Calculating power in dbm using NI USRP 2920

[MHantera]

Hello every body , could any body please provide me with VI that helps me in calculating the power in dbm for a certain bandwidth using NI USRP2920 and labview 2012 

Thanks 

[Leah-L]

Hi MHantera,

This looks like the same question you posted on this forum post. Did you try the examples that I suggested to you? I still think that the Power Spectrum Measurement example in the Example Finder does what you want. 

[Leah-L]

MHantera,

I believe there are also a couple examples that are available to you if you have the Spectrum Measurement Toolkit or the Sound and Vibration Toolkit. Do you have either of these toolkits?

[MHantera]

Unfortunately I have tried both examples and they didn't help me 

[MHantera]

And I'm also afraid that I make something wrong so could you provide me with example in order to check what's wrong and also unfortunately I doesn't have spectral measurement tool kit 

[cochenob]

Some tips from past experience....having wandered down this road before....

- The "Power Spectrum" VI used in many of the example programs returns the magnitude spectrum in Vrms^2 or dB.  WARNING...it's "dB"...not "dBm".  The next question might be, "dB relative to what?".  If I remember right...it's relative to 1Vrms into 1ohm...based on my digging down into the Power Spectrum VI.  Not really a particularly useful measure, but ok if you're just looking for relative measurements I suppose.  In any case...if you set the Power Spectrum input to return the spectrum in Vrms^2 (it's a boolean input to the VI), then you can convert Vrms^2 back to volts...then to power (into 50 ohms)...then to dBm (referenced to 1mW).

ANOTHER WARNING....the dBm that you calculate will not be the same as what you put in (say...if you're using a well calibrated RF synth).  This is because the driver doesn't calibrate for losses in the analog front end.  If you want to know exactly what the input power is based on your measurement in LabView...then you'll need to calibrate the radio over your band of interest with a known source.

---

Brandon

[MHantera]

Thanks a lot for this important tips, but could you support me with the modified VI that you have made. and thanks alot again for your reply  

[cochenob]

I didn't modify any of the standard VI's so to speak.  It sounds like you're unsure how to convert between units.  This may help...

dBm = 10*log10( Power/1mW )

...or...

dBm = 10*log10( (Vrms^2/50)/0.001) )

[MHantera]

Thansk for your effort, and i made what you described in the above steps but also it gives me false results

[cochenob]

It looks like you're multiplying the power in mW by dt....why?

Also...be aware that because you're summing...you're basically integrating up the total power within the USRP BW....which could be different than the power of the signal, depending on the nature of your signal and it's power level.  This may or may not be an issue for you.  If you're just looking to make an energy detector, than this is probably ok.  If you're trying to actually calibrate the radio somehow, you may need to be careful. 

For example....if you're input is a single tone....and you look at the power spectrum, then you'd say that you signal power is the level of the impulse peak in your spectrum.  If you calculate the power by summing over the entire band....you'll get a similar answer so long as your signal level was well above the noise floor (i.e. - high SNR).  At low SNR, where you signal and noise are comparable, you'd be integrating up both the signal and the noise floor, so you may get a slightly different answer between the "look at the peak" method and the "integrate everything up" method.

That said...I've used the "integrate everything up" method with a single tone input, and have been able to map out the entire 80-90dB dynamic range of the radio with sufficient accuracy.  Remember though...the radio isn't calibrated...so you'll be off a few dB from the input.

---

Brandon