Click here to go to our main page on microwave terminations
Click here to learn some ideas for designing broadband mismatched terminations
Click here to learn about VSWR
Click here to learn about phase stretchers
New for October 2020. Mismatched terminations are designed to reflect a near-exact amount of power back to a source. They are typically used in First Article and Group B testing of transmitters and power amplifiers (vacuum tubes and SSPAs). The design must meet a defined failure rate at full power into the specified mismatch, "at all phases".
To design a useful mismatch load, it must be of suitable power handling, and the overall design should include a phase stretcher. Whatever the loss of the phase structure is, it should be taken into account. If you need a 3:1 mismatch, you may have to use a 5:1 load, or even start with a short circuit and pad it down with attenuation. Keep in mind you will likely need a mismatched load with high power handling.
It goes without saying that one way to design a mismatched load is to use a resistive termination that differs from the system impedance. By looking at the ratio of the termination resistive value to the characteristic impedance you can calculate the VSWR in your head. 150 ohms, or 16.7 ohms (50/3), will result in 3:1 VSWR. Below is a simple Microwave Office schematic of a mismatched load that presents ~4.5:1 VSWR.
And here is that termination's VSWR over frequency, courtesy of Microwave Office. Very flat, but in practice flat frequency response is the devil in the details.
If you want to test a transmitter or amplifier and monitor it for failure during the test, you need a coupler and a way to vary the phase. If you have a big budget, buy a line stretcher. Be sure it can handle required power level and make sure it has low loss. There are many ways to make a line stretcher, here are a few examples. You will need at least 180 degrees of phase, whuch results in full 360-degree rotation because the signal travels throught the stretcher twice because it is reflected. In the schematic below we threw in a 1 dB attenuator, to account for the coupler and phase shifter loss.
Now you see why we picked 11.13 ohms for the mismatch. The mismatch load test system presents 3:1 to the device under test, which is a typical requirement.
Here is a plot where we phase-shifted around the Smith chart using the line stretcher. It can hit any phase but in practice you might only test eight or 16 points. At each point you will let the system soak up the reflection for an agreed-upon time, maybe a few minutes. At the end of the test you will need to verify that the tested unit has not permanently changed in any way. Start by checking how much DC current it draws...
Here's the coupled port. It does not behave perfectly because on port of the coupler does not see 50 ohms. You can calibrate out the small error over frequency or ignore it.
Much more to come, and soon!