USRP LO leak at carrier frequency

[xadhoom]

Hi
I'm trying to modify wideband spectrum analyzer vi to suit my needs. I'm having a problem with LO leakage. Due to length of my signals I cannot use high sampling rate and because of that I can see many unwatned peaks in my wide spectrum which makes it really hard to read. Can anyone tell me what are the most optimal ways to deal with this problem?

Paweł D.

[cochenob]

You can try to use two radios if coherence isn't an issue...though even if it is, there are ways to deal with that.

Even in this case, you may still end up with a bit of LO leakage, but you may find it more tolerable. 

You could also use two radios, but with one at an IF to the other.  That way your signal of interest is away from the leakage from the Tx side of the Rx radio.  I've done this before with good success, and just digitally BPF the downconvered signal to isolate the transmitted content.

The isolation between Tx and Rx in the same device is around 30dB or so.  There's no way to get rid of it completely.  Like death, taxes, and Cher...they will never go away.

---

Brandon

[torry]

What are the two radios??? How set the LO frequency in order to reduces LO leak???? Or what method you use?

[vvkakade]

Hi,

You can set the LO frequency to a value and then set the carrier frequency.

This will ensure that the LO feedthrough is reduced as you have effectively moved the LO away from the carrier.

For e.g. you want to operate at a carrier frequency of 2.02 GHz.

Set the LO frequency to 2 GHz and then set the carrier to 2.02 GHz. (The offset of 0.02 GHz will get implemented in DSP on the FPGA)

See the attached image to see how to set the LO frequency in case you are using the niUsrp API.