The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> How to use PORT with mismatched impedance!
https://designers-guide.org/forum/YaBB.pl?num=1146208378

Message started by plasma on Apr 28^th, 2006, 12:12am

Title: How to use PORT with mismatched impedance!
Post by plasma on Apr 28^th, 2006, 12:12am

I have to design a mixer whose Zin=1000 Ohm. It will be connected to a LNA whose Zout<<1000 Ohm, so their impendances are mismatched.

So when I simulate the mixer, I have to use a port with a power (dBm) in it. But the PORT is a strange element, which maybe only could be used for impedance matched system, in order to obtain accurate results.

Is there any good buddy who has the experience to use it in this case, and can help me?

Thanx, Merci, Grazie !
8-)

Title: Re: How to use PORT with mismatched impedance!
Post by ACWWong on Apr 28^th, 2006, 3:02am

Hi,
Because you are using the port in an unmatched system with Rs << Rl, you will get about double the voltage as a result. This is beacuse a port acts a source with an internal source resistance and what you specifiy is the power delivered to the load, when the load = Rs (ie matched).
This is the same way RF signal generators work, if you measure the power of a 50ohm Rs signal generator into a 50ohm Rin spectrum analyser, it will give the correct amaplitude/power level, if you alternatively measure it into a high impedence oscilloscope, you will see twice the voltage.

You can still use the port for characterising you mixer if you take the mismatch into account.

cheers

aw

Title: Re: How to use PORT with mismatched impedance!
Post by plasma on Apr 28^th, 2006, 3:44am

I know the twice effect of port, I think maybe I have to design the LNA and MIXER together. I can't trust the simulation results in the mismatched case:(

Title: Re: How to use PORT with mismatched impedance!
Post by ACWWong on Apr 28^th, 2006, 5:36am

The port can be trusted in the mismatch case, you just have to be careful what you are trying to measure. If you are using spectreRF, your small signal analyses (pac, pxf etc.) may incur the 6dB discrepency due to the mismatch voltage gain. You can always do transient analysis to compare.

Usually I work completely in the voltage domain, and specify my LNA & mixers in terms of voltage gains (& input referred noise, V1dB etc.), this is because I generally work on integrated RFIC, where signals are volatges or currents and not power. In your case I think this would make things clearer for you.

But yes you are correct, you can always do the LNA mixer design together to clarify the matter (assuming you have a conjugated matched input LNA, else you will have to again accommodate mismatch gain/loss).

anyway good luck with your design...

Title: Re: How to use PORT with mismatched impedance!
Post by plasma on Apr 28^th, 2006, 9:35pm

Thanks for reply :D

I think the pac result is the same as the transient's for mesuring the votage conversion gain:)
And I've seen the 6dB effect when I do a PXF analysis.

Maybe I should better use voltage source, there is not a source better than PORT....

Title: Re: How to use PORT with mismatched impedance!
Post by plasma on Apr 28^th, 2006, 11:21pm

If a system is mismatched like the mine, how to evaluate the POWER in dBm

in 50 Ohm or others(1K Ohm for example)?

It is quite troublesome, maybe input power is mesured at 1000 Ohm, and output is at 50 Ohm.

Thanks for discussion~!

Title: Re: How to use PORT with mismatched impedance!
Post by ACWWong on May 2^nd, 2006, 8:11am

If your LNA Zout is say 50 ohm, and the mixer Zin is 1k, then adding the mixer to the LNA output will do little to impact the LNA gain (voltage or power). The power gain calculation in your cadence (eg. S21) simulation will quote you the power into the 1k port and so will be a factor of 10*Log(1k/50)=13dB out, although the voltage gain is correct.

You can think of it as, say the LNA is a CS, then it converts the input voltage to a current which is developed into a load which is LNA Zout. If LNA Zout is 50, then when you shunt it with 1k (Mixer Zin) the net effect is to change Zload from 50 to 47.62.

Again, when designing RFIC front-ends, i rarely think in power except at the LNA input interface.

Hope this helps