How to measure the output impedance of a high power PA ?

[shengwuei]

Hi,

I need to do impedance matching between PA(300W, 10~50MHz) and load.

In my case, neither the impedance of PA nor the load is 50 Ohm, so my plan is to measure the impedance of both of them and do conjugate matching base on my measurement.

The problem is since PA output power is high, I cannot use a VNA to measure PA output impedance directly. I googled some posts and tried to add an attenuator between PA and VNA to do the measurement, though the measured result is always 50Ohm if attenuator is present. If I do VNA calibration with the attenuator , the reading of VNA would be totally wrong after calibration,so I think this is just not a right way to do this.

There are also some posts describing how to use a directional coupler or two different loads to do the measurement, but I cannot fully follow up those procedure because I am not an experience guy in RF area. I need some detailed suggestion, please help, thanks a lot !

 

[biff44]

No, you can not just put an attenuator at the output and measure impedance. The network analyzer does not really measure impedance, but instead measures something called reflecton coefficient. So, if you put a 20 dB attenuator at the output, your analyzer sends out a signal, that signal goes thru the attenuator and gets attenuated by 20 db, hits the amplifier output port and has perhaps a 5 db return loss, goes thru the attenuator AGAIN, and loses another 20 db. So the returning signal to the network analyzer is 20-5-20=-45 dB. A 45 dB return loss will look like exactly 50 ohms every time, no matter what the amplifier is doing!

You could connect the amplifier to the load with two directonal couplers in between. The couplers are installed in opposite directions. Then you put a signal on the input of the amplifier, and you use a power meter on the two coupled ports at the output. There will be a "standing wave" on the line between the load and the amplifier output. The power meter can measure forward travelling power, and reverse travelling power, from the two directional couplers. So you can easily figure out the MAGNITUDE of the mismatch between load and source.

If you actually want to know the source impedance, you need to do a "load pull" measurement. There are various ways, and all are complicated. Do a search.

 

[khouly]

there is an option in some network analyzers called Hot S paramter measurments , it can measure the S22 of the amplifier , but i don't know if this option can handel 200 Watts

Khouly

 

[BigBoss]

In power amplifiers, conjugate matching is not used.That's why you don't need to measure the output impedance of the PA...

 

[toonafishy]

You need to load pull the output so you can come up with a family of matching curves on the Smith Chart. But that is the least of your problems. If you expect to have a broadband match from 10-50MHz, that is over two octaves and it is unlikely you can do this with simple L/C matching techniques. You will have to build transmission line based transformers with the proper impedance ratio to get the power transfer. The techniques are found in the book by Norm Dye and Helge Granberg, "Radio Frequency Transistors: principles and practical applications". He published a lot of app notes.

https://www.ab4oj.com/dl/eb104.pdf