From: Ray Anderson ([email protected])
Date: Thu Mar 02 2000 - 11:12:22 PST
sweir wrote:
> The usual good intention of this is to keep noise from the planes out of
> the PLL, not the other way around. The fear is that the VCO will get
> thrown off by PS noise. It can be more dangerous to use an inductor,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> unless the filter is properly damped, than to use either a resistor or
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> direct connection. In either case, high quality bypass from the PLL VCC to
> the PLL ground is very important. This is also one of the places where
> judicious moating can be very useful.
>
Good point. I was involved in fixing a design that someone messed up
where they used one of the little integrated ferrite low pass filters
to feed the PLL VCC pin on an ASIC.
As it turns out, if you look at the attenuation curve in the vendor's
data book, the low pass characteristic appears well behaved and as
you might expect. The gotcha' is that the parts are characterized in
a 50 ohm system. Well if you are feeding the filter from a power plane
then the driving impedance is probably in the sub-ohm region and the
load presented by the PLL is in the couple hundred ohm region. Grossly
mismatching the little lpf part like that causes a real peaky response.
The one I saw had a peak of around 20 dB. The end result was that there
actually was more noise on the PLL VCC pin coming from the power plane
than there was on the plane. The filter didn't attenuate the noise but
it enhanced it.
I've seen this effect also with badly designed discrete LC filters used
in the same sort of applications. The cure is the judicious insertion of
a small bit of resistance to damp the response. You need enough to
tame the peaking but not enough to mess up DC conditions.
-Ray
**** To unsubscribe from si-list or si-list-digest: send e-mail to [email protected]. In the BODY of message put: UNSUBSCRIBE si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****
This archive was generated by hypermail 2b29 : Thu Apr 20 2000 - 11:35:16 PDT