The AM Forum

THE AM BULLETIN BOARD => Technical Forum => Topic started by: N1BCG on March 09, 2020, 10:34:54 AM



Title: Audio Peaks: Friend Or Foe?
Post by: N1BCG on March 09, 2020, 10:34:54 AM
While we all can agree that -100% modulation is a hard limit to be avoided, there are a surprising number of amateur radio ops who strive for positive modulation far above 100%. Why? What are the benefits of soaring positive peaks?

What makes this question even more meaningful is that the broadcast audio industry has perfected the reduction of peaks by increasing average modulation energy, so there are very stark contrasts in how peak energy is viewed between professional and amateur radio.

Peaks may be of high amplitude but are always of short duration and therefore do not contain useful energy. Conversely, average power is sustained energy and contributes significantly to perceived loudness.

A simple example of peak energy is within a static discharge between your finger and a wallplate. These sparks can be of 20-40kV but have a duration of just 1-100nS. It would make a significant difference if this energy had a longer duration of several seconds, right?

So what effect does the peak component of modulation have if it doesn't contribute to loudness?

In AM, negative peaks are the first part of the audio waveform to reach -100%, leaving average modulation reduced. Fortunately, negative peak clippers address this nicely and are widely used. Positive peaks are the first to reach 1500 Watts PEP even with modest carrier power. A 375 Watt carrier reaches 1500 Watts when modulated to +100%. 1500 Watts is also reached with a 240 Watt carrier modulated at +150% and just 167 Watts of carrier reaches legal limit at +200%. So why no positive peak clippers?

An equation for PEP of an AM signal: PEP = ((M+1)^2)C, where C is the carrier (Watts) and M is the modulation factor. (https://www.qsl.net/wa5bxo/amplitude-modulation-and%20pep.htm)

100% modulation would be M = 1, 200% would be M = 2, etc. So with 100% modulation, M = 1, M + 1 = 2, 2^2 = 4 (2^2 means 2 squared), so PEP = 4C (Carrier * 4).

With 200% modulation, M = 2, M + 1 = 2 + 1 = 3, 3^2 = 9, so PEP= 9C (Carrier * 9).

There would be a lot more flexibility if communicating were just about transmitting, but it's important to include the effects of propagation and signal detection in receivers. A highly asymmetric waveform is more susceptible to the effects of QSB than a symmetric waveform, resulting in increased distortion. Further, many detectors do not handle asymmetry well and will clip off peaks extending beyond a symmetric waveform, resulting in increased distortion. So much for that FB audio.

Even though SSB is properly measured in PEP, peaks are the first part of the audio waveform to trigger ALC action, leaving overall modulation reduced.

In any mode, excessive peaks increase the risk of arc-overs in RF and modulator components which can be particularly damaging to expensive mod iron.

This illustration shows the relationship between the various components of a signal:

(https://vertassets.blob.core.windows.net/image/c64b1890/c64b1890-be4b-49f3-b256-90cec7341170/lindsayfig1.png)

This display shows received audio with peaks left as-is and clipped off:

(http://www.internetwork.com/MAX/images/Peaks_vs_Average.jpg)

The latter contains significantly higher average audio energy and is dramatically louder even though the modulation level remained the same. Despite the extreme difference, there was no noticeable increase in distortion because only short duration peak energy was removed.

The goal of this post is to shed light on the difference between average and peak energy. As mentioned earlier, the broadcast audio industry has put a significant investment in competitive audio over the past half century and the lessons are there for the taking.


Title: Re: Audio Peaks: Friend Or Foe?
Post by: KK4YY on March 09, 2020, 11:17:42 AM
I totally agree. :)

But others may have a different reaction...


Title: Re: Audio Peaks: Friend Or Foe?
Post by: N1BCG on March 09, 2020, 12:23:45 PM
^^^ Hah, perfect!  That's what I'm going to hear from now on when the topic comes up.

It's strange jumping between Part 73 where the F.C.C. can put the hurt on you in epic ways and Part 97 where the rules get a wink and a giggle at best. I remember an AM station inspection where the agent noticed the negative peak indicator flash and began counting the number of flashes per minute. Two very different worlds.

"Legal limit", or whatever the interpretation is at the moment, is merely one point. Just last night, I heard an op demonstrate +300% on his Viking II. It was jarring to hear and more so to think about a rig of that vintage being put through hell. Shorted mod iron windings would have been a harsh lesson and I wish that on nobody.



Title: Re: Audio Peaks: Friend Or Foe?
Post by: KD6VXI on March 09, 2020, 01:24:27 PM
I completely agree with keeping the peaks limited to symmetrical audio levels.

Having your audio processed to remove the assymetry, compressing the frequencies needed and eqing the audio signal to produce more torque under the curve is the hot ticket.

Increasing positive peaks in the modulator to increase apparent loudness is not a bad thing, either, in my opinion.

I don't believe in just letting your peaks fall where they may.  This creates a natural sound, but avg power is WAY down typically.

'-Shane
KD6VXI


Title: Re: Audio Peaks: Friend Or Foe?
Post by: KK4YY on March 09, 2020, 01:26:55 PM
It's strange jumping between Part 73 where the F.C.C. can put the hurt on you in epic ways and Part 97 where the rules get a wink and a giggle at best.
It's more than a jump between Part 73 and Part 97... it's a leap of faith!


Title: Re: Audio Peaks: Friend Or Foe?
Post by: WBear2GCR on March 10, 2020, 01:31:26 PM
Two different things are being commingled here.

One is maximum peak audio modulation by %.
The other is audio "density".

They are not directly related and one does not preclude the other.

                       _-_-


Title: Re: Audio Peaks: Friend Or Foe?
Post by: N1BCG on March 10, 2020, 03:40:28 PM
Two different things are being commingled here.

One is maximum peak audio modulation by %.
The other is audio "density".

They are not directly related and one does not preclude the other.

                       _-_-

Correct, and understanding the relationship between the two is critically important for effective modulation control. For example, a compressor intended to increase average modulation by creating a “dense” sound can actually reduce overall modulation. This happens as a result of long attack times allowing waveform leading edges to pass unattenuated. Loads of peaks but at the expense of loudness.

Ironic, eh?

Response times, both attack and release, are also important factors. The key is how these affect peak and average modulation and knowing the differences between compression, limiting, and clipping. Each serve a purpose.


Title: Re: Audio Peaks: One Unit Does It ALL!!
Post by: WBear2GCR on March 10, 2020, 04:33:53 PM

What is needed is a circuit known in some circles as a Densocompypeaker.

It increases density up to 99.99% pure modulation (like fine Ivory soap) while boosting
the peaks (set by the peaker control knob) to between 100% and 199.99%. Both at
the same time!

If you have no peaks, it will generate them for you.

And, it adds a double BA without any "prox effect" at all, with the push of a
button!

Presents ride?
Yep, if you want "Presidents Rye's" (1 quart US bottles) you can get that by choosing one of the presets!
For a bit extra you can get the special DSP gold package of presets that includes:
- space shuttle audio
- breaker 1-9 audio
and importantly:
- Donald Duck audio

The product will be released and available according to company spokespersons on
April 1, 2020.

Look for it!

            _-_-
AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands