The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> Mixer simulation without input/output impedance matching
https://designers-guide.org/forum/YaBB.pl?num=1231854969

Message started by DoYouLinux on Jan 13^th, 2009, 5:56am

Title: Mixer simulation without input/output impedance matching
Post by DoYouLinux on Jan 13^th, 2009, 5:56am

Hi all,

I am trying to design a mixer for a receiver. In this mixer, it will be connected directly to the LNA and an IF amplifier at, respectively, the input and the output on chip. So, I don't need any impedance matching because the communication signals are voltages.

The problem is, when I simulate this mixer using SpectreRF, can I still use the port having 50ohm output impedance to ? Will this reveal a correct simulated result in this case of non-matched I/O mixer ?

Thank you very much in advance guys,

DYL

Title: Re: Mixer simulation without input/output impedance matching
Post by Andrew Beckett on Jan 13^th, 2009, 9:48am

When simulating with SpectreRF, you don't have to use ports. You can just as easily use voltage sources instead - which makes much more sense when you're dealing with a voltage based system. Use the "vsource" component in analogLib instead.

Ports are when you are working in a power-based system.

Regards,

Andrew.

Title: Re: Mixer simulation without input/output impedance matching
Post by DoYouLinux on Jan 14^th, 2009, 12:27am

Thank you very much Andrew :)

I just have a further question. If I use a normal voltage source in the simulation, can I still simulate the mixer's performances, e.g., NF, conversion gain, compression points using PSS, PSP, PXF, PAC and Pnoise ?

Thank you again for your suggestion :D

DYL

Title: Re: Mixer simulation without input/output impedance matching
Post by Andrew Beckett on Jan 14^th, 2009, 12:50am

Well, noise figure is defined as the ratio of signal-to-noise at the input to signal-to-noise at the output. The output would not be a problem if it was not terminated with a port, because all that happens is that the noise gets excluded from the load anyway. But the input is an issue - you need to have a noisy source at the input, which is what a port gives you (because of the resistor in the port). So I'm not convinced noise figure is a useful metric in a voltage-based system.

Conversion gain is straightforward, because you are just plotting voltage gain instead of power gain.

For compression point, I wrote a solution on this on sourcelink.cadence.com - number 11025262 - a while back. For this I used a vcvs and a port to transform dBV into dBm (effectively) and allow a port to be probed for measuring the output "power". The input sweep was a voltage source with the amplitude set to 10**(dbvrf/20) and you then sweep dbvrf. However, I think you could achieve the same thing by sweeping the voltage source the same way, but using the Net (dB, 1 ohm reference) choice for the Select field on the direct plot form. So no need to do the messing around with vcvs...

Regards,

Andrew.

Title: Re: Mixer simulation without input/output impedance matching
Post by didac on Jan 14^th, 2009, 2:06am

Hi,
I would say that you will need a port if system level metrics are the usual of RF design. For NF simulation you need a port with an output impedance equal to the LNA output impedance, if its a ressonant load you could make a "spot" simulation at one frequency, if its a narrowband system it won't change a lot within the band of interest, if its a broadband system you will need to run several simulations for the different impedances seen across the band.
For IIP3 and CP1dB if you use a port and want to have the actual input power you need to change the impedance of the input port to the complex conjugate of the input impedance of the Mixer, if not you will need to measure voltage levels and convert to dBm(knowing the impedance levels).
Hope it helps,

Title: Re: Mixer simulation without input/output impedance matching
Post by pancho_hideboo on Jan 17^th, 2009, 6:43am

Evaluation of RF blocks using ideal voltage source is not appropriate.

Higher harmonics are generated in input currents due to non-linearity of DUT.
But if you use ideal voltage source which has no impedance, higher harmonics voltages never exist in input of DUT.
This is very unreasonable.

Also see the following.
http://www.designers-guide.org/Forum/YaBB.pl?num=1193135745/3#3
http://www.designers-guide.org/Forum/YaBB.pl?num=1193135745/8#8
http://www.designers-guide.org/Forum/YaBB.pl?num=1052660826/1#1