliminating or detecting specific targets.

This is going to be a long project and will require you to do some local experimenting. Find a non metallic work surface and non metallic tools for holding your objects so you don't skew your results. Don't worry about soil conditions at this point, just detection patterns for select targets. I use a large cardboard box with plywood on the top for a table and plastic rulers for the objects. Elsewhere on this forum I have given the basic plans for a field detector that you can use to develop your Smith charts.

Google Smith chart for thousands of ways to draft them. Any coil and your microphone input on your computer can be used to sense the fields.

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,60153.msg295074.html#msg295074

There are a few things to remember the decay time of the coil is one of the most important factors, it also happens to be the main factor in a PI design (in my opinion) you also have the sample time and the delay  time to consider. If you try and make a detector that can go from 100Hz to 1Khz it will not work but on the other hand if you try and make one that will go from 100Hz to 200Hz you stand a damn good chance of it working. One thing that has be said on this forum many a time is build your detector a round your coil not the other way round. If you had a pulse width of 50us you would need a very long off time because if you had a 50/50 on/off time you would end up with a 10Khz detector and this will not work, your on time would be too short, a good on time to start with is 250us. I'm not an expert in this domain, what I'm telling you is what I have experienced so keep doing your researching although you'll probably end up doing your own thing in the end any way (there's loads of contradiction on the net)   

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,60153.msg295078.html#msg295078

Xavier

Well you've got enough info to get started now so all I can do is wish you luck and remember to show us your progress.

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,60153.msg295087.html#msg295087

Xavier

Yes I have made an idx pro and a couple of very simple pi circuits I have worked with electronics for years mainly in the security and automation fields with a focus on custom one off circuits and I have been programming as a hobby for some time now.
This will be my first real pi.

Linkback:

You are not allowed to view links.
Please Register or Login

http://www.thunting.com/smf/index.php/topic,60153.msg295118.html#msg295118