Shielding, connectors, filtering!
Brad
1.) The best case for shielding is a fully enclosed metal box. Would
this be equivalent to a shield on one side of the PCB with a ground
plane on the opposite side of the PCB?
2.) What is the best approach when you have a break in your metal
shielding for connectors? Should I bring the shielding all the way
around the connector? If I do this how should I route the lines?
Under the shielding with a via? Or through the shielding? Or just
keep the gap open for the connector?
3.) We had big ferrite clamps that we put on the signal cables and
power supply. I would not like to have these large clamps on the
cable and fix things inside of the device. The largest bandwidth for
our signals is 500 Hz. We were seeing problems above 30 MHz for our
emissions testing with the cables. Would it make sense to filter the
signals going out on the cables? Or is there a bead that is small that
would do the trick inside of the device?
Thanks in advance for your comments!
-Brad
Martin Griffith
amazon
Tim Williams
emc
search
press return key
buy
read
martin
sky4...@trline4.org
>Hello everyone! Thanks for your comments on my questions below!
>1.) The best case for shielding is a fully enclosed metal box. Would
>this be equivalent to a shield on one side of the PCB with a ground
>plane on the opposite side of the PCB?
For any faradays cage the holes should be less than half the wavelength.
(in microoven it's 1/64)
Any conductor in/out will work as an rf "tunnel".
>2.) What is the best approach when you have a break in your metal
>shielding for connectors? Should I bring the shielding all the way
>around the connector? If I do this how should I route the lines?
>Under the shielding with a via? Or through the shielding? Or just
>keep the gap open for the connector?
Shielding all the way I presume..
>3.) We had big ferrite clamps that we put on the signal cables and
>power supply. I would not like to have these large clamps on the
>cable and fix things inside of the device. The largest bandwidth for
>our signals is 500 Hz. We were seeing problems above 30 MHz for our
>emissions testing with the cables. Would it make sense to filter the
>signals going out on the cables? Or is there a bead that is small that
>would do the trick inside of the device?
Maybe some RLC stuff?
bill....@ieee.org
> In sci.electronics.design Brad <bradwilliams98...@yahoo.com> wrote:
>
> >Hello everyone! Thanks for your comments on my questions below!
<snip>
> >3.) We had big ferrite clamps that we put on the signal cables and
> >power supply. I would not like to have these large clamps on the
> >cable and fix things inside of the device. The largest bandwidth for
> >our signals is 500 Hz. We were seeing problems above 30 MHz for our
> >emissions testing with the cables. Would it make sense to filter the
> >signals going out on the cables? Or is there a bead that is small that
> >would do the trick inside of the device?
>
> Maybe some RLC stuff?
The "big ferrite clamps" are a quick and dirty way of making a "balun"
- otherwise a balanced transformer - which is used to block common
mode currents that would other wise circulate around the (common)
ground connection.
The clamps are big only because they have to fit around the cables -
the inductances involved aren't all that high - of the order of a
microhenry - and it is easy enough to to create comparable inductances
(and mutual inductances) on a printed circuit board with a suitable
two-hole ferrite bead. There are comparable surface mount ferrite
chips - Farnell lists a bunch of TDK and Murata SMD "suppression
chokes" while similar o parts from other manufacturers are listed in
the same section of their catalogue - section 8, EMC filters and
suppressors, as "common mode chokes" and "data line chokes".
--
Bill Sloman, Nijmegen
hrho...@att.net
You need the help and advice of a real EMC Engineer. Band-aid fixes
don't work very well and that is what you are exploring. Pay some $$
and buy a ciouple of good EMC books. Tim WIlliams' book is good, also
anything by Clayton Paul or Mark Montrose.
Bob Hofmann