Hi,
I am using the two tone test to measure the 3rd order IP of a passive double balanced mixer.
Conditions as follows:
25KHz separating F1 and F2. Input power to mixer = -15dBm (each tone)
Mixer conversion loss = 7 dB
INPUT 3rd order intercept point = 25dBm
FLO = 66MHz, 136MHz, 530MHz and 940MHz, 17dBm at the LO port.
There is a IF of 350 KHz for each LO frequency.
Equipment E4403B ESA-L SA
3 x 8642B signal generators
From theory I expect an IMR of 80dB, in practice I am measuring an IMR of 71 dB
I have noticed that the IMP's are sensitive to IF O/P impedance at 2x FLO, e.g. If I Null out the 2X FLO product (At the IF output) with a notch filter the IMP's reduce by approx 9dB.
I believe the IF port mismatch at 2XFLO is causing reflection of the 2XFLO product and further intermodulation in the mixer.
I have built a IF diplexer to dump the energy above 1MHz (IF passband) into a 50 ohm load. The circuit has a S11 of 40dB at 2xFLO, however this does not reduce the IMP level as expected.
I assume the mixer does not want to see a 50 ohm impedance at the IF output at 2xFLO.
I am also assuming that the IF port impedance at 2X FLO for each of the above LO frequencies will be different!
How can I trap the 2XFLO product at each LO frequency whilst maintaining 50ohm match in the passband and elsewhere ?
Thanks in advance for any help.
I am using the two tone test to measure the 3rd order IP of a passive double balanced mixer.
Conditions as follows:
25KHz separating F1 and F2. Input power to mixer = -15dBm (each tone)
Mixer conversion loss = 7 dB
INPUT 3rd order intercept point = 25dBm
FLO = 66MHz, 136MHz, 530MHz and 940MHz, 17dBm at the LO port.
There is a IF of 350 KHz for each LO frequency.
Equipment E4403B ESA-L SA
3 x 8642B signal generators
From theory I expect an IMR of 80dB, in practice I am measuring an IMR of 71 dB
I have noticed that the IMP's are sensitive to IF O/P impedance at 2x FLO, e.g. If I Null out the 2X FLO product (At the IF output) with a notch filter the IMP's reduce by approx 9dB.
I believe the IF port mismatch at 2XFLO is causing reflection of the 2XFLO product and further intermodulation in the mixer.
I have built a IF diplexer to dump the energy above 1MHz (IF passband) into a 50 ohm load. The circuit has a S11 of 40dB at 2xFLO, however this does not reduce the IMP level as expected.
I assume the mixer does not want to see a 50 ohm impedance at the IF output at 2xFLO.
I am also assuming that the IF port impedance at 2X FLO for each of the above LO frequencies will be different!
How can I trap the 2XFLO product at each LO frequency whilst maintaining 50ohm match in the passband and elsewhere ?
Thanks in advance for any help.
Since the E4403B spectrum analyzer also contains mixers, which are non-linear devices for the downconversion process, we measure and specify 3rd order intermodulation. The specification is < -75 dBc for two -30 dBm signals at the input mixer and >50 kHz separation. This equates to an intercept of +7.5 dBm. I don’t think you’re going to see -80 dBc performance on your device when the input tones are -15 dBm at the first mixer. Additionally, you have close signal separation, which could be problematic when trying to filter.
Just in case it is needed, you can find the procedure and test methodology for testing spectrum analyzer SHI and TOI performance on the web at www.agilent.com/find/esa_manuals and look for the Calibration Guide link.
Regards -