The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> ways to measure front-end performance
https://designers-guide.org/forum/YaBB.pl?num=1194331211

Message started by aaron_do on Nov 5^th, 2007, 10:40pm

Title: ways to measure front-end performance
Post by aaron_do on Nov 5^th, 2007, 10:40pm

Hi all,

I'm wondering what different ways there are to measure a RF front-end circuit for a low-IF system (2 MHz). Assume i'm downconverting from 2.4 GHz to 2 MHz, and i have an LNA, a Q-mixer, and a channel select filter with image rejection. The LO is provided off chip. I intend to do on-wafer measurements unless i really need to make the chip+PCB.

oh yeah, i'm interested in NF, voltage gain, and IIP3...

thanks,
Aaron

Title: Re: ways to measure front-end performance
Post by didac on Nov 6^th, 2007, 1:08am

Hi,
Few months ago a did more or less the same measurements that you did. We were able to do all the measurement on-waffer, the materials we used were:
3 RF generators:1 for the LO, and 2 for the RF inputs for measuring IIP3(if you have a signal generator capable of generating two tones is good as well,although maybe you should correct the input power because it splits equally the power to the two tones).
2 baluns:1 for LO and the other for IF(our output was differential).
1 power combiner: necessary to combine the two RF tones for IIP3 measurement if it's not generated by a single source.
1 calibrated noise source for NF measurements
1 spectrum analyzer with NF options(or a separate NF meter if you have access to that).
1 Vector Network Analyzer for measuring the input matching of the front-end.
And nothing else that I can remember right now,
Hope it helps
PD:if you are interested in some general procedures I think I can arrange a document eliminating all "sensitive"(NDA) information.

Title: Re: ways to measure front-end performance
Post by aaron_do on Nov 6^th, 2007, 2:19am

Hi didac,

thanks for the help. That document would be great if it isn't too much trouble but if its troublesome then don't worry about it. Its only for research. I'm roughly aware of the procedures but I don't want to miss anything out since it might mean re-fabricating the design.

thanks,
Aaron