The Mixer 10 Commandments

By Marki Microwave, Posted Thu Apr 29 2010 19:00:00 GMT+0000 (Coordinated Universal Time)

I’m the first person to admit that mixers are confusing. Even drawing a mixer schematic can be challenging—what with all the crossing over of lines and the 4 and sometimes 8 diodes configurations. For better or worse, the complexity of mixers means (a) most companies don’t want to design mixers, and (b) Marki Microwave’s customers sometimes need a lot of coaching and advice. To put things in perspective, there are literally hundreds of texts relating to RF and microwave amplifier design, and about 3 relating to mixer design. It is not your fault you are confused!

To help you, I have come up with a list of “The Mixer 10 Commandments”. If you can follow these simple rules, I promise they’ll help to make your life a little easier when it comes to using mixers.

  1.    Thou shalt not starve the mixer of LO drive.
  2.    Thou shalt not blame conversion loss ripple on VSWR problems when reflective filtering is present at the IF and/or RF ports.
  3.    Thou shall carefully follow the recommended solder reflow temperature profile when mounting surface mount mixers.
  4.    Thou shalt not measure phase noise using mixers made using GaAs devices (FETs, diodes, or otherwise). Silicon schottky diodes are preferred for phase noise measurements.
  5.    Thou shall phase lock your synthesizers to a reference oscillator when making mixer measurements.
  6.    Thou shall test mixer performance in a broadband 50 Ohm system.
  7.    Thou shall place the LO driver amplifier as close to the mixer as possible when laying out a PCB.
  8.    Thou shalt not provide attenuation on the LO port unless it is absolutely necessary.
  9.    Thou shall assume mixer simulations are always wrong, unless corroborated with measured data. This is especially true when simulating mixer nonlinear performance (i.e. single tone and multi-tone IMD). (This is subject to change as nonlinear modeling matures in the coming years).
  10. Thou shalt not starve the mixer of LO drive.

Here’s a hint, #1 and #10 are by far the most important. Too often engineers believe they are saving board space, power and money by under-driving the mixer, only to find later in development that they have unintentionally caused catastrophic penalties in terms of conversion loss, isolation and IMD. If you add up all the time and money wasted in order to correct this relatively straightforward mistake, you quickly find that the safest, fastest, most economical approach is to drive the mixer at the recommended LO level. Think of it this way: you wouldn’t expect an amplifier to work properly if you don’t bias it to the manufacturer’s recommended DC levels, so why should you expect the mixer to “turn on” with too little LO drive?

To learn more about how mixers work and how to use them, check out ourMixer Tutorial.

Related Application Notes

Advanced Calibration Techniques for IQ Mixers: Enhancing Performance Across Variable Conditions

By Doug Jorgesen

Published on: Sat Apr 27 2024 00:26:00 GMT+0000 (Coordinated Universal Time)

Designing a Compact, Filtered mm-Wave Multiplier with IADA-2050CH

By Doug Jorgesen

Published on: Thu Aug 07 2025 01:00:00 GMT+0000 (Coordinated Universal Time)

Comment

Andrew, Fri Mar 15 2024 16:14:47 GMT+0000 (Coordinated Universal Time)

Looking for the description of the different drive mixers. Just need SSM = diode A, HSM = diode B, .... Basic information on your mixer series is not easy to find.

Humberto Sarinana, Sun Apr 20 2025 14:48:41 GMT+0000 (Coordinated Universal Time)

Do you guys have more info on the following "commandments"? Thou shall place the LO driver amplifier as close to the mixer as possible when laying out a PCB. Thou shalt not provide attenuation on the LO port unless it is absolutely necessary. As a CCA designer, these are extremely intriguing. I have always been told to include a attenuator between the LO port and the LO amp to create a better match.

Doug, Mon Apr 21 2025 16:20:52 GMT+0000 (Coordinated Universal Time)

Humberto - Placing an attenuator between the LO amplifier and the mixer is a very common approach, but we don't generally recommend it. The return loss on the LO port (and the RF port for that matter) is usually not great, less than 10 dB and as bad as 5 dB for a typical mixer. The reflections should be resolved inside the LO amplifier due to the reverse isolation, however, and they should not propagate backwards through the system. As long as the distance between the LO buffer amplifier and the mixer is as short as possible, there should be minimal standing waves in the band of operation. By having as much LO power as possible the linearity of the mixer will be maximized. The caveat to this is that the intermodulation suppression (both single tone spurious suppression and two tone intermodulation) of the mixer is dependent on the terminations of the mixer in unpredictable ways. Sometimes the linearity is better with an attenuator on the LO port of the mixer due to the termination of the particular nonlinear products, so the best approach is to experimentally test whether an attenuator improves the performance of any specific nonlinear tone when possible. If it isn't possible then put the LO amplifier as close to the mixer as possible and it should work well.

LEAVE A COMMENT

<< Back to Application Notes