I'm trying to pull IQ data out of an EXA N9010B for post-processing. I've managed to capture the I/Q samples and dump them to a file and analysed its FFT spectra. The signal is a LTE, QPSK in a 20MHz signal generated by Signal Studio.

I used Excel (limited to 4096 FFT points) to display the FFT spectra.

It appears that it is not correct.

1. I was actually expecting a LTE signal in 20MHz bandwidth after applying FFT. On the spectrum analyser, it actually displays a LTE 20MHz signal.

2. The frequency component is wrong - it should be 28GHz.

Can anyone help me where I am going wrong? Is it related to the EXA settings?

The sampling time and number of samples of the EXA are:

Sample time 7.33333E-07

No of Samples 2728

The EXA have not got the VSA enabled to measure EVM. So, I thought capture the IQ data and analyse it in math tool such as Excel to calculate EVM.

Hello –

I do not think that your issues are related to the EXA settings, but more on the principle idea of what you are expecting to view with you IQ data analysis in excel.

It seems like with what you are describing, 28 GHz is your carrier frequency. IQ data is baseband data. Your EXA is returning raw IQ (baseband) data. The method to convert it to a format that will return EVM is a complex procedure that requires more than just using FFT to calculate.

Here’s an example I found that covers a brief description of the complexity of the process using Matlab: https://www.mathworks.com/help/comm/ug/error-vector-magnitude-evm.html

VSA software truly makes this a whole lot easier – you can request a free 30 day trial here: https://www.keysight.com/main/software.jspx?cc=US&lc=eng&ckey=2013472&nid=-32806.0.02&id=2013472&cmpid=zzfind89600_updates

Hope this helps!