Dual-channel receivers

Updated June 28, 2008

Click here to go to our main page on receivers

Page is under construction!

New for July 2008! Often in microwave systems there is a need for two or more receive channels. This could be because dual polarization is required, or in a monopulse radar, when sum and azimuth or elevation are required on at the same time (often the AZ/EL channels can be switched into a single receive chain).

The image below of a four-channel receiver was used by permission from Liam Devlin, who also contributed some of the discussion on this page. Thanks!

Designing for high isolation

One requirement for dual-channel receivers is channel-to-channel isolation. This isolation is limited by the antenna isolation for a dual-pol system; typically the "system guy" will try to specify the receiver isolation to be 10 dB better than the antenna, for example, 35 dB id the antenna provides 25 dB. If the system guy want 40 dB isolation between receiver channels, tell him (her) to go take a hike!

One "sneak path" associated with dual receiver channels is the LO signal path because a mixer can leak its RF signal out the LO port. The two LOs must come from the same source, by way of a splitter. If a Wilkinson splitter is used, the isolation of the two branches is determined by your ability to control the mismatch on the common port. The way to fix this is to add LO amplifiers on each LO path, or isolators, either way the leakage out of either mixer will be blocked! (We'll soon add a picture to help you understand this...)

In order to save space, dual-path MMICs are often employed (limiters, LNAs, phase shifters, etc..), but these are the culprits in designing for high isolation. Here's a Microwaves101 rule of thumb:

Generally, channel isolation is limited to around 30 dB for dual channel MMICs. For a full receive chain, 25 dB isolation is a good goal.

Measuring isolation

The general procedure is to terminate input of B and output of B and measure gain through channel A as a reference. The input to B while terminating input A, take the "gain" (the cross-channel transmission) data and SUBTRACT the second measurement from the first measurement. (so the "isolation" is positive in dB).

To help visualize how this works, imagine each channel had a gain of 50dB. Assume the channel to channel isolation is 30 dB. We then measure from input B to output A (input A and output B terminated in 50 ohms).

If we consider the case where all of the coupling (which limits the isolation) takes place at the input. The measured transmission (from input B to output A) is thus made up of -30 dB isolation plus 50 dB gain resulting in 20 dB measured transmission. To get your 30 dB isolation you take the 50dB gain and subtract the 20 dB cross channel transmission. (If you do it the other way around you get -30 dB).