-Opprinnelig melding-
Fra: Wan Juang Foo [mailto:f...@np.edu.sg]
Sendt: 20. april 2002 06:30
Til: am...@westin-emission.no
Kopi: ieee pstc list
Emne: Re: Decoupling - capacitor values
Amund,
Cortland may be right, a chamber 'may not' be needed, high ambient
considerations to be put aside for the moment
-Opprinnelig melding-
Fra: Wan Juang Foo [mailto:f...@np.edu.sg]
Sendt: 20. april 2002 06:30
Til: am...@westin-emission.no
Kopi: ieee pstc list
Emne: Re: Decoupling - capacitor values
Amund,
Cortland may be right, a chamber 'may not' be needed, high ambient
considerations to be put aside for the moment
Kopi: ieee pstc list
Emne: Re: Decoupling - capacitor values
Amund,
Cortland may be right, a chamber 'may not' be needed, high ambient
considerations to be put aside for the moment, even if the emission is
measured to be on the 24dBuV/m @ 3m, freq.band 155MHz-165MHz. I am just
curious, what
Amund,
Cortland may be right, a chamber 'may not' be needed, high ambient
considerations to be put aside for the moment, even if the emission is
measured to be on the 24dBuV/m @ 3m, freq.band 155MHz-165MHz. I am just
curious, what are the limits (or standards) are you trying to meet? From
Dear Bob
Just a point - smaller is not always better.
AVX and maybe some others make 0612 capacitors for decoupling. They are the
same size as the normal 1206 size but are metallised along their long edges.
This reduces their ESL considerably, to the extent that (according to
manufacturer's
I have found removal of inductance at power pins often REDUCES emissions.
When this is so, it is because designers who put it in did not insure there
was sufficient local capacitance to hold up Vcc while the isolated device
switches. This has happened to me often enough that when I see series
to
accepted theory.
Richard Woods
Sensormatic Electronics
Tyco International
-Original Message-
From: Robert Wilson [mailto:robert_wil...@tirsys.com]
Sent: Thursday, April 18, 2002 11:57 AM
To: emc-p...@majordomo.ieee.org
Subject: RE: Decoupling - capacitor values
Not sure this would
...@tycoint.com [mailto:richwo...@tycoint.com]
Sent: April 18, 2002 4:57 AM
To: emc-p...@majordomo.ieee.org
Subject: RE: Decoupling - capacitor values
Consider adding a ferrite bead in the 5V trace to the microprocessor.
Richard Woods
Sensormatic Electronics
Tyco International
-Original Message
decoupling I assume that reducing loop area
is the most important.
Amund
-Opprinnelig melding-
Fra: Cortland Richmond [mailto:72146@compuserve.com]
Sendt: 18. april 2002 00:54
Til: am...@westin-emission.no; ieee pstc list
Emne: Re: Decoupling - capacitor values
Yes, it makes sense
Hmmm.
I've read a couple of other replies on this...good suggestions. I have
also read your re-replies which suggest that you have a two layer
planeless board.
I understand that you are trying to get an unwanted 156Mhz off of the 5V
line. A couple of thoughts come to mind.
First 156Mhz
Consider adding a ferrite bead in the 5V trace to the microprocessor.
Richard Woods
Sensormatic Electronics
Tyco International
-Original Message-
From: am...@westin-emission.no [mailto:am...@westin-emission.no]
Sent: Wednesday, April 17, 2002 4:50 PM
To: emc-p...@majordomo.ieee.org
Amund,
In the absence of any layout information that can evaluate the loop
inductances, I suggest you read what I wrote about placing two capacitor in
parallel which can be found on Chapter 3 of my thesis. A link is available
in:
http://www.geocities.com/timfoo6143/index.html
BTW, how about
I read in !emc-pstc that Robert Wilson robert_wil...@tirsys.com wrote
(in 3FF57405336C9B4C976A1819F860A2560F696F@xng_tirsys.TIRSYS.COM)
about 'Decoupling - capacitor values', on Wed, 17 Apr 2002:
The main
reason is that the ESR of the larger cap begins to rise to unacceptable
levels as frequency
Capacitors can continue to function quite well above their self
resonance. You should not be worrying about choosing a cap based on self
resonance per se.
However, it is standard practice to parallel a 0,1 uF bulk bypass cap
with a much smaller NP0 or C0G type (say 470 pF or 1 nF), especially
Yes, it makes sense. But the goal here is preventing or reducing Vcc drop
during the time the microprocessor is switching. You need not only low
reactance, but *also* enough capacitance to supply the current needed
_while it is switching_. You have not given enough information here to tell
if
15 matches
Mail list logo
