Post subject: Unwanted Modulation - Nasty Stuff!
Posted: Fri Jun 12, 2009 10:59 am
Mon Feb 09, 2009 12:26 pm
I'm sure that at some
time(s) in the past and if not then definately in the future you
lucky RF people will experience unwanted modulation on your transmitter
carrier / receiver local oscillator. This nasty issue has reared
it head (again) on a recent design I am working on so it seems appropriate
to bang in a post on this subject. Maybe there is a design gem or
two out there in one of your brains that may help me and others
out now and in the future?
My particular nasty unwanted
modulation is due to a cellular telephone nearby.
1. VCO Isolation is not the cure all
Adding a cascode
amplifier stage(s) between VCO output and final PA is always cited
as a good idea. However, you may still find a boat load of modulation
on your output signal even with these additional amplifiers in place.
In practice I have found that although cascodes and other amplifier
stages between a VCO and the antenna do provide lots of isolation
you can still end up in a world of unwanted modulation pain. I have
found in the past that I have got rid of FM modulation due to VCO
pulling to end up with AM modulation due to variation in bias on
my buffer and ampifier stages caused by the AM modulated interference
signal (GSM is one such animal). If you can run amplifiers with
a reasonable bias current the problem is eased somewhat. However,
in practice it is often the case that you want to run with just
a sniff of bias to get decent amplifier efficiency. In this case
the AM on your transistor junctions reaps havoc!
2. SAW it
It's a no brainer but if you can slap a decent filter
on the output this will gnaw off a great deal of the out of band
signal energy. Of course it doesnt help if the QRM signal is in
3. Isolators - fantastic
If you have the room
an isolator or two is definately a must have on the output of the
transmitter. I think there are now some chip parts around TOKO or
Murata I think?
5. Synthesizer loop bandwidth
running a synthesized frequency source, you can dramatically improve
unwanted VCO modulation by simply running with a wide loop bandwidth.
Of course this has trade-offs elsewhere by worth considering if
your spectrum is a comb of unwanted garbage!
6. Avoid running
your oscillator at F.
Running your oscillator at frequeny
F when you know that this is likely to have a large interference
source near or in band is not a good idea. You should consider running
your oscillator at a harmonic or sub-harmonic of F and then dividing
or multiplying to get your wanted frequency F. The divider version
of this is quite common on a few COTs radio IC's.
a good start I think!
Your thoughts on this subject would
be much appreciated.
My VCO is definately getting pulled around.
Post subject: Re: Unwanted Modulation - Nasty Stuff!Posted:
Sun Jun 14, 2009 9:44 pm
Jun 27, 2005 2:02 pm
this is a great and challenging design issue.
From my own
experience, most of the unwanted modulation problems were solved
by improving the filtering of the VCO supply voltage. A careful
selection of capacitors and ferrite beads is a first step.
In addition, there is a well-known circuit implemented by an
NPN transistor with a voltage divider in its base that drops the
supply voltage to the the voltage required by the VCO (This requires
a higher voltage source for the VCO). This circuit has shown a significant
improvement on the VCO's supply.
Other issues are careful
PCB layout and shielding.
Concerning buffer stage, this
is of course helpful but the buffers need to have a considerable
S21-S12. There are specific amplifier models for this purpose.