The procedure is explained in detail, in an application note on
"Mini-Circuits" web site,
at: http://www.minicircuits.com/appnote/mxr1-18.pdf

Regards,

Frank


"Airy R.Bean" wrote in message
...
I do NOT have experience of this, but using one's
background training as an aspiring engineer, I would
use the output of a calibrated signal generator as the input
signal, together with an accurate - and variable - attenuator.

Connect an RF Voltmeter to the output of the mixer.

Now you can plot input power against output power, assuming
that it is within your capabilities to assess the output impedance of the
mixer.

This should give you sufficient points to plot the response curve up
to the 1dB compression point.

However, there is a problem here, and that is, at the point at which
the straight line starts to curve off, you are not just measuring the
output power of the desired output, but also are including that of the
newly-arrived distortion products.

However, as I said, I don't have the answer for you (We are, after all,
amateurs and not professionals), but perhaps I have given you food for
thought?

"Jason" wrote in message
oups.com...
Hello All

May I know if there is anyone here really have experience in
calculating or estimateing IP3 intercept point for any mixer real
case,(cmos or bipolar mixer)?
I need some advice on this topic
Thank you and hear from you all

rgds and thanks
Jason