Adjusting Matlab for 50 ohm impedance

I am trying to measure power of a complex waveform with power in multiple frequency bins.

Matlab gave a good example on their topic 'Measuring the Power of Deterministic Periodic Signals'

http://www.mathworks.com/help/signal/examples/measuring-the-power-of-deterministic-periodic-signals.html

I think in both examples they assume a resistance of 1 ohm, but my system is RF and has a 50 ohm impedance.  For the time domain input I was able to use P=V^2/R and compute power correctly.  I would like to get the same answer in the frequency domain as they have shown with the example from Mathworks.com.  My problem is when using the two functions they use 

[Pxx, F] = periodogram(x, hamming(length(x)), [], Fs, 'centered', 'psd');
power_freqdomain = bandpower(Pxx, F, 'psd')

I don't see how you can specify a 50 ohm impedance.  I assume you must adjust the input before you pass them into Matlab but I am not sure.

I would welcome any ideas on what I am missing.

Thanks.

On Friday, March 28, 2014 2:54:18 PM UTC-4, BH wrote:
> 
> I don't see how you can specify a 50 ohm impedance.  I assume you must adjust the input before you pass them into Matlab but I am not sure.
> 

MATLAB operates in sampled discrete time, so when it computes power the unit of that power is Blah^2, where Blah is whatever unit that sampled waveform is. And when it computes density, the unit is Blah^2/rad, again in discrete time. 

When you call functions such as Periodogram with a sampling rate Fs argument, it normalizes the output so that it is given in Blah^2/Hz. If your input is in Blah=Volts, and you want the output to be Volts^2/Ohm aka Watts, then you just need to divide the output by impedance in Ohm. 

And don't feel bad, the first time I tried to do the same years ago I spent a full day stressed out because I wasn't sure what was the way, there was no signal processing person to ask, and all the EEs were looking at me like I was retarded. It is good to be careful than to be careless.