Hi

Just as with PM noise, AM noise is a “that depends” sort of thing. The first
dependency is on your test gear. If the AM noise was well below the PM
noise, would you be able to measure it? Equally, if you are doing something
like a diode detector - can it measure AM below XXX dbc? If the detector is
only good to -120 dbc/Hz, you aren’t going to “see” -130 ….

Noise on output state supply lines can generate AM. In some cases the stage
is a better AM modulator than PM modulator. Limiting (like with a logic gate)
is not going to pass AM noise well.  That combination can make for some
“interesting” noise profiles.

One “old time” assumption for wide band noise is that there is no process that
will generate AM independent of PM. Thus they will always be in an equal power
relationship. The key phrase here is wide band. Think in terms of notching out the
carrier and looking at the result on a spectrum analyzer in this case.

One of the classic arguments for an AGC or for certain types of limiters in
oscillators is AM to PM conversion. The concern here is “close to carrier”
rather than far removed. There certainly are examples tossed around where
AM noise is significantly higher than PM noise at low offsets.

So, no data that I can share. Maybe a few things in there will be of help.

Bob

Attila,

Since we talk background noise and white noise, the amplitude is the
same for AM and PM. This is part of the AM/PM lecture of NIST that I
know you have participated in at least once.

Under the assumption of low modulation index, which is fair assumption
for background noise compared to most carriers, both the AM and PM noise
of a certain side-offset has two side-band peaks, a lower and an upper.
The big difference is that for AM they have the same polarity and for PM
they have opposite polarity. Thus, they are as orthogonal as common mode
and differential mode. Noise on both absolute frequencies will combine
and contribute to the same AM and PM levels. It's that simple.

So, for that scenario you know one you know the other.

Now, for actual sources this is no longer true. The AM noise can be much
higher, which is why it can be a real danger to the PM noise if there is
a AM to PM noise conversion. One source of such conversion can be the
amplification stage, but another could be a mistuned filter, which have
different amplitudes of the side-bands, which can create conversion as
the balance does not balance the same way anymore.

Also turns out that the nitty-gritty of cross-correlation spectrum
analysis also occurs in AM-to-PM conversion and cancellation. This I had
not paid full attention to, but got reminded off at the workshop. Thus,
they are tied together and should be measured and understood together.

Cheers,
Magnus

On 01/02/2018 08:55 PM, Attila Kinali wrote:

On Tue, 2 Jan 2018 23:34:18 +0100
Magnus Danielson magnus@rubidium.dyndns.org wrote:

[About AM noise being of equal power as PM noise]

Yes, exactly. I am currently trying to understand how noise affects
circuits an how input and circuit noise get converted to output noise.
First assumption that needs to be dropped is that the noise processes
is purely additive and independent of the signal. This means that a
noise process does not anymore produce equal AM and PM power.

I think I have a 90% solution of the noise processes and conversions
in a sine-to-square converter (aka zero-crossing detector, aka comparator).
But there is one process that keeps puzzling me. I think I know where in
the circuit it must come from, but I have no explanation as to how it happens.

			Attila Kinali

--
It is upon moral qualities that a society is ultimately founded. All
the prosperity and technological sophistication in the world is of no
use without that foundation.
-- Miss Matheson, The Diamond Age, Neil Stephenson

Hi

If I pass both a sine wave tone and a pile of audio noise through a perfectly
linear circuit, I get no AM or PM noise sidebands on the signal. The only way
they combine is if the circuit is non-linear. There are a lot of ways to model
this non-linearity. The “old school” approach is with a polynomial function. That
dates back at least into the 1930’s. The textbooks I used learning it in the 1970’s
were written in the 1950’s. There are many decades of papers on this stuff.

Simple answer is that some types of non-linearity transfer AM others transfer PM.
Some transfer both. In some cases the spectrum of the modulation is preserved.
In some cases the spectrum is re-shaped by the modulation process. As I recall
we spend a semester going over the basics of what does what.

These days, you have the wonders of non-linear circuit analysis. To the degree
that your models are accurate and that the methods used work, I’m sure it will
give you similar data compared to the “old school” stuff.

Bob

Is this an audio tone, summed with audio noise whose spectrum surrounds
that of the tone?

Dana

On Fri, Jan 5, 2018 at 9:56 AM, Bob kb8tq kb8tq@n1k.org wrote:

Hi

The audio (or RF) tone is summed with “baseband" noise. 1/F noise seems
to be the flavor of the day in recent postings. The only reason to use audio
in the example is that it is really easy to demonstrate things at audio with
a sound card.

Bob

But what I'm wondering, because this is important to the discussion, is the
tone at a frequency encompassed on both sides by the noise band?  Or
is the tone outside the noise band?

Dana

On Fri, Jan 5, 2018 at 1:35 PM, Bob kb8tq kb8tq@n1k.org wrote:

Hi

It can be either. The easy example is a tone that is outside the entire band of the noise.
If it is a “real” noise spectrum, that’s never going to be the case. There will always be
some noise at the tone frequency in a real system.

Bob

Hi Attila,

On 01/05/2018 12:27 PM, Attila Kinali wrote:

I'd say that first assumption should be that you are not linear.
Still, linear systems can already do conversion of AM to/from PM. The
lack of balance keeps being ignored for so many cases.

Then with nonlinearity things intermodulate and you get all the fun you
want.

OK, let me see the notes on that...

Cheers,
Magnus