cascaded iip3
of calculation if one block preceded a filter. Because a filter can filter out
the interference.
The bigger issue is if the filter attenuates one or both of the two tones used in the testing. If so, the apparent IP3 of the system will be higher
Filter attenuation can affect the IIP3 and the insert loss also has influence.
Thanks for your reply, another question is may I release the linearity requirement
of the stage following filters? And, how much I can release?
If the filter rejects a frequency, and if you are looking for the IIP3 of the following stage at that frequency, you can subtract from the gain of the stage the rejection of the filter. Lower Gain means higher IIP3.
Thanks,
By the way, we often see the IP3 specification of a filter is divided into in-band
IP3 and out-band IP3, why need this? Why we need out-band IP3?
Thanks,
Out-of-Band IIP3 is necessarily to determine the filter rejection in your system analysis.
An example using ADS you can find here starting with page 28:
http://eesof.tm.agilent.com/pdf/blue_trans.pdf
The out-of-band rejection is sometime more important than the in-band esp. for the filter because in communication system we usually have to deal with very large out-of-band signals that can translate into in-band interferences. Therefore, all the stages before filters should have a good out-of-band properties.
I have another question, once if we use integrated active filter, how to deal
with it? Because the linearity is limited and not like passive filter. How do we
characterize the IP3 of the filter itself?
For high-order state variable filter e.g. leapfrog, the typical way to optimize the linearity is doing internal node voltage scaling. This prevents any internal node to have too large voltage swing than the other hence no any node has been clipped too fast. Please refer to the textbook in active filter e.g. one written by Schuamann, in dynamic range optimization section.
any filter is powerless,it has nothing affect on cascaded ip3 output
cascaded 相关文章:
