Showing results for 
Search instead for 
Did you mean: 

Share LO

Hi All,


I am using two generators with same LO(Sharing Local Oscillator). When I am generating two signals with the same frequency,everything is fine, But in case of generating two different frequencies in the two generators I am getting wrong values? Can anybody helps me to know why is that happining and how exactly the LO works? thanks in advance


Best regards 

0 Kudos
Message 1 of 5

Hi Al-rawachy,


Can you post what generators you are using, what frequencies you are trying to generate, and perhaps some screenshots of the results? 


What do you mean by "wrong values"? Are the frequencies incorrect? By how much? And what are you using to measure the output of the generators?


Are you sharing reference clocks as well? The Local Oscillator is used for upconverting your signal to RF frequencies, but the reference clock is what would be driving the FPGA onboard and the DACs. 


Finally, what is your end goal with this application? 


Details will hopefully help you get a better answer from the community.

Eric H.
Senior Field Applications Engineer
0 Kudos
Message 2 of 5

HI Eric,


I am using Two generators NI PXIe-5673E with share LO, the first genrator is being used to inject two tone signals from the source side the spacing between them is only I khz which are 1Ghz and 1.000001000 hz with -10 and -25 dbm respectively. at mean time I am injecting two other signals from the load side  1Ghz and 1.000003000 hz with -10 and -25 dbm respectively. I am using thru as a DUT. what I am expecting to get is that the a1 should be similar to b2 and a2 should be similar to b1 becasue it is just thru. but unfortunately I am getting totally different valsue fo both b2 and b1.

I am using 4 receivers NI PXIe-5663E to measure the four S-parameters. and also checking the result by using power meter to check the magnitude. 


Yeah, I am using shareing reference clock. once I get the right result , I will be able to replace the thru with the real device and starting do my measurements.


Thanks in advance





0 Kudos
Message 3 of 5

Hi Al-rawachy,

Can you provide the measurements that you are getting? Does the spectrum look wrong? Are the power levels incorrect? By how much? Can you provide a sketch or diagram of how you have everything connected? I'm guessing you have some couplers for making the reference (a1/a2) and reflected (b1/b2) measurements, can you include all components in the sketch?


It sounds like you are generating from both 5673E's at the same time. Is this correct? From my understanding, that is not how S-parameter testing is usually completed. If you are generating simultaneously, is there a specific reason? Here is a diagram from the Introduction to Network Analyzer Measurements document - VNA_Block_Diagram.PNG

Based on this, only one side should be generating at a time, and the other port should be terminated. Generating from both devices at the same time is double-driving the lines which could overwhelm the 5663E's, depending on the reference level you set. This can also introduce unwanted constructive and destructive interference between the 4 tones.

Eric H.
Senior Field Applications Engineer
0 Kudos
Message 4 of 5

Hi Eric, 


I am using loadpull method , thus, I am injecting the signal from both sides. 

But yesterday I found the problem, it's from my code . I wish I could explain more but it takes too long time to explain it. anyway, the harware is working propely and thank you so much for your support.


Best regards

0 Kudos
Message 5 of 5