Having problem in RFSignal generation

[MVK]

Hello,

I want test one Transceiver device. There are different variety of test colud be done. for that case i have to generate an RF signal. The signal specification is as follows.

Carrier frequency = above 113MHz to 399.975MHz (with this range i will use any frequency)

Message frequency = 1KHz

Power level = -101dBm

Modulation depth = 30% (i mean modulation index is 0.3)

Then this signal should be given to UUT (unit under test... i mean Transceiver). UUT will do demodulation and it will output audio signal (i mean message signal). Then i have to measure the power of that audio signal. This is the procedure. Actually i could achieved this test using stand alone instrument like R & S..etc. same thing i am trying to do using NI modular instrument. For that case i am using PXI 5670 to generate RF signal. But when i generate signal with above mentioned specification and feed into UUT, no output from UUT device. I mean no audio signal output from UUT. But when I generated RF signal with above mentioned specification using standalone instrument (R & S) and then I feed into UUT, now audio signal coming out from UUT. Then I cross check my RF signal generation. Whatever I have generated using RFSG with same specification , given into standalone, then I analyzed… every thing is fine in my generation except strength of the signal. I mean I am generating –101dBm right, but when I measured the value is around –105dBm. But for that amplitude also UUT will give audio output. But my case nothing coming. Then I increased power level of my RFSG from-101 to –95dBm…now I am getting audio output from UUT. But that audio output power also not correct. So I dam sure problem not in the UUT side. It is lying some where in NI RFSG side only. In the software I have to do some correct settings.. say some thing like gain setting. But as such upto my knowledge I am doing every thing correct in the LabVIEW coding side. Please give me some idea to rectify this problem. See I am generating very low signal I mean –101dBm(almost 2 microvolts) for this case should I do any thing special type of setting in my software? Say for example in the RFSA side if I want to measure low power let say –101dBm signal means, I have to give reference level around –110 or less than that. Then only I can able to measure that low level signal. Like that any other special settings should I do in the RFSG side? Please let me know more.

Note: For your info i have attached vi which i have sued for RF signal generation.

[Andy_Hinde]

This could be a dynamic range/noise floor situation. At -101 dBm output power, the PXI-567x will have a noise floor of around -160 dBm. This is 14 dB higher than the ideal noise floor of a 50 ohm source of -174 dBm. The added noise could be the problem. The solution other customers have used has been to generate a stronger signal and use an external attenuator to bring both the signal level and the noise down, allowing the noise to reach a -174 dBm level. I recommend trying this first. Generate a -81 dBm signal and use an external 20 dB attenuator, which will bring the signal down to -101 dBm and the noise down to -174 dBm.

Regards,
Andy Hinde
National Instruments

[MVK]

Hello Andy Hinde,

I have tried as per your suggesstion. But still my problem not solved. I have generated -71 dBm signal then externaly attenuated by 30 dB attenuator. then i fed into UUT(unit under test) but no response. any other way to bring down noise floor?... or any other solution for my problem.?

[MVK]

Hi andy,

sorry for wrong reply. as per your suggesstion i tried. first time i have generated -81 dBm signal then attenuated by 20dB so -101dBm signal given to UUT. but no response from UUT. then i have changed attenuation as well as generation. -71dBm generated attenuated by 30dB attenuator so -101dBm signal feeded into UUt now it's responding. Thank you for your help. but i want to know the exact reason why it so...?

our PXI 5670 will able to generate even -110 dBm signal. but how the stength of the signal getting reduced ? where is the flaw?

Thank you.

vijay

[Andy_Hinde]

Hello MVK,
This isn't a flaw in the PXI-567x, simply a design difference from traditional box instruments. Traditional box signal generators have a lower noise floor than the PXI-567x when generating low power signals, which gives them greater dynamic range or SNR. The PXI-567x noise floor bottoms out around -160 dBm/Hz, which is about 14 dB higher than the theoretical limit of -174 dBm/Hz which the boxes are probably reaching. This is merely a tradeoff for working within the PXI form factor and has a readily available workaround as you have seen yourself. If you generate a higher power signal with the 567x, your dynamic range will increase, and when an external attenuator is placed at the output, the signal and noise are both reduced, in effect letting you achieve a similar -174 dBm/Hz noise density.

Assuming you are using some DUT with an automatic gain control system (AGC) like a radio receiver of some type, the DUT will be receiving the same AM signal from both sig gens, but with the 567x, the noise power will be larger for the same signal power, so the AGC will use less gain and the noise could be overpowering the UUT. With the box sig gen, the noise is lower so the AGC uses its gain to amplify more of the signal.

Please see the PXI-567x Noise Floor specs and look at how the noise floor varies in comparison to signal output:
http://www.ni.com/pdf/manuals/371355b.pdf

You can also verify this by acquiring some power spectrums of (a) the box sig gen output, making sure the noise floor of the spectrum analyzer is around the limit of -174 dBm if possible, and then (b) doing the same with the 567x with and without the external attenuator.

P.S. I have been assuming you are using Double Sideband (DSB) AM modulation for this, although it is immaterial for the behavior in question in this thread.

Message Edited by Andy Hinde on 07-18-2005 12:04 PM