Implement Metal detector circuitry in LabVIEW

[DPARK]

Hello togehter,

Currently, I build up coils for a VLF-metal detector, because I want to determine and discriminate between different metals. My problem is:

I just built up the coil and not the whole detector. Therefore I wanna try to create the circuitry in LabVIEW and support the coils via a DAQ board.

My question is: Do I have the possibility to create a "known electrical circuit" for a metal detector in LabVIEW? 

The basic set-up that I use is:

1. Power a transmitter coil with a signal generator

2. The induced signal from the receiver coil is the input to the metal detector circuitry (composed of: pre-amp, synchronous demodulator, filter, etc...)

I am very delighted to all advices. Thank you!

[F._Schubert]

LabVIEW is no circuit design software but a programming language.

As such you can use it to set the output voltage of your DAQ, aquire signals from the DAQ or talk to the signal generator via RS232. And then do all the cool stuff of processing the data, saving it to files, displaying it to users...

But the whole detector still needs to be soldered.

Felix

[DPARK]

Hmm... okay.. Thank's a lot. So, I've never really worked with LabVIEW but I have access to all the DAQ devices etc. Therefore it would be great to use that.

Actually I supposed that I need the whole circuitry... Do you know a software to create "electrical circuitry"? And isn't it possible to connect the "circuitry programm" with LabVIEW? I just wanna aviod the "hands-on" work and do the stuff as much as possible digital.

Please give me a request if you have an idea for that.! I really apprechiate it.

Thank you!

[F._Schubert]

There is a lot of software available to model an electrical circuit. This depends on the size of the circuity you need to model and on your budget. But you also can just draw it on a paper and get a prototype working. This depends on the scope of your project.

The realization of the circuit is then done by soldering the pieces together. Again depending on the scope of your project you can do this in your garage or spend some money to get professional products.

I will need more information of the kind of project you are doing (hobby, student, professional) and as I never build a metaldetecor myself, I don't know how much hardware you need.

Connecting the circuit designer to LabVIEW is a wrong idea of the system. The electric circuit software is just giving you a model. This allows you or others to create the circuit with "hands-on" work (using a soldering iron, wires and all this). This you connect to the DAQ board.

LabVIEW isn't just giving you a model, but a concrete software that operates. It's on the same abstraction level as the circuit.

If you can detail the complete concept of your model detector more precisely, I can also provide you some ideas which functions you could do in LabVIEW instead of electronics (some function block diagram maybe?).

Felix

[DPARK]

That sounds great! Thank you for that! So, I this is a student project and therefore far away from professional work... 

My problem is: At least I want to determine metal with a (Very low frequency)-detector and I built already up the coils for the project.

I used a signal generator to create a sinodial signal into the transmitter coil. With the inductance principle I get a tiny "induced voltage" in the receiver coil. I connected the received signal to a oscilloscope and compared it with the original signal from the signal generator.

So, I could determine a vary in the amplitude and the phase-shift of the received signal if I move metal about the two coils.

The first problem is: According to some guys in a "metal-detection-board" I have to use the whole metal detector (not only the coils) and so I think my received signal isn't the signal I have to use... 

But let's try with my signal just from the coils. Therefore I have a basic set-up picture for you: (please see the attached picture to this time)

If I can create the pre-amp, the synchronous demodulator and different filters in LabVIEW, then I don't need the whole detector. (That's my opinion  ) At least I wanna do FFT and WAVELET to discriminate between different metals and use a "case study" in LabVIEW. For each detected metal one case.... That's my opinion but at least I am not sure If i have the right signal to do this...

If it is not the right signal, I have to build up the "whole metal detector" by hands-on and transmit then the signal via DAQ... But it must be possible just with the coils....

Please let me know if you have experience with create the blocks for a "pre-amp", "synchronous demodulator"etc... This is the next step after I have the "right signal"... 

Thank you so much and I really need help with that project...

DPARK

[F._Schubert]

Looks cool for a student project. Here just a quick brain-storm:

* Preamp needs to be electronics. You need to get the signal level in the range of the DAQ. So how high is the amplitude you got for now? It might be possible to go without a preamp if the signal is strong enough.

* about your demodulator, you are interested in phase and amplitude when the metal is present; that's why you have the phase shifter there; And it's just AM demodulation? Then this is basically a lock-in amplifier. This can be done by software, search ni.com for lockin as there is some code already written (and wikipedia->external links for some good explanations by manufacturers how they work).

* FFT doesn't make sense in the post-processing, or do you expect the frequency could change. All you have is amplitude and phase.

Felix

[DPARK]

Okay Felix... Very helpful to me! Thank you!

Currently, I got an aswer that I definitely CAN'T use my signal. So I have to make the whole detector. (Just for your information)

But if I have the right signal of the receiver coil, does it make sense to implement this into LabVIEW? I think, I can already determine

and discriminate with the detector...

The received vlotage is very tiny. (10...50 [mV]) I think in this range I definitely need the preamp...

The programm is designated to vary the frequency, but I can't do this because the search coil and also the capacity depends on the

value of frequency. [   f=1/(2*pi*sqrt(L*C))   ]

Sorry I can't answer your question about the AM demodulator. Therefore I need to read your link and get further information...

Okay, no FFT... This is very helpful for me... You're right. All I have is amplitude and frequency... Have you ever heard about the

"Lissajous pattern" ? I think I can also discriminate between metals with this. Do you know if that is possible to implement in LabVIEW?

Thank you!

Simon

[F._Schubert]

Lissajous pattern:

Bingo! It did that when still in school, my project was to measure LRC resonant circuits and plot the amplitude and phase for each frequency. The trick for getting the phase was based on this. The signal was connected to the X-Input of an (analog) oszi and the ref-signal to the Y-Input. Thus I got an ellipsoid. From the crossing of X- and Y-axis I could calculated the phase. I don't remember the correct details, but it's not difficult to calculate it (basic math like sine...).

I think you can implement it easy with an XY-plot. But I wouldn't use it to extract the phase, more for learning and demonstrating the principles of you detector. This was a good way in the old analog-age, but we are now in the digital information age. The demodulator makes use of all the data points you sampled, not just of those that are at the axis crossing -> gives you much better signal-to-noise ratio.

Try to search about AM modulation, demodulation. At the core of it, it's just multiplying the signals, so pretty easy to program.

Felix

[DPARK]

Hi Felix,

I am back with more questions now. Unfortunately, I cannot design the whole metal detector in my short research time and so I have to work with the signal of the coil.

With this signal I cannot distinguish between different metals but I can determine it. So I only have the vary of the amplitude from the signal.

I read a little bit through the isue "amplitude modulation" but to be hornest, I am not sure if i need that for my project.

The problem is:

I want to create a VI which can at least determine metals over the coil with the help of LED or a sound (whatever). Because signal analysis is totally new for me,

I don't know how to work with the received signal and wich blocks (komponents) should be used in the VI.

Currently, I get the received signal from the coil via DAQ. I hope for a few advices or at least a direction how to work.

Thank you so much!

Simon

[johnsold]

Simon,

You commented that "signal analysis is totally new for me..."  You really should learn at least the basics of the kind of analysis you will need to do.  If you do not understand the physical basis of the process you are trying to measure or simulate, you cannot effectively evaluate the performance of your system or debug it.

What do you expect the signal to do when a metal comes near the coil?  Is the behavior different if the metal is ferrous (or magnetic)?  Once you can answer questions like these, then you can begin to look at the software components to implement your VI.

Lynn