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

[Gerhard Hoffmann]

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.

The FFT analyzer can do it anyway, a second AF pre amp costs nothing,  
and other than that

there are only another mixer and 2 power dividers on the BOM.

In addition, each mixer gets 3 dB less power which helps the survival rate.

One can always crank it up for bragging rights.


Also, high level mixers (type 2 and 3 in [1] ) have resistors in series 
to each diode for biasing

so their lower efficiency should not come as a surprise.  And, the 
resistors in the ring generate

AF noise. With DC across them also 1/f, esp. for thick film.

The usual 1nV/rt Hz op amps (LT1028, AD797, ADA4898) have the voltage 
noise of a 60 Ohm resistor.

That probably brings them to the point of diminishing return.

The diodes in the ring themselves create only half-thermal noise.


Methinks one would be better off with many low level mixers and power 
dividers,

adding up the demodulated AF voltages. There is no point in power 
matching the AF side.

Maybe for RF with a RC high pass.

SMD Wilkinson dividers are $2.50 or so now from Macom, MCL or Pulse and

low level mixers are also cheap, or even 2 1:4 SMD transformers + diode 
ring.

Easily done when one does not have to wind the coils.


When verifying my links below, I stumbled across M9H  for $333.

The price for two of them is the number of the beast.  =8-) )


[1]

< 
https://www.jlab.org/uspas11/Reading/RF/Mixers%20-%20phase%20detectors.pdf >


< 
https://www.google.de/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=2ahUKEwiu8Zi7pa_mAhUKQUEAHb1pAyAQFjAAegQIAxAJ&url=http%3A%2F%2Fwww.rfcafe.com%2Freferences%2Farticles%2Fwj-tech-notes%2Fmixers-theory-technology-p2-v8-3.pdf&usg=AOvVaw3fc3OZgN2H5jzPx6jXDBh9 
    >


Also interesting:

< 
https://www.google.de/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=2ahUKEwiFuqCvpq_mAhVTh1wKHcc2ALQQFjAAegQIAxAJ&url=http%3A%2F%2Fliterature.cdn.keysight.com%2Flitweb%2Fpdf%2F5989-8999EN.pdf&usg=AOvVaw3RFawCNg543QEv1hUsscbY 
   >


regards,

Gerhard,  DK4XP




Am 11.12.19 um 20:29 schrieb Bob kb8tq:
> Hi
>
> There are some famous name papers showing data taken on the 3048. When
> asked about the levels involved, the next question inevitably was - “How many
> mixers did you fry running the test?”.
>
> Bob
>
>> On Dec 11, 2019, at 10:54 AM, Richard (Rick) Karlquist <richard at karlquist.com> wrote:
>>
>>
>> The HP 3048A phase noise system has a phase
>> detector in the HP 11848A chassis that was
>> originally a Watkins-Johnson (the original
>> company) M9H.  The M9H lives on now sold by
>> MaCom Technology Solutions.  "Someone" (unknown)
>> established that the Mini-Circuits JMS-5H
>> is an acceptable substitute, although not
>> an exact replacement.
>>
>> The tradeoff in picking a mixer is the LO
>> level.  A high LO drive level such as the above
>> mixer have, gives higher performance, provided
>> you have a driving amplifier that has both the
>> necessary output power is itself low phase noise.
>> Depending on the test set up, some of the driver
>> amplifier phase noise will common mode out.
>>
>> Also, high drive level mixers have a narrow window
>> between high enough drive to operate vs the damage
>> level.  So a third requirement on the driver
>> amplifier is that it's maximum output level is
>> less than the damage level.  (Don't ask how I
>> know this :-).  You can always put a pad between the
>> drive amplifier and the LO input of the mixer used
>> as phase detector to adjust the maximum possible
>> drive.
>>
>> Rick N6RK
>>