Parasitic AM during BPSK ?

[tigr]

hi

I'm trying BPSK modulator example from FPGA RF Communications Library 2.1 and the output waveform is like this

Here carrier freq = 25MHz and bitrate is approx 3Mbps

It seems there is a parasitic AM here and I cannot understand its origin

By the way, the same phenomenon occurs in "ni5640R Analog Input and Output" from PXI-5641 example and looks like following

There is visible phase flip-flop on both waveforms, but as far as I know, AM is prohibited in pure PSK

 thanx in advance

[Elizabeth_K.]

Hi tigr,

One thing I'm not sure your considering if the fact that the IF-RIO takes a baseband signal and modulates it onto an IF frequency. The "Parasitic AM" that you are witnessing is the carrier signal. This is expected. What are you expecting to see?

Regards,

Elizabeth K.

[dsangree]

To clarify, the 25MHz oscillation you see is the carrier tone. The BPSK data is mixed with the carrier.

[tigr]

of course, 25MHz carrier tone is evident. I reffer to "parasitic AM" not to carrier, but to envelope (having approx. 300 nsec duration). I expect to have something like this

[Elizabeth_K.]

Hi tigr,

I apologize, what I meant to say is that the envelope you are witnessing is the message, not the carrier. Your message is going at 3 Mbps = 333 ns per bit, which is what you are seeing. I have attached a simulation VI that shows the behavior you are seeing, which is expected.

The graph that you provided is what the message signal looks like, but it gets modulated onto a carrier frequency (in this case 25 MHz). You should not expect to see the message signal equal the modulated signal.

Regards,

Elizabeth

[JerryAL · Accepted solution]

Hi Tigr

I’m assuming that you are using the ni5641R DBPSK Modulator.lvproj example.

If you look at the FPGA code, there are two things to point out.

1. The data after being modulated is filtered with a pulse shaping filter.  This is to prevent the abrupt 180 phase changes in the carrier signal that you have in your image above in your previous post.  Transitions as such are not very tolerated in the real world, unwanted side effects will usually occur.  The pulse shaping filter is affecting the amplitude of the I part of the modulated signal by minimizing the amplitude of the transition point.

2. The example only uses the I component.  The Q component is zero’d out.  Only using the I component of a complex IQ signal is also a direct way of implementing AM modulation on a carrier when using IQ modulated data to digitally upconvert to an IF carrier.  Whatever amplitude is in the I component of the IQ signal will show up as an envelope on the carrier.

Your example image is a perfect example of a BPSK signal without pulse shaping that is required in a real transmission link.

Jerry

[ChenYong]

Hi jerry ,

   I 'm looking for an example about 5641R with BPSK modulate.  Could  you upload the " ni5641R DBPSK Modulator.lvproj " example here?   Thanks very much!

[Elizabeth_K.]

The example that Jerry is talking about can be found here - NI Community: Differential M-PSK Demodulator. Note that you will need the LabVIEW FPGA RF Communications Library 3.0 to run the example.

[ChenYong]

I'm using 5641R to make a DBPSK modulator. But  I only find the example using 7965R to make a modulator with a RFSG.

If you could help to find this example " ni5641R DBPSK Modulator.lvproj "  it will be great.