I have a requirement to quickly Enable the 5385 and begin modulation. I have been failing EVM tests during the first few hundred uS after I Enable the chip. I have found 2 root causes of my EVM failures.
During turn on my 5V supply gets a step up function 200mA of current added to it. I get a 0.1V droop on the 5V supply at the point of Enabling the 5385. The 5V supply comes halfway back to 5V (0.05V droop) during the transmission and the EVM improves over time. How sensitive is the 5385 IQ accuracy to power supply variation? I assume that the modulator uses internal references to desensitize itself to power supply variations but I do not see any discussion of this in the data sheet.
The next item I am seeing is in regard to the LO input impedance variation as the part goes from Disabled to Enabled. The part is speced for 20 dB return loss but I can only assume that is tested while the part is Enabled. As I Enable the 5385 it pulls my unbuffered VCO so much that my PLL "rings" for a few hundred uS and causes EVM errors at the start of the transmission. I am adding a pad and buffer amp to the VCO input to overcome this but wanted to know if anybody else has been affected by this behavior.