Yaesu C4FM vs Plain Analog FM bandwidth - the same?

[xrunner]

I got a wild hair and wanted to compare the FM bandwidth of a Yaesu C4FM emission to plain analog FM (+/- 5 kHz deviation). I used a Yaesu FT-70D for this test. The stated bandwidth of C4FM is supposed to be 12.5 kHz, whereas the bandwidth of analog FM is about 16 - 18 kHz or so - and can be easily calculated using Carson's Rule. However; what I observed was that they are nearly identical (see pic). In the pic the C4FM is purple and analog FM green.

Research on the internet revealed other people had the same spectrum plots as I have done. I could find NO spectrum plots at all of Yaesu C4FM showing the bandwidth as 12.5 kHz - yet this is the stated specification for Yaesu C4FM everywhere you look. I posed this dilemma on a ham forum but got no one that could explain it. The question is - what am I misunderstanding about what I am seeing vs. the C4FM specification? The "bandwidth" of 12.5 kHz must be referring to some other aspect of the system design or ... what?

[radioactive]

If you are talking FCC language...

(d)Emission Mask D - 12.5 kHz channel bandwidth equipment. For transmitters designed to operate with a 12.5 kHz channel bandwidth, any emission must be attenuated below the power (P) of the highest emission contained within the authorized bandwidth as follows:

(1) On any frequency from the center of the authorized bandwidth f0 to 5.625 kHz removed from f0: Zero dB.

(2) On any frequency removed from the center of the authorized bandwidth by a displacement frequency (fd in kHz) of more than 5.625 kHz but no more than 12.5 kHz: At least 7.27(fd−2.88 kHz) dB.

(3) On any frequency removed from the center of the authorized bandwidth by a displacement frequency (fd in kHz) of more than 12.5 kHz: At least 50 10 log (P) dB or 70 dB, whichever is the lesser attenuation.

(4) The reference level for showing compliance with the emission mask shall be established using a resolution bandwidth sufficiently wide (usually two or three times the channel bandwidth) to capture the true peak emission of the equipment under test. In order to show compliance with the emission mask up to and including 50 kHz removed from the edge of the authorized bandwidth, adjust the resolution bandwidth to 100 Hz with the measuring instrument in a peak hold mode. A sufficient number of sweeps must be measured to insure that the emission profile is developed. If video filtering is used, its bandwidth must not be less than the instrument resolution bandwidth. For emissions beyond 50 kHz from the edge of the authorized bandwidth, see paragraph (o) of this section. If it can be shown that use of the above instrumentation settings do not accurately represent the true interference potential of the equipment under test, an alternate procedure may be used provided prior Commission approval is obtained.

[xrunner]

Hmmm ... Ok interesting, it has to conform to a mask. Never thought of that. I'm going to re-do the trace according to their recommended settings, but sort of roughly I guess it conforms to it, but it's not that easy to just look at and see that it does. Without that information the 12.k kHz makes no sense. I appreciate the info.

What about this from (3) -

At least 50 10 log (P) dB

is a math operator missing from between the 50 and 10 log ... like maybe a "+"? It can't be multiply can it?

[radioactive]

Yes,  I believe the operator should be + .  P is power in Watts ( I always shoot for the 70dB, even though highest power amp I designed was 30W).  The most interesting thing for you might be the RBW which is set to 100 Hz.  I am looking forward to what you measure for 100 Hz RBW.  This might explain the differences you are reading about?

[edited]

[xrunner]

OK let's see if I got this about right ... I have a new plot from the Yaesu HT set to VW C4FM mode. The analyzer set to a res. BW of 100 Hz, and it's 100 kHz wide so as to compare better to the example plot I got for Mask D for 12.5 kHz BW digital (now that I know what I'm looking for).

(d)Emission Mask D - 12.5 kHz channel bandwidth equipment. For transmitters designed to operate with a 12.5 kHz channel bandwidth, any emission must be attenuated below the power (P) of the highest emission contained within the authorized bandwidth as follows:

(1) On any frequency from the center of the authorized bandwidth f0 to 5.625 kHz removed from f0: Zero dB.

What I interpret this to mean is that, inside these limits of +/- 5.625 kHz, there is no requirement for the emmision to have any attenuation. however as seen it does drop off a little bit but that is OK.

(2) On any frequency removed from the center of the authorized bandwidth by a displacement frequency (fd in kHz) of more than 5.625 kHz but no more than 12.5 kHz: At least 7.27(fd−2.88 kHz) dB.

Now it gets more complicated because I can't generate a mask for an overlay, so taking only one point - my marker 2 is 6.0 kHz from Fo which is greater than 5.625 kHz. According to the equation -

7.27(fd−2.88 kHz) dB

we have 7.27 * (6.0 - 2.88) = 22.68 dB

My marker says -22.45, so close enough.

(3) On any frequency removed from the center of the authorized bandwidth by a displacement frequency (fd in kHz) of more than 12.5 kHz: At least 50 10 log (P) dB or 70 dB, whichever is the lesser attenuation.

Well, the HT is set to a Po of 0.5 W, however it's going through a 40 dB attenuator before the analyzer, but is it simply:

50 + 10 log(0.5) = 47 dB

So between 70 dB and 47 dB, 47 dB is the lesser attenuation.

So that looks about right. Not sure why they say "the lesser attenuation" seems like more attenuation would be better.

Any mistakes in understanding?

[radioactive]

Going back to your original question,  I think you might want to measure occupied bandwidth as the 20dB bandwidth for comparing the two modulations (even though you need to use the D mask for determining compliance in this case).  Find the peak, and move your markers to 20dB down from there and measure that bandwidth.  Just kind of eyeballing the first plot (RBW=1khz) you posted, that looks like might explain the difference in BW between modulations that you are trying to verify?  20dB BW looks like ~12KHz vs ~16KHz to me. 

[xrunner]

Going back to your original question,  I think you might want to measure occupied bandwidth as the 20dB bandwidth for comparing the two modulations (even though you need to use the D mask for determining compliance in this case).  Find the peak, and move your markers to 20dB down from there and measure that bandwidth.  Just kind of eyeballing the first plot (RBW=1khz) you posted, that looks like might explain the difference in BW between modulations that you are trying to verify?  20dB BW looks like ~12KHz vs ~16KHz to me.

OK I'll check that out some more, thanks for the assistance.