Granted, the answer is to read the datasheet. Maybe Prahlad had read it but was not understanding how it works, thus the question. I think in the circumstances that maybe Jez just simply shouldn't have responded - when I read the first post my response was the same as Jez's!

Anyway, back to the topic.

No, I haven't had any direct exposure to the AD637 myself, but reading the datasheet and app notes explains the operation of the device and more importantly, tells you about the averaging capacitor - this sets the response time of the measurement. If Prahlad wants to do a cycle by cycle measurement of RMS, then this may not be the best solution and his current method may be more applicable. At about $10USD per device, that buys a lot of micro, so Prahlad's current solution may be the most economical.

There may be some means of using a sample and hold with the AD637 to give a cycle by cycle solution but I expect the usual application is to average the value over a number of cycles.