|
 |
PLL ( Offset Loop ) configuration - RF Cafe Forums
|
mac
Post subject: PLL ( Offset Loop ) configuration Posted: Tue
Dec 09, 2008 3:22 am
Captain
Joined: Fri Nov 02,
2007 3:45 am
Posts: 8
Location: germany
Hi all,
I am in
a PLL design process,
trying to tame a VCO 1.5-3.6GHz with a PLL.
Main goal is low phase noise -
cause the frequency range
is that wide i thougt about dividing vco signal down first by variable
N (maybe N= 6-16 to get to Frequency Range 250MHz - 292MHz) to narrow
frequency range and mixing it down another some MHz afterwards ( e.g
LO Frequ = 300MHz ), all in the same feedback path supplying phase detector
(with this IF 8MHZ- 50MHz ).
I have often seen it the other
way round, mixer at first then dividing down and feeding pd - but then
i get another wide frequency range after the mixing process
anyone done this configuration before, or seen it ?
or is it out
of question - worrying about spurs especially from divider harmonics
feeding the mixer -
mny. thanks for comments
mac
_________________
regards
mac
Top
markapexwireless
Post subject: Posted: Wed Dec 10, 2008 3:38 pm
Captain
Joined: Tue Jul 08, 2008 7:15 pm
Posts: 11
Location: Boulder
Colorado
Mac,
Most up mixing PLL synthesizers were used
to generate small frequency increments at the output, which is no longer
much of a problem with Fractional N synthesizers and DDS driven PLLs.
If your VCO must tune over an octave, you will probably be stuck
with fairly poor phase noise outside the PLL bandwidth, which practically
is over 200KHz or so from the carrier. A fundamental problem with VCOs
is the wider they tune the higher the phase noise. If you mix a poor
phase noise VCO with a clean one, you will get a poor phase noise VCO.
If you do not need to sweep entire band quickly, you could switch
in multiple narrow band VCOs, each with better phase noise performance
than one that tunes an entire octave. If you can design VCO, you could
switch in multiple resonators essentially giving you lower noise sub-bands.
Good luck on your project.
- Mark -
Top
mac
Post subject: Posted: Thu Dec 11, 2008 3:22 am
Captain
Joined: Fri Nov 02, 2007 3:45 am
Posts: 8
Location: germany
Dear Mark,
thanks for your reply -
You're right, if i use a wideband vco phase noise gets worse cause
of high
Kv from the vco etc. - i build some systems with switched
vcos or resonators in the past but that was not what i am for with my
question.
Maybe i missed to say that tuning speed is important,
too and there is the old problem - Phase noise / Loop bandwith / switching
speed all relates to each other.
I use this wideband vco to design
a wideband synthesizer dividing down for the lower frequencies, with
low noise and fast tuning capability and knowingly choose a wide bandwith
pll.
I ve seen what all the other guys do from agilent/ rohde schwartz
etc
with their wideband synthesizers using a mix of DDS/ Frac -N
and convetional PLL Techniques and they use wideband vcos, yigs etc
aswell
in past times as fast dividers were not available it
was common to design plls with a down conversion mixer to improve phase
noise because of running the phase detector on a lower frequency or
using harmonic samplers combining downconversion mixer and phase detector
ref signal -- pd --- filter --- vco
\divider -- mixer
/
\lo signal ( or harmonics, with vco presteering )
but
now fast dividers are available and i wondered if i cant interchange
divider and mixer in the feedback path or if i will get problems with
this configuration
ref signal - pd - filter - vco
\ mixer
- divider /
regardless of answers think i have to set this
up and
evaluate it in the lab
mny thanks
mac
_________________
regards
mac
Top
markapexwireless
Post subject: Posted: Fri Dec 12, 2008 1:06 pm
Captain
Joined: Tue Jul 08, 2008 7:15 pm
Posts: 11
Location: Boulder
Colorado
Mac,
Please post your results. We are always looking
for techniques to reduce phase noise in fast tuning wideband synthesizers.
Thanks,
- Mark -
Posted 11/12/2012
|
|
|
|
|
