[time-nuts] Thunderbolt spurs on 10MHz output at 100Hz and 200Hz from signal.

[Bruce Griffiths]

On Wednesday, September 21, 2016 08:59:18 PM Attila Kinali wrote:
> On Wed, 21 Sep 2016 09:16:18 -0700
> 
> Dan Rae <danrae at verizon.net> wrote:
> > On 9/21/2016 9:01 AM, David C. Partridge wrote:
> > > Yes,  I fear a timepod is a bit out of my budget (I can dream tho'). 
> > > Over here in the UK PN measurement kit is a bit thin on the ground 
too.
> > > 
> > > So I looked at 
<http://www.wenzel.com/documents/measuringphasenoise.htm>
> > > but that seemed a little thin on details.  Or were you referring to
> > > something else?> 
> > The Wenzel note is at: http://www.wenzel.com/documents/circuits1.htm
> > 
> > as a pdf under link: "Low noise amplifier for Phase Noise 
measurements"
> > or try:  www.wenzel.com/wp-content/uploads/lowamp.pdf
> > 
> > Pretty much all there, but the FET is hard to find now.
> 
> If you want to build such an amplifier, then the first stage (aka the
> jFET input and the first opamp) is the critical component. There are
> other, more modern descriptions for this kind of application, in case
> you don't want to just replace the jFET by one if its modern cousins.
> 
> One well known is AN124[1] and it's more current cousin AN159[2].
> For the later, Bruce commented that an jFET input stage would
> probably be more quiet. Gerhard Hoffman has designed a similar
> system[3] that uses a couple of paralel low noise opamps instead
> of a jfet stage and claims a noise floor of 220pV/sqrt(Hz). All of
> these would need to be adapted to imporve thier frequency response
> up to 10MHz (or actually a bit higher), but then the input stage
> gets also a bit simpler as you don't need the huge capacitors anymore.
> 
> I also recommend a look at [4] which lists a couple of issues with jFET
> input stages for opamps and how to get around them.
> 
> 			Attila Kinali
> 
> 
> [1] "775 Nanovolt Noise Measurement for A Low Noise Voltage 
Reference",
> by Jim Williams, Linear AN124, 2009
> 
> [2] "Measuring 2nV/sqrt(Hz) Noise and 120dB Supply Rejection
> on Linear Regulators", by Todd Owen and Amit Patel, Linear AN159, 2016
> 
> [3] "A 220 pV/sqrt(Hz) low noise preamplifier", by Gerhard Hoffman, 2014
> http://www.hoffmann-hochfrequenz.de/downloads/lono.pdf
> 
> [4] "Some Tips on Making a FETching Discrete Amplifier", by George
> Alexandrov and Nathan Carter, Analog Dialog 47-10, 2013
> 
http://www.analog.com/library/analogdialogue/archives/47-10/discrete_ampl
ifi
> er.html

2sk369BL are still available from:
http://www.ampslab.com/trans_2sk369.htm

I obtained some from them, they met the noise specs so are probably not 
counterfeit.

One issue with the Wenzel preamp is that it has poor PSRR.
This can easily be improved substantially by redesigning the bias circuitry 
for the opamp noninverting input.

Another issue is the requirement to trim the current sink for low output 
offset. If one takes advantage of the fact that the PLL imposes a low 
frequency cutoff to the PN measurements, the amplifier input   can be AC 
coupled, allowing dc bias feedback to be applied to the input device gate.
Use a parallel dc coupled JFET input opamp for the PLL

Using a single ended input stage is significantly quieter than using a 
differential pair.

Paralleled BF862's can be substituted for the input JFET.

Bruce