[time-nuts] Phase Detectors/Mixers for DMTD and PN measurements

[Attila Kinali]

On Thu, 12 Dec 2019 06:53:41 +0100
Gerhard Hoffmann <ghf at hoffmann-hochfrequenz.de> wrote:

> It seems the mixer noise cannot be ignored.
> 
> I wonder then  why nobody  takes the mixer to cross correlation land, 
> and maybe even the driver amplifier.

Mostly because it doesn't do as much as you think it does.

If you want to do cross-correlation you have to duplicate the
whole chain from the mixer onward, which adds a lot of cost.
Otherwise you gain at most 3dB by splitting the mixer and
combining the signal immediately. At the same time you get
into trouble because now the IF outputs of the mixers are
now coupled which leads to periodically chaning impedance
as the mixers are not perfectly identical, which will
detoriate LO isolation. 

It is by far easier and more fruitful to work on the mixer itself
instead. E.g. the output capacitor, that Tobias asked about, smoothes
out the waveform of the IF signal, thus leading to a more linear
behavior of the mixer and lower harmonics (less noise down-folding).
Much simplified, the goal is to have a high impedance at the IF
frequency (maximizes output voltage, or rather minimizes loss)
and have a low impedance at LO, RF and 2LO/2RF/LO+RF frequencies
(to minimize harmonics and improve linearity). Of course there are
lots of other factors that come into play, like LO to RF isolation
(back feeding a signal close to the oscillator frequency is never
a good idea) and these are contradictory.

Another trick you can employ is to pass the LO through a limiting
amplifier. The idea is that the main contributor to noise in a
diode ring mixer is the noise (or uncertainty) of switching delay
of the diodes (which is not symmetric, btw). By having an LO with
a high slope, this uncertainty is reduced. But high slope means
high power, which might damage the mixer. Using a clamped signal
instead limits the power into the mixer and thus prevents it from
frying while giving the advantage of having fast rise/fall times
for the diodes to switch with (note that the swtiching time is
limited by how fast the space charge zone can be built up and removed).
Big disadvantage of this is, that now the LO has lots of harmonics
which will lead to noise down-folding. As 1/f noise is the limiting
factor for DMTD, especially even harmonics have to be kept as low as
possible, ie the duty cycle (or DC offset) of the LO signal should
as close to 50% (or 0, respectively) as possible. A nice side effect
of this is that the switch-on-switch-off asymmetry of the diodes is
reduced by having a higher slope, thus reducing even harmonics.

For DMTD applications, another limiting factor is phase shift
variation through the mixer. I haven't looked yet at which factors
contribute to this, but from the data I have seen, my guess it's
mostly temperature induced. A way to combat this is to inject a
pilot signal into both paths of the DMTD and use that as a reference
to cancel differences in the phase shift. With this most of the
temperature dependence can be removed. But this will introduce
a leakage path from one channel to the other, which will cause
problems if not properly handled. For details see [1].

			Attila Kinali

[1] "2π Low Drift Phase Detector for High-Precision Measurements"
by Jablonski, Czuba, Ludwik and Schlarb, 2015
https://doi.org/10.1109/TNS.2015.2425733
-- 
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, Neal Stephenson