[time-nuts] Distribution amplifier for 10MHz and 1 pps

[Bob kb8tq]

Hi

How low is low ? :) 

Based on the plot that Bill provides in the write up you linked to, he measured 
the insertion phase noise of the unit. His description is (as always) accurate:

“….this level of performance isn’t exceptionally good, it is quite adequate for 
its intended purpose …”

 = “PLENTY low enough for me”

The Rb sources he’s playing with limit what he needs in an amplifier. 

=========

So - how low does this or that *need* to be?  Is  -153 at 10 KHz good enough 
or does it need to be at -173? In some systems applications, the magic number 
is in the -180’s (with a price that is a bit insane ….). Only the system designer
can answer that question. There is no absolute limit to “low”. 

You can buy OCXO’s with phase noise in the -165 to -170 dbc/Hz range at 10KHz 
offset on eBay for not a crazy lot of money. Some are at 10 MHz others at odd 
frequencies. Quite a few are floating around at 100 MHz out of a Harris Farinon
box being parted out.

Why might you want a 100 MHz output? A 10 MHz unit will be 20 db worse at 
100 MHz.  Phase noise goes up 20 log N (N is the multiplier). A lot of people
go that way for microwave work. Some do GPSDO’s at 100 MHz. 

If you have an OCXO with noise at -173 dbc/Hz, it still will be degraded a bit by
a buffer that is at -173.  Your net output would be at -170 dbc/Hz. That’s just the
way noise works. At -163 on the OCXO and -173 on the buffer, the impact is small
enough that it really doesn’t count. 

If that all sounds like the whole story - nope. You can then get into various offsets
from carrier and go round and round some more.  The same 100 MHz OCXO
that “wins” over the 10 MHz unit above likely looses badly if you look at 10 Hz
offset. 

Lots of fun !!!!

Bob

> On Mar 21, 2019, at 5:01 AM, Tobias Pluess <tobias.pluess at xwmail.ch> wrote:
> 
> Hi,
> 
> if it is of interest for someone:
> I have made my own distribution amplifier for this exact purpose. Since I want sine waves, not square, I used some OpAmps. Hamilton Technical Services has published a design for a distributor here
> 
> http://www.stable32.com/An%20Octal%2010%20MHz%20Distribution%20Amplifier.pdf
> 
> and I have designed my own one which is quite similar.
> 
> Schematic of one channel:
> https://hb9fsx.ch/wordpress/wp-content/uploads/2018/11/ad9631-10MHz-driver.png
> 
> Frequency response, input matching and reverse isolation:
> https://hb9fsx.ch/wordpress/wp-content/uploads/2018/11/distrib-amp-bode2.png
> 
> Distortion:
> https://hb9fsx.ch/wordpress/wp-content/uploads/2018/11/distrib-amp-fft.png
> 
> 
> Since it uses OpAmps it is quite easy to design one stage for a particular desired input impedance. So one of the channels has 200 Ohms input impedance, and which results in 50 Ohms if 4 channels are paralleled - and thus the input and output impedances stay at 50 Ohms. So there is no need for a real "splitter".
> 
> So far I was not able to measure the residual phase noise of this distributor because I am still lacking a proper phase noise test set. (but it becomes better. I am currently making one. Hopefully I will finish someday :-)).
> 
> 
> Tobias
> HB9FSX
> 
> 
> ________________________________________
> From: time-nuts [time-nuts-bounces at lists.febo.com] on behalf of Perry Sandeen via time-nuts [time-nuts at lists.febo.com]
> Sent: Wednesday, March 20, 2019 23:50
> To: time-nuts at lists.febo.com
> Cc: Perry Sandeen
> Subject: Re: [time-nuts] Distribution amplifier for 10MHz and 1 pps
> 
> Yo Bubba Dudes!
> Wrote:
> Gosh. This topic comes up at least once every 6 months. Adiligent search of the time-nuts database would turn up amazing things. . .
> Yep, it does and it always will for at least two reasons.
> One, we have new people joining the list.  Which is great. They're at the beginning of a learning curve and most aren't in the upper echelons of TN knowledge.  It's a hobby for them not an extension of their professional lives.  But they want to learn and make progress
> 
> New members aren't aware of the past data base of info, and speaking for myself I never found an easy way to search.  If there is an easy way to do so I'm willing to be taught.
> I learn a lot from the posters even if it is far above my technical skills.  If I have no understanding or interest of post(s), I just delete it and keep what interests or is helpful to me.
> That's as it should be, and hopefully always will be, on a list like this with such a diverse mix of members.
> Now we all started spending 9 months in a nice warm water bath happily sucking our thumb.
> Then some dude takes us out, turns us upside down, and slaps us on our butt while saying *Welcome to the world*!
> Now what we think is: Say what sucker! I was perfectly happy where I was and minding by own business.
> However, like it or not, we are then at the start of the unending learning curve until our lights goes out figuratively or literally.
> Regards,
> Perrier
> 
> 
> 
> 
> 
> 
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.