Unless there has been some newer material published in the last 20 years, that document is pretty all-encompassing on the subject.
On the subject of carrier pinch-off:
"First, the intermittent disappearance of the carrier, which is believed by many broadcast engineers to be the cause of splatter, actually has very little to do with it."
Now, on the subject of negative peak clipping:
"Most important, it is apparent from Figure 3 that negative peak clipping is the major cause of splatter and that, theoretically, it matters little whether the clipping occurs as carrier pinchoff in the transmitter or as clipping in the audio prior to modulation. Comparing the spectra of
3 dB overmodulation in Figures 3f through 31, there is almost no difference between carrier pinchoff and audio clipping. The conditions of carrier pinchoff, -99% audio clipping, and -95% audio clipping, produce spectral
components which are stronger than -35 dB out to the fourth harmonic of the modulating frequency, and produce additional significant components even farther out."
It goes on from there, but you can go read the details for yourself.
I started this thread 5 years ago and would like to think that I've grown in my understanding of some things since then. Keep in mind that the first part of this thread is really chopped up bad and is missing quite a few posts due to system conversions. However, I wouldn't try to prove the same point again by using those methods.
First, I think we should stop using the word "distortion". That word has many different applications. Second, I think everyone can agree that what we are trying to avoid is splatter, IMD, and producing an all around "crappy" sounding device. The last part of that is subjective and can also mean a lot of different things to a lot of different people.
I get what CBDoctor is saying and what this mod is attempting to accomplish, and in a lot of cases is accomplishing. But there is nothing new about this technique...as W5LZ reminds us, people have been doing this since the inception of AM. This mod is just a basic negative clipper circuit with the built in modulation clipper disabled. The basic negative clipper circuit is widely available and is even shown here:
Clipper circuits : DIODES AND RECTIFIERS
Look familiar?
In principle, I don't have a problem with limiting the negative peaks to just under -100% and allowing the positive peaks to extend beyond %100. Ham AMers and all commercial AM stations do this. Of course, they don't get crazy with it, either because of the splatter it can cause. As NU9N eloquently put it,
"Of course, like anything else, there are some purity issues and tradeoff when using real aggressive peak-limiting. There will always be a fine balance between low distortion and aggressive peak limiting. The trick is finding the balance that will satisfy both sides of the equation - Loudness vs. Cleanness. A little bit of peak limiting will go a long way, so don't get too carried away!"
The biggest problem that I have with this type of mod in a CB environment is that there are too many other factors and variables happening that are not being accounted for. On the test bench, we send a 1Khz tone into the radio, look at it on the scope and say, "see -95%/+150% (or whatever) modulation...nice and loud, sounds good". In the real world, what we DON'T know is how the operator is going to use his radio. What mic is he using? Is it a power mic cranked all the way to oblivion? What are the audio characteristics of the mic? What are the audio characterics of the person using the mic? Where is this operator going to set the mic gain? What other equipment is going to be put inline that will affect the overall audio quality and spectral purity? Without observing the radio on a scope (or other equipment) while it is in use by the eventual operator, you have no idea what the end result will be.
Again, quoting NU9N,
"Even at low levels of speech, the dynamic content contained in normal speech is more than any transmitter can handle on its own! This is where vocal dynamic processing (Compressors, AGC Gain Riders, Peak Limiters, etc...) come to our rescue. They do exactly what is implied by their names; compress or limit the amount of the signal amplitude coming out as opposed to the signal amplitude that was going in."
[and]
"With good EQing, compression and tight asymmetrical peak limiting, even the most outrageous peaks cannot escape, insuring that a constant and controlled audio signal will be delivered to your transmitter, giving you the freedom to talk the way you want without worrying about excessive amplitude or distortion as a result of poorly controlled dynamics."
The standard CB station has no EQing, no compression, and I wouldn't call a simple negative clipper circuit "tight asymmetrical peak limiting".
When you hand the typical CB operator a radio that has the stock modulation limiter disabled with an NPC/RC mod introduced, they will hook it up with their power meter in line and proceed to crank the MIC Gain (and also crank the power mic up) until they see the most power output on their meter. So now we have a radio with no control on the audio side that is sending a lot of crap into a radio that is capable of modulating that crap at +125% to +200% while hoping our negative clipper circuit contains the crap to -95% modulation. So now we've created the potential for a loud, overmodulated, splattering monster. Even if they are careful with their mic settings, there still is very little control over the audio input to the radio other than whatever dynamics are introduced by combining some random mic with some random human voice characteristic. This is why some radios with mods like this sound "good" to our ears with some operators and the same radio with another operator's voice/mic will sound terrible.
So what do we really accomplish by doing this? For one, we've made it terribly easy for the CB operator to screw up the resulting output even easier than it normally would be. Second, we might have made it possble to be LOUD. Because of the operating habits of a lot of operators, it just ends up sounding like LOUD crap.
If it were so easy to just add a simple negative clipper circuit to an AM radio to make it sound good, do you think NU9N would need all this equipment in the stack prior to the rig itself?
Now, I realize that the CB operator isn't attempting to put together an AM station like that...but I hope you can see my point. Doing a mod like this and saying it doesn't
cause splatter when you're not accounting for all of the other things going on is like handing a loaded gun to an untrained kid and then saying it's not the gun's fault that it went off. OK, maybe that's a bit of an extreme analogy, but it makes my point. The end result is often a net LOSE.
BTW, here is a link to NU9N's AM site, which is a great tutorial:
NU9N AM Transmitter Hi Fi Audio Modulation Processing