[hpsdr] Level calculation error in Two Tone Test generation routine

[Scott Traurig]

This is a really great thread, and I've enjoyed reading it so far!

I had long since noticed that the two-tone signal displayed each tone 6dB
down from an equivalent total power single-tone tune signal. This has
bothered me for quite some time, for I also, like John, believe that the
two-tone signal should have each tone shown -3dB down from the equivalent
total power single tone signal. Why I never questioned this, I'm not
certain, however I suspect that I continue to be more concerned with UI and
audio processing issues than with signal processing issues, as the latter
are in generally excellent shape!

After reading this thread, it would appear to me that you are both correct,
but that the display of the two-tone power at -6dB remains a bit of a
mystery.

Warren quite correctly points out that, in the voltage domain, a single
tone has a peak-to-average ratio (PAR, or crest factor) of 3dBVoltage, and
the two-tone signal a PAR of 6dBVoltage. The conversion to decibel units in
the voltage domain is 20log(voltage).

At the same time, John also quite correctly points out that, in the power
domain, a single tone as a PAR of 0dBPower, and the two-tone signal a PAR
of 3dBPower. The conversion to decibel units in the power domain is
10log(power).

I am not surprised that you think more naturally in terms of voltage,
Warren, as one of your primary goals is to obtain a particular set of
voltage conditions on a DAC.

Similarly, I am not surprised that John and I think in terms of RF output
power, because, like most RF engineers, we think more naturally in those
terms. Indeed, I almost never think in dBuV, dBmV or any other set of
voltage units, only in terms of dBm (dBmW, to be more precise, but "dBm" is
the convention in the RF engineering lexicon).

So, if my understanding that both positions are correct is correct, that
the math depends on whether one is in the voltage or power domain, there
nevertheless remains a mystery: why does the two-tone signal show each tone
down 6dB from the single tone equivalent total power condition on the
transmit spectral display? My best guess now is that the transmit spectral
display (normal or DUP mode) is displaying dBVoltage. If so, this is quite
disconcerting, because spectrum analyzers universally display in units of
power, typically dBm.

Thanks & 73,

Scott/w-u-2-o
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20161124/5d89aa28/attachment.htm>