Implement Metal detector circuitry in LabVIEW

[Henrik_Volkers]

OK, not much time...

I saw a USB 6212and I assume you have a university license and can install the sound and vibration toolkid

And you really should use the amplifier

Start with the example SVXMPL_Averaged Frequency Response (DAQmx).vi   using a chirp signal with a no window at the FRF express vi.  

Oh signal buried in noise....     try  Lock-In technique see here

Now you have hopefully signals to analyse.

Another technique: Use a pulse as a stimulation signal and add multiple measurements (noise add with sqrt(n), signal with n 🙂 )

Here is a blind try (have no DAQ here) with an example as starting point and added 'adding' , however the synchronizing might not work..... 

[DPARK]

Hello Henrik,

thank you. That program overwhelmes me. 🙂 I can try it, but actually I have another point of view.

Currently, I use the MOSFET as a "switch" and connected a 12V battery to drive the coil and the MOSFET. 

The problem is that I still don't have as much as current I need to drive my coil. (please the the attached png)

1. load is my transmitter coil

2. I also use the flywheel diode

3. I also use the 1k resistor

4. micro should be a digital output of the daq board

I thought I can use a the puls signal from the DIGITAL output of the DAQ board and connect the signal to 

the gate of the MOSFET. Please let us try this first, that would be easier to understand for me.

And so I don't need the amplifier...

Do you think this circuitry will work? I definitely need the puls from the digital output. Can you help me to

get this into LV? Otherwise I can't drive my circuitry to if if it works. 

Just a few facts: The resistance of my transmitter coil is 22,5 Ohm and I drive it with the 12 V battery now.

Thank you!

Simon

[DPARK]

see png below.

I also attached the terminal plan of the NI USB 6212.Which channel should be used for the digital output and do I also need GND for a digital output?

[johnsold]

Simon,

What is your pulse width?  You should be able to get about 0.5 A with the circuit you have shown.  Can your battery provide that much current?

In an earlier post you mentioned an inductance of 6.3 mH.  With a resistance of 22.5 ohms you have a time constant (T = L/R) of 0.28 ms.  If your pulse is less than about 1 ms duration, your current will not rise to the full value.  Henrik was recommending 10 us pulses.  With pulses that short your current will be quite small. The diode, which protects the transistor, also lengthens the pulse!  Calculating its effect is much trickier.

Many options have been discussed.  Please describe completely what you are presently using: Transmitter coil inductance and resistance, pulse duration and repetition rate, estimated or measured current; Receiver coil parameters, analog sensitivity, measured signal and noise, sampling rate; and anything else which you think might be relevant.  This way we can tell where you are and not be guessing.

Lynn

[DPARK]

Lynn,

my problem is I don't know how to create a puls signal in LV. I didn't get response by searching for this issue...

That's what I want to try:

1. Send a ditial puls signal to the DAQ board DO 

2. the signal passes the 1k resistor and drive the gate of my MOSFET

3. I use the MOSFET as a switch and provide power from a external 12 V and 1,3Ah battery. (see the png before) (I hope the current is enough)

The inductance of my transmitter (TX) coil is calculated and maybe not as much accurate. The calculated inductance of the receiver (RX) coil is 6.5mH and the resistance 24.7 Ohm

I can't give you the other parameters because I don't know how to create puls. Sorry if it's so basic for you...

Thank you!

Simon

[johnsold]

The USB-6212 only supports software timed digital I/O. This means that your pulse will be in the millisecond range, not microseconds.  You should be able to use the counters for generating shorter pulses.  Look at the examples which come with LV.

The battery should be sufficient.  As I said earlier your inductance and resistance may preclude the kind of pulsing we have been discussing.

Lynn 

[DPARK]

Hmm okay... But that's the problem with milliseconds? Is then too much current in the circuitry?

I tried to find the emaple, but I only found this:

??? Do I really need so much software to produce the digital signal???

[johnsold]

I cannot say whether millisecond pulses are a problem or not.  It depends on your coils.  How much current can they handle without overheating?  What size wire is used in the windings? What pulse duty cycle do you plan on using?

To produce a pulse with controllable frequency and width requires two counters.  Each need to be configured, then run and possibly reset after each pulse, and shut down after the task is complete.  That is not much software.

Lynn

[DPARK]

I used a 30AWG with resin insulation.

I got another question. What's the difference between "Counter Output" and "Digital Output"? With the Counter output I already got a great puls to the oscilloskope. 🙂

Thank you!

[DPARK]

Okay... some new data.

I created a "counter output"signal. (see .png 01 and .jpg 02) With that properties I got the following values of puls width and frequency (see .jpg 03 yellow channel).

I also measured on the DAQ output: 2,5 V and 0.05 A.

The voltage from start to the end of the transmitter coil is just 5V (I don't know why?!) I can't measure the current in the coil. Otherwise the MOSFET becomes very hot!!! 🙂

The blue channel is the receiver coil, but I can't get result if I move metal about the coils... 😞 😞 😞

Picture 04 and 05 are my "small" circuit board and the connection of the coils. Do you have further advices how to use my circuit? Do I have a mistake in my circuit?

I don't understand that... 😞

Simon