I'm working with Ni5663e with LabVIEW code (example code by NI). While testing the device with an external Frequency Signal Generator I found a small frequency shift in the resulting spectrum. The shift is around 100 Hz with a IQ-Carrier Frequency of around 200 MHz. I checked the value of the IQ Carrier Frequency before it was set to the device . Additionaly I also used the RFSA property node to read this property from the device to double check this value. So far I cannot see a problem here.
While reading for such a thing I found an effect called Carrier Frequency Offset.
I wonder if this shift I found this Carrier Frequency Offset effect and if it is how I can correct for it? Or is maybe the device just broken?
sofar no solution. I beginning to wonder if the device is really just broken .. it cannot be that bad...
The first (and easiest) thing you can try is to connect the "10MHz REF Out" of one device to the "10MHz REF In" of the other with a short coax cable. My educated guess is that it should get rid of your offset.
Your two devices are operating using their own separate internal references, and each device has its own small frequency error. I don't have an NI5663e. I got this spec for the internal frequency reference for a PXIe-5663E VSA from its users manual. At 200 MHz, your 100 Hz offset is 0.5 ppm and is well within spec, so I don't think your device is broken.
thanks for your reply.
The 10MHz Ref is already connected. So this is not the cause for the offset.
About the internal frequency reference. You seem to be right .. but then it is realy not a good device ... We also have a NI6674t module which has a very good clock (ocxo) with a accuracy of 80ppb (http://www.ni.com/pdf/manuals/377485a.pdf, p4). I tried to use it for the 5663E. I routed this clock to the PXIe backplane and set the RFSA set reference clock to PXIe backplane. This reduced this offset to 60 Hz.
60 Hz / 200 MHz = 0.3 ppm = 300ppb. The OCXO clock is referenced with a accuracy of 80 ppb. So I gues there is another way to use the ocxo clock for the 5663e?