NEW !!! Full differential PI front end.

[moodz]

how simple can it be ?

Here is the vga circuit ...you can assign any pins as video interface.

Spartan 3e 3a ... whatever ... it will still work !

moodz.

[moodz]

FPGA update

As a sidebar I have posted a surface mount process in
http://www.geotech1.com/forums/showt...218#post113218

Works for me anyway ..

On the FPGA side of things I have now taken delivery of the rotary encoders, 2 x 16 LCD and MOSFET drivers so watch this space !

One thing I did not have in the design is the audio channel. Of course I could have written a routine to generate audio ( tone etc ) using the CPU however there is still heaps of space in the FPGA so I have added a 3 channel synthesiser and noise source module which will be modulated by the CPU but will not involve any messy waveform synthesis by the CPU. Even better the volume control is also fully digital and the audio output is PWM ( so it uses only one output pin on the FPGA ) A simple resistor and cap form a low pass filter and the audio resolution is 16 bit. In this way very complex sounds can be generated to indicate all sorts of things at the same time.




Some of you are probably wondering when I am actually going to get around to demonstrating an actual metal detector .... please be patient ... there are about 30 'chips' ( modules ) in this design and each one has to be designed or modified then added to the total design then the whole thing retested to make sure that I did not break anything. However it is getting there. So now I have added the sound module I have basically got everything I need functionwise. I now need to code the user menus and fiddle the basic ground balance. Then I have to build a basic front end .. put the whole thing in a box and rod and get out in dirt. ( My bench testing days are over ... that is why I put the VGA scope interface in this design )

Hope this explains more ....

Moodz.

[doncaffe]

Hi moodz !

I am very interested with what kind of waveform processing you plan to do ?
This prototype is multi frequency or single frequency system ?
Spectral analysis was done in multi periods of waveforms or not ?
If you plan extreme accuracy of measurements you need some scrupulous selection of data tapering windows and appropriate averaging.
Tell me more ok!

Reg. Don

[moodz]

Hi moodz !

I am very interested with what kind of waveform processing you plan to do ?
This prototype is multi frequency or single frequency system ?
Spectral analysis was done in multi periods of waveforms or not ?
If you plan extreme accuracy of measurements you need some scrupulous selection of data tapering windows and appropriate averaging.
Tell me more ok!

Reg. Don

Hi Don .... I did consider doing a complex domain analysis and I certainly have the code blocks to do it however the conventional DSP techniques are overly obsessed with obtaining results in utilising these techniques IMHO. Mainly because I am focussing on a PI detector all the waveform processing is done in the time domain using the techniques utilised in the conventional analogue processing circuitry. So I am just implementing the same devices like sample and hold, averaging, gain, adders and subtractors etc in the digital domain ... the technique avoids noise and drift.
Some of the old analogue ways have been forgotten and AFAIK are no longer taught in engineering.

This is a definition of the technique ... ( this site is excellent resource BTW )
http://cnyack.homestead.com/files/aconv/convfil1.htm

I will be making a VLF / IB version probably in the same chip and that will be using the Frequency domain processing. ( multi & single ).

Also in one of my past lives I worked with wavelets so I will be coding one of those just for the hell of it ... if I have time.

Regards,

Moodz.

[doncaffe]

Hi Moodz

I will preformulate a question to be much explicite.
Is this a multi period sampling PI or not ?

For quite complete understanding of pulse waveforms in PI we need viewing the pulse in terms of its frequency component. Coz, we have to use Fourier analysis.

Mathematical apparatus is well known in the prior art of DSP.
If you have developed quite a new mathematical expressions let me know.

And another question is:
Do you considering bipolar or unipolar transmit system and multi period rectangular waveforms ?.

Given link is good for students but for engineers there is excellent book
published by CRC Press LLC in cooperation with IEEE press.

Google it.
Key words "Digital Signal Processing Handbook","Vijay K.Madisetti and Douglas B.Williams"

Reg. Don

[moodz]

Hi Moodz

I will preformulate a question to be much explicite.
Is this a multi period sampling PI or not ?
Yes this code version is.

For quite complete understanding of pulse waveforms in PI we need viewing the pulse in terms of its frequency component. Coz, we have to use Fourier analysis.
I do not agree .... the frequency domain can describe a signal in phase frequency and amplitude but not time location ...this was one of the reasons wavelet transforms were developed. Also the TD and FD are interchangeable ...proven by operator transforms equivalence.... eg a multiplication in the time domain is the same as addition in the frequency domain and vice versa .. similiar for division / add / subtract. When you understand this you can process stuff in time domain or frequency domain ..just depends what is easiest and most appropriate.

Mathematical apparatus is well known in the prior art of DSP.
If you have developed quite a new mathematical expressions let me know.
I have not developed new techniques ...just using old analogue techniques but in digital domain. You have heard of analogue computer ... same thing .. old music synthesisers use same techniques.

And another question is:
Do you considering bipolar or unipolar transmit system and multi period rectangular waveforms ?.

All of the above the pulse generator will generate what you program it and connect appropriate cct.

Given link is good for students but for engineers there is excellent book
published by CRC Press LLC in cooperation with IEEE press.

Not all the readers are up to speed on this so I publish appropriate links.

Google it.
Key words "Digital Signal Processing Handbook","Vijay K.Madisetti and Douglas B.Williams"

Reg. Don

answers above ...regards. M.

[Tinkerer]

Hi Moodz

I will preformulate a question to be much explicite.
Is this a multi period sampling PI or not ?

For quite complete understanding of pulse waveforms in PI we need viewing the pulse in terms of its frequency component. Coz, we have to use Fourier analysis.

Mathematical apparatus is well known in the prior art of DSP.
If you have developed quite a new mathematical expressions let me know.

And another question is:
Do you considering bipolar or unipolar transmit system and multi period rectangular waveforms ?.

Given link is good for students but for engineers there is excellent book
published by CRC Press LLC in cooperation with IEEE press.

Google it.
Key words "Digital Signal Processing Handbook","Vijay K.Madisetti and Douglas B.Williams"

Reg. Don

Don,

I think the posts below might be relevant to your questions:

http://www.geotech1.com/forums/showthread.php?t=16793

http://www.geotech1.com/forums/showp...0&postcount=17

http://www.geotech1.com/forums/showp...5&postcount=20

Tinkerer

[simonbaker]

Thanks for the detailed explaination. It seems reasonable, the difference between motivating the current with an applied voltage versus a magnetic field. I'll have to believe Spice and reality as I'll probably never work the equations to my satisfaction.

I think the disconnect I still have trouble with is: if the second coil suddenly has the same current as the first coil, you would expect twice the magnetic field compared to the first coil alone -- yet that I don't believe happens. I guess I'll just gnaw on it all for a while.

Cheers!

-SB

Hi Moodz:

(still gnawing on it...)

You mentioned a spice simulation helped show the currents in the two bifilar coils during MOSFET shutoff.

I'm still befuddled by it, but I made this little LTSpice circuit to try to help. It seems to show current in the energized coil decreasing while the current in the other coil increases, then they meet and both decay together. I have no idea if this is realistic or even represents your circuit.

Can you comment?

Regards,

-SB

P.S. Nice stuff Moodz, Aziz, Tinkerer... I think

[simonbaker]

gnawing

Continued from prev post... More interesting graph showing sum of currents in coils too. Derivitive of sum gives induced current in target???? (of course this is without MOSFET breakdown).

-SB

[moodz]

Hi Moodz:

(still gnawing on it...)

You mentioned a spice simulation helped show the currents in the two bifilar coils during MOSFET shutoff.

I'm still befuddled by it, but I made this little LTSpice circuit to try to help. It seems to show current in the energized coil decreasing while the current in the other coil increases, then they meet and both decay together. I have no idea if this is realistic or even represents your circuit.

Can you comment?

Regards,

-SB

P.S. Nice stuff Moodz, Aziz, Tinkerer... I think

Thanks Simon ...your simulation matches experimental results. The coils are coupled quite tightly together .... any target within range is also coupled. So energy is transferred to target during flux build up and released on field collapse. The point of collapse is a discontinuity in the signal ( this is why I feel spectral analysis is inappropriate ) however you have pretty much nailed it in your next post regarding differential analysis ..( which is also the key to descrimination IMHO).

Regards,

Paul.

[doncaffe]

The Geotech forum definitely need good mathematic equation editor !!!

[Qiaozhi]

The Geotech forum definitely need good mathematic equation editor !!!

Try using the latex editor ->
http://www.geotech1.com/forums/showt...ighlight=latex

[moodz]

Well thanks Don .... I do appreciate your reply ...actually I am not sure yet what will eventually deliver the descriminating PI detector .... Maybe it will be FFT ... if there was a recipe to do this I would just build one and save my time and money.

However there is no project of that will suit as test bed to test all possible detector theory. ... This is a problem ...everyone talks about theories on this board ... but there is no general purpose test bed.

The nearest we have to a test bed is LTspice and maybe breadboard and Oscilloscope.

All I am trying to do with the FPGA is create a container with all the tools you need to test out ideas that can be expressed in mathematics ( or DSP if you like ).
So I have the following things in the chip.

[Aziz]

Hi Don,

I am a bit confused by your naming convention:
square wave and rectangular wave

What do you mean by that?

Aziz

[moodz]

Well thanks Don .... I do appreciate your reply ...actually I am not sure yet what will eventually deliver the descriminating PI detector .... Maybe it will be FFT ... if there was a recipe to do this I would just build one and save my time and money.

However there is no project of that will suit as test bed to test all possible detector theory. ... This is a problem ...everyone talks about theories on this forum ... but there is no general purpose test bed.

The nearest we have to a test bed is LTspice and maybe breadboard and Oscilloscope.

All I am trying to do with the FPGA is create a container with all the tools you need to test out ideas that can be expressed in mathematics ( or DSP if you like ).
So I have the following things in the chip.

and the following things are ( because I accidently pressed submit too early ... he he )

1. One or more pulse generators that can be numerically controlled ( duty cycle and frequency ). Obviously need this to control the MOSFET switching. However my initial design will be very similiar to an older minelab.

2. One or more ADC drivers to get analogue data into the system.

3. One or more DAC drivers to generate analogue signals either for feed back to amplifier front end, coil balancing, meters .... whatever.

3. One or more rotary knob controller so you have user controls.

4. A sound synthesiser for user alerts and maybe music during lunch breaks.

5. one or more LCD driver so you can display stuff . ( 2 x 16 char currently )

6. Multichannel VGA digital oscilloscope so you can see signals not only what the ADC captures but also results of DSP computations. Also very useful for checking coil performance etc. This single feature alone is the most useful to have as you can see if your design works or not ... in the field.

7. VGA also displays text console so you dump lots of info about things that wont fit on LCD.

8. serial port so you can connect to computer. On the larger Xilinx Board there is ethernet and more RAM / Flash also.

9. A 32 bit MIPS compatible RISC processor on the FPGA for user code. ( the fast DSP is done in FPGA though ... more efficient) This does the menus, reads the rotary knob values, controls the displays and maybe does the high level descrimination.

10. DSP code blocks that can switched into the signal chain ...

11. All of the above is coded inside the FPGA, tested and works and I have only used up 59% of the resources of a cheap ( $20 ) FPGA.

moodz.