RFSA FSK Demodulation: Output bits stream received without emission

[Guetteur]

Hi All,

I'm trying to receive an ouput bit stream, by demodulating an RF signal.

The signal is a fsk modulated 272 bits frame.

The emission frequency is 433.92 MHz. The flow rate transmission is 227.27 us.

The frequency deviation is between 30 KHz.

The retreive the bits emmited, i'm using: MT RFSA FSK Demod With Power in Band Measurement.vi

The FSK IQ Spectrum is correct. Not the output bit Stream.

The main big problem is that i'm receiving an output bit stream even with no emission.

Below, three pics for more information.

Any help will be great.

Thanks in advance

P.S: I'm using PXI 5600/5620 with an antenna

[Yardov]

Hi Guetteur,

The MT VIs are always trying to demodulate whatever you tell them to. Every time you perform an acquisition there could be emission (very low power) so the VI is just doing what you told him to do.

Looks like you have a frame duration (227.27 us??) meaning that you should have a "burst" there. The PXI-5660 (5600 + 5620) does not have IQ power trigger (like the 5661 and 5663 does) and therefore you will have to implement that in software. You will have to program a step that detects the burst (what you call emissions) and make sure that you pass to the FSK demodulator only the burst and nothing else.

To make your program more robust, you will also need to wire the input "synchronization parameters" to the FSK demodulator. You probably have a preamble in those bits that define the beginning of the burst or valid data. This is very important to get the best results out of the demodulator.

The rest of your program I believe is OK.

Hope this helps,

[Guetteur]

Hi,

Thank you very much Yardov for your answer.

What you wrote, is what i understood when i combined two VIs, one to detect the first peak of the modulated signal, and the second to begin to acquisition, once the first peak detected. But, even with this solution, i wasn't able to "catch" all the 272 bits of the initial frame.

I'll try to wire the input "synchronization parameters" to the FSK demodulator as you suggested and reply you back again.

Thanks again for your answer !

[Yardov]

Hi Guetteur,

For the burst detection you can reuse some of the code in the shipping examples:

MT ni5660 MSK Demod (burst detection).vi

This example has a VI to detect the burst in software. I would make sure you did something similar to detect the burst and then parse only the FSK section you need.

Regards,

[Guetteur]

Hi Yardov,

Again thanks for your answer. It has been very useful. I actually did what you suggested. In other words, i've used "MT ni5660 MSK Demod (burst detection).vi"  in order to "catch" my signal.

I was able to receive the emmitted bits (every single one), when it's the 5610/5441 that are performing the emission. But, when i tried to catch the same bits, but emitted by an industrial emmitter (not the 5610/54410) i wasn't able to retreive the original bit stream.

P.S: The only difference between to the emitted signal is the power level. 

Any idea why this is happening ?

Very big thanks in advance.

[Guetteur]

Hi again,

I doubt that the problem comes from the fact that the filters used in the two transmitters are not the same. So logically, If i don't change the filter type of the demodulation depending on which filter is used on the emission, the results won't be the same.

So, my question is: is there an exemple that shows a demodulation of an fsk signal which independant on the filter used in the emission (by performing a level detection for example) ?

Regards,

[Yardov]

Hi Guetteu,

"P.S: The only difference between to the emitted signal is the power level. "

I think this is where the problem is since this is very critical to a demodulator.

1) Make sure your burst is detected correctly. Make a graph of power versus time to make sure you "see" it.

2) If that is correct, then you can simply FM demodulate the data just to make sure you see the analog message signal which should be +Frequency Deviation and -Frequency deviation values only (since is FSK). For example, look at this spectrogram of FSK, you can see how the frequency changes over time:

If you have the generator working with our instruments, then you need to make sure your configuration to capture the DUT signal gives you similar results.

I am not sure what filter you are talking about (instrument or demodulator) but visualizing the waveform will help you debug this. You can also post some of your code or your IQ raw data here.

[Guetteur]

Hi Yardov,

Thank you again for your answer.

Actually i'm receiving the signal from my receiver (please see the picture 1.gif). But i wasn't able to retrevie the bit stream. You can pick a look to my code (picture 2.gif).

I was able to retreive every signle bit, when the transmission is performed by 5610/5441, but not in the other case.

Just to mention that the power emitted by the 5610/5441 is 10 times higher...

The values used are: Bandwidth:0.1MHz, burst duration 63 ms, trigger level 1.5 mV, FSK deviation: 15KHz

Is there anything wrong in my code ?

Big thanks in advance !

[Yardov]

Hi Guetteur,

Your signal power seems very low, I would even think is burried in the noise floor.What is the PXI-5660 reference level you are using? This should be about -30 dBm for this signal. If you already have it like that, do you have a PXI-5690 power amplifier so you can amplify the input signal?

The goal is to amplify the input power of the signal.

[Guetteur]

Hi Yardov,

Thank your for the quick answer.

Actually, my reference level is 0 dBm.

I also think that the signal is buried in the noise floor. To answer tyour question, i don't have the 5690 card.

Is there any way to filter or to minimize the noise floor ? ( like a digital signal processing technique handled by a vi) ?

Thanks in advance !