Is it correct method to find the differental port impedance?

[yuhohang]

Dear All
I want to find out the differental port impedance by the following method, please anyone give me the comment.
1. Use the battery to power on the circuit
2. Isolate the ground bewteen of the circuit and the network analyzer
3. Connect the "Port 1" of the network analyzer the either one of the differental port.
4. The other one differental port connect to the network analyzer ground of the "Port 1".

Is it the S11 could be equal to the differental port impedance?

 

[flatulent]

it all depends

This should work as long as you do not overload the circuit you are measuring with the network analyzer signal. Try a measurement and see if it is what you expect. Then lower the analyzer signal level with the options on the analyzer and see if the reading is the same.

 

[toonafishy]

Port 1 of network analyzer to one end of diff circuit. Port 2 of network analyzer to other end of diff circuit. Take the S-Parameters. Put this S-parameter block in your favorite simulator. If it is @/D/S, then terminate the block with a 50 ohm termination with both ends ungrounded. Then you can read the differential S-parameters return loss. If your simulator only has grounded terminations (M/W/O), then use an ideal 1:1 transformer between the S-parameter block to the termination in order to do the diff to single ended conversion.
If you don't have a simulator, you can also calculate the diff return loss using the single ended S parameters taken above.
See R/F D/e/s/i/g/n M/a/g/a/z/i/n/e Jan 1999 page 24.

 

[yingyang]

differential port

What if you have four ports and how would you measured the differential port using the VNA???

:roll: :roll: :roll:

 

[cray]

>Port 1 of network analyzer to one end of diff circuit. Port 2 of network analyzer to other end of diff circuit. Take the S-Parameters. Put this S-parameter block in your favorite simulator. If it is @/D/S, then terminate the block with a 50 ohm termination with both ends ungrounded. Then you can read the differential S-parameters return loss.

if the termination port was ungrounded, how does ads defined it's port characteristic impedance? in smith chart what is our Zo?
maybe we can use 1 termination port and defined the termination port Zc=100 ohm and connect one end to port 1 of the s-parameter block and another end(terminator) to port 2 of the block. that should gives differential S-parameter return loss.

anyone know how to find a common S-parameter return loss?

can i see the following articles?
See R/F D/e/s/i/g/n M/a/g/a/z/i/n/e Jan 1999 page 24.

 

[toonafishy]

I don't see the R/F/ design issue on their web site, it only goes back to 2000. I'll scan it in and post it later.

In @DS, if you connect a Z=x ohm termination, the resulting smith chart is normalized to x ohms. It doesn't matter if the termination is grounded or not.

 

[jwei363]

two way

1. use 1:1 transform or BALUN to measure
2. Agilent new VNA has that feature to test diff network

 

[Rayengine]

The topic is quite interesting to me. I have read @gilent application notes on use balun before but how can we measure it on circuit with PCB size like GSM board. I think balun case is only valid on isolated component test. Have someone the RF design describe in previous mail, I also very interest on it!

Or Give us some more idea on this topic!

 

[toonafishy]

I don't like the balun idea. It just introduces another uncertainty in the measurement. Sometimes I will use a precision 180 degree hybrid for this purpose but baluns are not very good.