[time-nuts] Measuring Osc Noise

[WarrenS]

Noise Nuts

I'm the first to admit that my explanations are confusing, but I hope a few 
will find my confusing better than others misleading or wrong.

Some seem to believe or suggest that there is some magic properties to the 
different types of Oscillator noise that makes it hard to measure.
Osc noise is just a signal, much like any other signal one would want to 
measure.
It has an amplitude and a frequency component, and depending on the type of 
noise, the signal amplitude either goes up with freq, down with freq, or 
stays the same over freq. The amount it goes up or down is generally 3 or 6 
db / octave depending on the type of noise.
The fact that the frequency content and amplitude is random for any short 
time period, does not change the fact it is still JUST a variable signal 
that covers a wide bandwidth range.
If one wants to make a tester that measures the 'noise signal' accurately, 
the tester has to be able to do the same as if it where any other signal, 
that is to have a flat amplitude response over the bandwidth of interest. 
That is about it.
The TPLL has a near perfect flat signal response from DC out to > 1KHz, only 
limited by any low pass RC filter(s) that one may want to add (and limited 
by the Ref osc). As such it does not know or care what type of osc noise it 
is measuring. It is not Prejudice. IF the noise is within it's bandpass 
range and above it's noise level, then it is going to measure it accurately.

Some are confusing measuring  noise with a flat, high resolution, low noise, 
method that in effect adds no significant noise or coloration of its own 
with what happens when a signal is measured with a device that as limited 
resolution and has a high noise level of its own.  The results of the 
limited device gives a 'noisy' answer even for a 'clean signal'.
Big difference between measuring  the noise of a signal and adding noise 
with the measuring device.
And an even  bigger difference in how one goes about to try and tell what 
the original signal was.

Maybe once I point out his obvious difference between signal noise and 
tester noise then others will explain it better.
Summery
Osc noise is not a mystery. Because the TPLL can measure any osc freq 
difference signal (including 'noise') over a wide bandwidth range, that 
includes DC, and at a resolution well below the noise level of the 
oscillators it is testing, with no dead time between readings, it's responds 
is the same to all noise types and levels. To think or clam otherwise is ... 
... unbelievable.

ws

****************
[time-nuts] Advantages & Disadvantages of the TPLL Method
WarrenS warrensjmail-one at yahoo.com
Tue Jun 15 21:43:25 UTC 2010

Didier

Please consider just because I can not explain it in words that you can
understand does make it not work.
The dynamic high gain EFC feedback is so much greater then the static
injection lock signal, by typ >100db, the injection lock just does not have
enough effect to change anything enough to be significant. My test prove
that (to my satisfaction).
 If you have a paper or theory that  says otherwise, it should include a
reason (besides Luck) that says why my test show otherwise.

Maybe best proof of that is the results from TPLL testing.
If you assume it is not all just luck,

Check out how close the simple TPLL matches the TSC5120A over the full tau
range,
OR for an even better example, see how good the simple TPLL does with the
swinging OSC test.
The advantage of the swinging Osc test is that it can easily be seen what
the waveform should look like, so no need to guess if it is right or not.
Hey, it works, better than ANYTHING else so far on that test, what else is
there to prove?  (besides that I know how to integrate and add)

ws
**********
[time-nuts] Advantages & Disadvantages of the TPLL Method
Didier Juges didier at cox.net
Tue Jun 15 19:54:43 UTC 2010
I promised myself I would not get into this any more, but here we go
again...

---- WarrenS <warrensjmail-one at yahoo.com> wrote:
>
> Charles posted:
> but the locked frequency will be different from both oscillators'
> free-running frequency and
> the EFC will not correctly indicate the test oscillator deviation
> because it isn't the only control input in the system.
>
> Good point and No argument  (except for the deviation part)
> Because the EFC is the only control input THAT IS VARYING.
>

Any parasitic control input is a problem in that system, like any other
system.
I thought the point of all this was to measure the noise of an oscillator?
If it is noisy (and they all are, to some level, otherwise you would not
need to measure it), then its frequency (or phase) is varying.

If the test oscillator is coupled (via injection locking) to the reference
oscillator, the test oscillator will force the ref oscillator to follow its
noise without the need to move the EFC. The EFC voltage will be stable
(because the oscillators move together), while you have two synchronously
noisy oscillators. If you measure the EFC, you will be left to believe your
oscillator is better than it is.

Please note that the effect is not simply a scaling factor, because
injection locking is a non-linear effect, or rather it is a mostly linear
effect over a typically very limited dynamic range. Small variations will be
totally coupled, where larger ones could possibly unlock the oscillators,
producing steps in the EFC voltage. Said another way, you cannot eliminate
the effects of injection locking by post-processing the data.

Injection locking is a parasitic control input and it is a problem with ANY
method that purports to measure noise. Ignore it at your own risk, but don't
say it does not matter, unless you want to prove something we already know.

Didier
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Swinging OscWS3c.gif
Type: image/gif
Size: 66544 bytes
Desc: not available
URL: 
http://www.febo.com/pipermail/time-nuts/attachments/20100615/e4c25279/attachment-0001.gif