[time-nuts] 10MHz Square to Sine Wave Conversion

Attila Kinali attila at kinali.ch
Fri Jul 17 22:50:53 UTC 2015


On Fri, 17 Jul 2015 18:23:51 +0000 (UTC)
Michael <mikenet213 at comcast.net> wrote:

> Before I waste time simulating too much...
> 
> Does anyone have any intuition around temperature dependence
> of these designs? Is one 'style' significantly better than another?
> 
> I'm far more concerned about phase shift of the fundamental during
> temperature swings than I am relative harmonic levels/phase moving around. 

Temperature dependence of phase (mostly) happens when you have filters
that have temperature dependent components, which in turn shift the
resonance frequency. When you look at the Bode plot of a first order
filter, you see that the phase changes from 0° to +/-90° around the
corner frequency. How fast the change (as a function of temperature) is
depends on the quality factor of the filter. The wikipedia page for
Bode plots[1] give a quite good intuition how steep the phase change is.

The only thing left is, to figure out how much the component values
change with temperature. Unfortunately, this is highly material
dependent and you need to check this for each component individually,
then calculate from this the frequency change.

A good engineering approach is to make the filters you have as
wide as resonably possible (low Q), to ensure that you stay in
a region of low where the rate of phase change per temperature
(i.e. d(phase)/d(T) ) is low. Or if you have bandpass filters,
make them wide enough that the upper and lower corner frequencies
never come close to the signal of interest.

If you have taken care of all the filters and ensured that they
do not interfere in any way with the phase of the signal, then
you can go to the next level ;-)
Temperature affects the dimensions of all materials. But a change
in length also means that the phase at the output of the cable/wire/...
changes. Fortunately, this effect is very low (measured in ppm/K,
and e.g. ~20ppm/K for copper, 10-30ppm/K for most elemental metals).
Which means, for most practical devices, you do not care about this,
unless you do high precision measurements, with low noise sources.
It's definitly not an issue when you are dealing with signals comming
out of the aether. (Phase shifts due to atmospheric effects will
dominate this by many orders of magnitude)

HTH

			Attila Kinali


[1] https://en.wikipedia.org/wiki/Bode_plot#Rules_for_handmade_Bode_plot

-- 
I must not become metastable. 
Metastability is the mind-killer.
Metastability is the little-death that brings total obliteration.
I will face my metastability. 
I will permit it to pass over me and through me. 
And when it has gone past I will turn the inner eye to see its path. 
Where the metastability has gone there will be nothing. Only I will remain.




More information about the Time-nuts_lists.febo.com mailing list