What is 0dbm?

[Bob Bruhns]

Terry -

Using P = E^2 / R, 0 dBm (0.001W) into 50 ohms will be approximately 0.2236V RMS [SQRT(0.05)].
With a sine wave, this will be approximately 0.2236 * 2.828 = 0.632V P-P  [SQRT(0.05) * 2 * SQRT(2)].
But when checking these levels with a scope or hi-Z meter, you must use a load resistor of the specified impedance.

Some test generators are essentially zero-ohm source devices, so their output level will not be different if no external load is presented.  But more precise equipment tends to have the specified output impedance, which means the generator is in effect a perfect zero impedance source producing exactly twice the rated output into an open circuit, with a series resistance of exactly the rated impedance (50 ohms in this case) between the source and the output.

With the specified load on such a generator (50 ohms in this case), you will see the calculated RMS and P-P voltages.  But with no load (hi-Z scope, meter, etc) you will see exactly twice the calculated levels.

This can be fun when testing loads that are less than precise, such as phone lines, etc.  The impedance of typical phone lines would produce readings that were off by 2dB or so using instruments calibrated for 600.0 ohms.  True line loss was difficult to define.

I just ran into this problem with an HP function generator.  Without reading the manual (oops) I expected either 0 ohms or 600 ohms output Z.  Watching the scope, I placed a 10K resistor (it happened to be on the benchtop) across the generator output, and I did not notice the tiny reduction in signal level, so I assumed (oops) the generator had zero ohms output Z...  But the readings were off by 2:1.  Then I read the manual...  That function generator had a 50 ohm output.  With a 50 ohm load, all was well.

  Bob, WA3WDR


Terry Fox wrote:
> 
> In playing around again, I am trying to figure out how many millivolts
> peak-to-peak is 0dbm into 50 ohm load.  I have a chart from Andre that
> indicates 0dbm should be .225V, but is that rms, peak, or peak-to-peak.
> I did this last year, and convinced myself that it is rms, and 0dbm
> should be about .636V p-p, but in testing with real-world equipment I am
> now not so sure.  Can somebody take a relatively accurate 0dm LF/HF
> signal source, put it into a known 50 ohm load, and read off the
> peak-to-peak voltage ON A SCOPE for me?
> 
> I have done the above, but at a frequency where I don't trust the scope
> (50 MHz).  In addition, my Advantest spectrum analyzer from work says
> one thing, and an HP RF power meter (microwave) says something else.
> 
> Sorry to ask such a basic question, but it will help me calibrate
> stuff.  I guess 0dbm isn't as important as some accurate level that I
> can see on a scope and analyzer (+3, +10, +17 dbm).  I should know this,
> but my head starts hurting whenever I try to figure it out.
> 
> Terry  (tfox@erols.com)
> 
> _______________________________________________
> Tacos mailing list
> Tacos@amrad.org
> http://www.amrad.org/mailman/listinfo/tacos