hi,
Today, I measure the 2-stream Tx EVM of WLAN(802.11n) by using the Litepoint for 802.11n. The data is HT40 format.
When I have checked option for channel estimation by long training sequence, the evm fluctuation is high and not good(-25~-29, rarely down to -22! It increase the Re-transmission of the packet)
but i checked option for channel estimation by full packet and recalculate the EVM. The value of evm is very good!! (-30~-33). Is it normal?
So, I guess that channel estimation error cause the evm degradation.
I have question at this point.
What makes the channel estimation error by using LTS? (Phase noise error by PLL?/ or not good baseband signal? or other?)
So, if i want reduce the channel estimation error, what can i do? anything i want to do for enhancing the EVM performance, board-level, RF-chip level, Digital-level, Software. (for information, I have checked the CCDF,Chanel flatness,coupling, I/Q mismatch, LO leakage, and so on. They are all ok. and SISO measurement is done, the fluctuation of SISO EVM is smaller than that of the MIMO measurement)
please help me.
Thank you.