[time-nuts] fast freq. synthesis schemes

EWKehren at aol.com EWKehren at aol.com
Thu Oct 15 13:21:37 UTC 2009


Luis
On the second one at 50 MHz compare you will be able to do 25 Hz steps  at 
18 GHZ since this is fractional. They will not be exactly 25 Hz because of  
the fractional nature but if exact steps are required you use a reference 
that  is a binary multiple, nothing major, I have done it to get exact 1 Hz 
steps.  Takes an AD4157 with a Xtal VCXO, in my case 134.217728 MHz which 
actually  gives me .33333 Hz steps. You should be able to get better than -90 
dbc  phase noise 20 KHz off. If you want to go that way I will be glad to get 
you  data for your requirement.   Bert
 
 
In a message dated 10/15/2009 8:56:31 A.M. Eastern Daylight Time,  
cupido at mail.ua.pt writes:

Hi  Bert,

Thanks for the input.

briefly;
-Phase noise is not too  important
-It is to make just a few (and cost is not a major  issue)

The suggestions using prescalers or any other high  speed
digital chips may bring simplicity while compared with
the design  using multiple loops and harmonic mixers and samplers
that is something I  had in mind when I place the question out.
I was wondering if recent  technology would make possible
to have a simpler approach.

I like  your first suggestion but I fear the spurious...
I have no clue how bad it  will be, but I guess I can only
be sure if I make a prototype of  that.
The second one I must check how small would be the step...
1 MHz  would be enough for a start.


Thanks
Luis  Cupido.


EWKehren at aol.com wrote:
> Luis
> There is not  really enough information to make a good recommendation. Is 
 
>  this a one off or is the plan to make multiple units, what are the phase 
 
> noise  requirements, what are the cost limitations? If cost is a  major 
issue 
> there are  in my opinion only two  options.
>  In both cases a VCO has to be used presetting with a  fast  D/A.  In the 
> first case you pre scale by 16 and feed  an AD 9910 DDS as a  
programmable 
> divider using a 100 MHz output  in to a Hittite HNC 439 phase  comparator 
with 
> all the  shortcomings of a DDS multiplied by the total divider  ratio. 
>  The second option is a Hittite HMC 700 fractional PLL and use 50 or 60  
MHZ  
> compare frequency. You have to pre scale by 2 . Having used  the all the 
> above  components I can tell you the DDS works  reliable at 1200 MHz in 
and the 
> 700  works great at 9  GHz.
>  I have not gone for switching speed but in my case emphasis  is on  
phase 
> noise and resolution. As a matter of fact a friend  of mine and I are 
doing  a 
> low cost signal generator from 100KHz  to 26 GHz in one Hz resolution 
using 
> the  Hittite 700 part.  Phase noise goal at 8 GHz is -100 dbc, 10 KHz 
offset. 
> We use   surplus yigs.
> Bert Kehren WB5MZJ
>  
>  
>  In a message dated 10/14/2009 3:25:33 P.M. Eastern Daylight Time,   
> cupido at mail.ua.pt writes:
> 
> Hi,
> 
> I'm  looking for the schemes used on the
> frequency  synthesizers that  change frequency
> in few microseconds time (or less)
> at   microwaves lets say circa 12-18Ghz)
> Obviously with some resolution  (let's  say 100Khz step
> or in that order)(otherwise it  would
> be a trivial  exercise in the BW of the PLL loop  filter)
> 
> Does anyone know of some  paper or tech notes  from
> some instrument or modules that show block  diagrams
>  of such?
> 
> I have many info on mw synthesizers but all  fall  into
> microwave radio style of things with much  higher
> resolution (khz  and less) and much longer
>  switching times (millisencods or more).
> Info  on fast stuff I  can't really find.
> 
> 
> I'm thinking of experimenting  on  something along those
> lines so... all comments are   welcome.
> 
> Thanks.
> 
> Luis  Cupido.
>  ct1dmk.
> 
>  _______________________________________________
> time-nuts   mailing list -- time-nuts at febo.com
> To unsubscribe, go to   
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and  follow the  instructions there.
> 
> 
>  _______________________________________________
> time-nuts mailing list  -- time-nuts at febo.com
> To unsubscribe, go to  
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the  instructions there.
>  

_______________________________________________
time-nuts mailing  list -- time-nuts at febo.com
To unsubscribe, go to  
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the  instructions there.




More information about the time-nuts mailing list