Single tone Tx/Rx and DC offsets

[cochenob]

Attached is some LV code that tries the method suggested above.  I'm using the code on a single 2920 USRP, with the Tx connected into the Rx (with some attenuation).  The Tx program can be set to a certain carrier frequency and generates DC I's and Q's (i.e. - no modulation...just the carrier).  On the Rx side...one can set the same carrier...but with a slightly different LO (within the BW of the sample rate I choose).  According to the suggestions above...the CORDIC should remove the IF caused by the offset LO. 

This seems almost true.  When using the attached VI's...I see a very slowly varying I and Q on the Rx end...suggesting that the CORDIC doesn't remove all of the IF.  Wondering if someone can verify using the same code. 

Doesn't seem to be too dependent on carrier/IF.  I've tried a smattering all over the 0.05-2.2GHz range. 

---

Brandon

[James-K]

Brandon,

I think I'm seeing the same thing here.  Can you provide a screenshot?

[cochenob]

I could...but I'm not sure how helpful it would be.  Maybe I can figure out how to do a movie.

Basically though...if you capture, say, 10,000 pts....I and Q will look constant....but fetch over and over again in a loop, and you'll see those I and Q levels drift very slowly....indicating that there's still some IF hanging around. 

Not sure to what extent the CORDIC can be expected to remove all of the offset.  Less than 1 Hz?  0.1 Hz?  0.01 Hz?

[cochenob]

James-

Just checking in again to see if you guys have any other insight here as to what might be happening.  Also...I'm still curious as to why I can mix down to DC at some frequencies and not others (using the same LO and Cordic).

---

Brandon

[cochenob]

Hi James-

Just checking in again to see if the NI folks have any idea why this method doesn't seem to remove all of the IF.  Both the NI and Ettus message boards have suggested using the same approach to covert a single tone down to DC, but it seems as if no one's actually done it for themselves to notice the slow drift we're both seeing.

Let me know what you think.

---

Brandon