Mixer Level/Reference Level

[cochenob]

I feel like I'm starting to become a regular here on the RF discussion board.  I'm not sure if that's a good thing or not...  🙂 

I'm making some measurements using the 5660...and the results seem unnecessarily noisy.  I'm concerned that perhaps I'm not setting the "Reference Level" correctly.  Could somebody clear up the definition of the mixer/reference levels?  In "traditional" mixer notation, there's the RF (input) port, the LO (reference) port, and the IF (output) port.  How do the 566x's define this in relation to the mixer/reference level?  All of the help documents just list how the attenuation is set based on the relative levels...but it's still a bit confusing as to what I'm setting when I input a level for the "Reference Level" field, and furthermore, what the optimum level should be based on a particular application.

---
Brandon

[Yardov]

Hi Brandon,

Is always good to fully understand what is going on in this field of RF (and yet it is hard to tell sometimes).

The objective is this. Whatever your signal amplitude or power level is, it better be satisfying the needs of the mixer. That is where, the mixer level comes into picture. Our goal is to bring our signal level up or down so that it meets the effective signal level.
Look at the following drawing and I hope your questions are cleared with this.



Hope this helps,

Message Edited by Yardov on 05-16-2007 12:29 AM

[Andy_Hinde]

Hi Brandon,
You should set the Reference Level to the maximum expected total integrated power you expect on the input to the PXI-566x RF VSA. For example, if you have a modulated signal with an integrated power of 20 dBm across the bandwidth of the signal, you should set the Reference Level to 20 dBm.

Regards,
Andy Hinde
National Instruments

[cochenob]

Thanks Andy.  I was afraid of this.  I'm measuring signals orders of magnitude lower than your example.  Likely less (sometimes much less) than -30dBm.  At what point do I need to worry about there not being enough drive power for the mixers?  Can I go all the way down to the noise floor and trust the measurement, or would I need to amplify externally?

[Andy_Hinde]

Whenever the Reference Level is set <= -30 dBm, all attenuation in the PXI-5660 is turned off. This includes RF and IF attenuation. This is desirable since attenuation raises the noise floor.

At this point, the noise floor of your power spectrum will eb determined by your Resolution Bandwidth (RBW). Starting with the noise density specs of the PXI-5660, which is -135 dBm/Hz for frequencies < 1 GHz, you add the amount of RF attenuation being applied as well as 10*Log (RBW) to get the final noise floor. For example, if you had 10 dB RF attenuation and were using a 10 kHz RBW, you would add 10 + 10*Log(10,000) = 50 dB to -135 to get a noise floor of -85 dBm. If your signal has enough SNR above this noise floor (6-10 dB is a good rule of thumb) then you should have no problem.

In your case for low level signals, if you have no attenuation and acquire a power spectrum with a RBW of 1 kHz for example, you should get a noise floor of -105 dBm.

Regards,
Andy Hinde
National Instruments

[cochenob]

Good!