WBX configuration

[leonell]

Is it possible to configure the WBX 50-2200 MHz daughterboard of the USRP to transmit AM signals ranging from 535-1705 KHz? If so, how?

[muffin.vi]

leonell,

The WBX daughterboard will not transmit below 50 MHz, so this is likely not the best choice for AM transmission. You could probably use external circuitry to drop the frequency, but I'm guessing you would need to do that before remodulating the signal. 

[leonell]

muffin.vi,

what possible external circuitry could we use to drop the frequency and how are we going to connect it to the USRP?

[User002]

I imagine muffin.vi is refering to a downconverter.  Since your USRP requires you to broadcast your AM station at at least 50 MHz, you would need a downconverter capable of dropping the carrier frequency down to the kHz range.

[leonell]

what if I try upconverting the signal from my signal processing blocks in GNU to reach 50 MHz so that I can transmit using the wbx daugtherboard, is it possible? If so, can I use the multiply and multiply constant blocks to upconvert the signal I am transmitting?

[smoores]

I'm confused.  Are you trying to broadcast an AM signal with a carrier in the kHz range?  If so, that won't be possible with the WBX daughterboard by itself.  If you want to broadcast your AM signal (lets say a 1 kHz sine wave) with a carrier at 50 MHz, that won't be a problem at all.  I'm not too familiar with GNU, but if it works like any other USRP I've worked with all you have to do is modulate your AM waveform, write it to the USRP, then tell it what center frequency you want.  Trying to upconvert and downconvert at baseband becomes an issue of sample rate.  In order to satisfy nyquist, the sample rate must be greater than 2*desired carrier frequency.  So in order to upconvert a baseband signal to a carrier at 50 MHz, your USRP would need to be able to sample at > 100MHz, which it is not capable of.

[AdeelAnwar]

Another option is to use Ettus Research Basic Tx and Rx boards. See
http://forums.ni.com/t5/USRP-Software-Radio/Can-the-NI-USRP-295xR-RIO-use-an-Ettus-BasicRX-front-end...

-Adeel

[leonell]

smoores,

I am making a DSB transmitter-receiver in GNU. The carrier freq of my transmitter is 4000 kHz while the other signal is at 1512 kHz, I multiplied both the signals then use a multiply constant to reach the 50 MHz freq of the daughterboard. I use a frequency counter to test if I transmitted 50 MHz, freq counter shows about 40-49 MHz but no stable frequency was really transmitted. I also used a frequency spectrum to see if I transmitted someting, but it only generated spikes more like noise. Am I doing it correctly? If not, how can I transmit the signal?

[smoores]

leonell,

My experience with the WBX is with the NI-USRP driver.  I'm going to assume that with GNU the concepts are the same.  Also, it sounds like you are just trying to broadcast a tone that has a frequency of 1512 kHz.  In that case, I would modulate the 1512kHz wave into AM DSB.  Here, your I values will be the various points in your waveform array and the Q values will be the desired phase of the carrier.  Of course, you will need to do other things, such as adjust for the modulation index and the peak message amplitude, but that is beside the point.  Hopefully GNU has some built in functionality that will allow you to modulate your signal into AM DSB without actually having to program the modulation much yourself.  After doing that, all you should need to do is write the complex data to the WBX and specify to use a carrier of 50MHz.  The mixer and LO on the WBX should then upconvert your signal to 50 MHz.  If you feed this signal into an RF analyzer, you should see power in the band where your signal is being broadcast (50 MHz).

[leonell]

smoores,

I want to transmit 50 MHz signal via DSB modulation using USRP.  I need at least > 100 MHz sampling frequency but the USRP only supports 8 MHz limit for sampling rate, how can I transmit the said signal? Works fine in GNU simulation but testing if I transmitted 50 MHz using a spectrum analyzer was a fail. I've attached the signal processing blocks I used in GNU.