Re: [Discuss-gnuradio] sending single tone Am modulated using usrp1+wbx
Ok, I understand your example, but before work with complex examples I want to experiment with a simple circuit as a simple DSB AM modulator based on the theory that I wrote in the past mail, Im aware about realistic situations like noise and other factors and it was the reason I connected directly a wbx output to wbx input (different wbx) in the same usrp trying to avoid this realistic situations. Now I know that is necessary an attenuator (in this moment I dont have an attenuator) and I use two antennas, but I need to know if the usrp mathematically works like Im believing . ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] sending single tone Am modulated using usrp1+wbx
On 08 Jan 2013 08:48, José María Valencia wrote: Ok, I understand your example, but before work with complex examples I want to experiment with a simple circuit as a simple DSB AM modulator based on the theory that I wrote in the past mail, I'm aware about realistic situations like noise and other factors and it was the reason I connected directly a wbx output to wbx input (different wbx) in the same usrp trying to avoid this realistic situations. Now I know that is necessary an attenuator (in this moment I don't have an attenuator) and I use two antennas, but I need to know if the usrp mathematically works like I'm believing…. I previously observed that direct connection between TX and RX has a *high* probability of damaging the LNA in the RX. Further, a direct connection will *not* remove all the noise. Analog electronics don't work that way. There is *always* noise in an electronic circuit at ordinary temperatures. 1/f noise, shot noise, various quantum effects. The only reason we don't have to think about that too much in the *digital* domain is that we quantize, so that there are voltage zones that represent 0s and 1s, and we choose those zones such that the various circuit noise features never (or almost never, statistically) cause us to make a wrong decision with respect to 1 and 0. In the analog domain, the noise matters and is always present. ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] sending single tone Am modulated using usrp1+wbx
Ok, Im aware about the damage as a result of direct connection, But now the question is How to recovery the m(t) signal in spite of noise? That is, finally the I and Q in rx should be equal to I and Q in tx. ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] sending single tone Am modulated using usrp1+wbx
On 08 Jan 2013 13:49, José María Valencia wrote: Ok, I'm aware about the damage as a result of direct connection, But now the question is How to recovery the m(t) signal in spite of noise? That is, finally the I and Q in rx should be equal to I and Q in tx. The values of I/Q at the receiver will, hopefully, be linearly-proportional to the values as transmitted. When you use radio systems to convey *information*, you use those signals to *encode* that information, using modulation schemes that only require that the signals be linearly-proportional (and sometimes not even that) to the signals as sent. That's how radio works. You modulate information onto a carrier, and demodulate that carrier to recover the original (or a reasonable facsimile of the original) information. The Universe conspires to mangle electromagnetic waves, to a greater or lesser extent, depending on a number of factors including wavelength, distance, interference, etc, etc, etc. If you rely on the received signal to be *identical* to the transmitted signal, then you are living in a conceptual state of sin, and must revise your thinking. ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] sending single tone Am modulated using usrp1+wbx
Ok, I will try with an attenuator, but I suppose I can work too(while I get an attenuator) with two antennas setting the proper gain for each one, but now I have the next question: Independently if I work with attenuator or antennas, why the gain affects the spectrum and shape of the received signal? Thanks. ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] sending single tone Am modulated using usrp1+wbx
Ok, I will try with an attenuator, but Isuppose I can work too(while I getan attenuator)with two antennassetting the proper gain for each one, but now I have the nextquestion: Independently if Iwork with attenuator or antennas,whythe gain affects the spectrumand shape of the received signal? Thanks. Without seeing the details, I can't comment in detail. But as you increase the RX gain, more of the weaker components of the signal start to rise above the noise floor of your configuration, so you can see them. As you increase gain further, depending on the strength of the TX signal, you end up with the amplifier operating in a non-linear fashion, thus distorting the signal. -- Marcus Leech Principal Investigator Shirleys Bay Radio Astronomy Consortium http://www.sbrac.org ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] sending single tone Am modulated using usrp1+wbx
Marcus thanks for your help. I was trying with 2 antennas (WA5VJB) varying the usrp tx gain but the results are the same (In this moment I dont have an attenuator). I was reading a past mail from this list (March 25 2012) and there is a similar problem that I have: the q component is not zero, I send a mail to the list asking for it . thanks ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] sending single tone Am modulated using usrp1+wbx
I will write my theory about usrp working and some math in Am modulation, because I want to be sure before continuing: If I want to modulate in DSB amplitude a carrier of saying 1GHz ,with a m(t) signal (in this case a simple 1kHz cosine tone) the carrier is sourcing by USRP module, and I have to set up the freq of usrp in 1GHz and based on Leon Couch Book the complex envelope for DSB AM should be I=Ac(1+m(t)) Q=0. (In my case Ac=1). This is equivalent to (1+m(t))(e^0) eq.1 I and Q are a complex signal that I connect to the usrp sink. Internally the usrp multiplies eq.1 by e^(2*pi*1GHz*t) resulting (1+m(t))e^(2*pi*1GHz*t) eq.2 Eq2. travels by the air to the receiver. In the USRP receiver, Eq. 2 is multiplied by e^-(2*pi*1GHz*t) resulting (1+m(t))(e^0) whose I and Q components are 1+m(t) and 0 respectively , and I can obtain the magnitude with a complex to magnitude block, this magnitude is 1+m(t). This my theory and I want to be sure if this is correct . Thanks. ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] sending single tone Am modulated using usrp1+wbx
On 01/04/2013 08:24 PM, José María Valencia wrote: Hi, I want to transmitand receivea1kHz cosinetoneusing dsb am modulation and usrp1 with2wbx dbs,I have connectedfromtx/rx output wbx-ato rx2 inputwbx-b with a 50ohmsma cable, I'mexpectingin scope sinkthe cosine tone with 1KHz butI have another signal, can you tell me what could be the error? I'm based on the complex envelope theory: I=Ac(1+m(t))Q=0 You can see the diagram and the signals onthe attached image. Thanks. ... ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio So, did you use an attenuator, or did you connect directly? I suggest that you put in a 30dB attenuator in-line, and carefully adjust both the TX and RX gain until you see the spectral component you need. Connecting a transmitter directly to a receiver (whether it's a USRP or any other piece of radio equipment) is generally a bad idea -- there's a *strong* risk of damaging the front-end of the receiver. More so when the receiver (such as with WBX) has a low-noise front-end. So, when you're doing laboratory loopback tests, you use an attenuator. The WBX receiver is easily capable of seeing extremely weak narrowband signals, so there's no reason not to put in a fair amount of attenuation and experiment with gain levels on both the RX and TX side. -- Marcus Leech Principal Investigator Shirleys Bay Radio Astronomy Consortium http://www.sbrac.org ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
