|
Home > Archive > Electrical code Compliance > April 2007 > How to reduce radiated emissions
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
How to reduce radiated emissions
|
|
| raki1998@gmail.com 2007-03-29, 9:25 am |
| Hi everyone,
we have developed a telematics prototype pcb which has to meet the
class B requirements of EU.It has many cables(like
RS232,USB,DB15,Audio jacks) connected to it.
But certain frequencies emitted by the pcb are very high.
I have tried options like shielding,enclosing with metallic
box,ferrite cores on the cables etc.. but in vain.
Can anyone please suggest the effective ways of reducing these
radiations????
| |
| Maurice Janssen 2007-03-29, 1:25 pm |
| On 29 Mar 2007 02:24:05 -0700, raki1998@gmail.com wrote:
>Hi everyone,
>we have developed a telematics prototype pcb which has to meet the
>class B requirements of EU.It has many cables(like
>RS232,USB,DB15,Audio jacks) connected to it.
>But certain frequencies emitted by the pcb are very high.
>I have tried options like shielding,enclosing with metallic
>box,ferrite cores on the cables etc.. but in vain.
>Can anyone please suggest the effective ways of reducing these
>radiations????
Fixing a design with ferrite and shielding is often quite difficult.
There's no easy answer that can reduce the emission further when you've
taken the obvious steps.
It's better to take measures in the original design (separation of HF
and LF parts on the pcb, careful routing of data buses, location of
external interfaces, using Vcc/GND planes, decoupling of Vcc near IC's,
etc.), but that's a bit late in this case.
Sorry that I can't give a useful answer.
--
Maurice
| |
| Paul E. Bennett 2007-03-29, 5:25 pm |
| raki1998@gmail.com wrote:
> Hi everyone,
> we have developed a telematics prototype pcb which has to meet the
> class B requirements of EU.It has many cables(like
> RS232,USB,DB15,Audio jacks) connected to it.
> But certain frequencies emitted by the pcb are very high.
> I have tried options like shielding,enclosing with metallic
> box,ferrite cores on the cables etc.. but in vain.
> Can anyone please suggest the effective ways of reducing these
> radiations????
It is useless trying to retro-fix a poor design when you have such issues to
worry about. If your current design does not meet the EU requirements when
you have tried a range of quick fixes then you be better off taking a bite
of the bullet and re-design the product.
When you have high frequency oscillators on the board then you should start
thinking of a boxes within boxes approach to the design of the product.
Some of the circuits will be quite clean in operation and will cause you
very little trouble. Other circuits (usually those containing square-wave
oscillators) radiate and conduct a large amount of noise. These need proper
decoupling of the supply lines, in-line on-board filtering and maybe even
some metallic shielding around the circuit.
There are some good guides on circuit board layout and noise reduction
techniques. I will leave it to others more into doing the PCB layout
currently to advise on the best of the current texts available and maybe
guide you further in the techniques.
--
********************************************************************
Paul E. Bennett ....................<email://peb@amleth.demon.co.uk>
Forth based HIDECS Consultancy .....<http://www.amleth.demon.co.uk/>
Mob: +44 (0)7811-639972
Tel: +44 (0)1235-811095
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
********************************************************************
| |
| Peter_X@nospam.com 2007-04-03, 9:25 am |
|
raki1998@gmail.com wrote:
>Hi everyone,
>we have developed a telematics prototype pcb which has to meet the
>class B requirements of EU.It has many cables(like
>RS232,USB,DB15,Audio jacks) connected to it.
>But certain frequencies emitted by the pcb are very high.
>I have tried options like shielding,enclosing with metallic
>box,ferrite cores on the cables etc.. but in vain.
>Can anyone please suggest the effective ways of reducing these
>radiations????
The answer depends on where the radiation is coming from.
You can get a feel for it by slowing down the edges of the main clock
signals, by hanging a capacitor (say 10pF) on their sources one at a
time, and watching the emission spectrum as you do it. This is very
quick to do.
Often, an RC lowpass filter on a clock line is enough to get you in
spec - and that is all that's needed; you don't get a Queens Award for
doing better!
Then you should find that other emissions fall off also.
Ferrite connectors are expensive and to be honest don't do a great
deal. They are used as a last resort when a product is finished, you
have orders, you have to send some out, you have 2 weeks to fix it,
and you aren't worried about another $100 spent on filter connectors
Same comment for shielding. You can spray nickel paint inside the box,
or even stick some metallised self adhesive paper inside (a LOT of
consumer electronics in plastic cases does cheap stuff like that). But
proper metallisation of plastic mouldings is expensive, and doesn't do
a great deal.
The #1 thing is to use a 4-layer PCB with a ground plane, and a VCC
plane. That alone does far more than anything else you can do. If your
PCB is 2-layer, just sticking the extra layers in there is really easy
and nowadays (Chinese manufactured PCBs - far better quality than UK
ones) the extra cost is hardly significant.
SO:
1) Identify the worst radiators and reduce the slew rate on them, with
an RC filter close to the driver(s).
2) Use a 4-layer PCB
3) Use slower logic if possible e.g. 74HC instead of fast stuff. The
old 74AC logic was "impossible" for getting approvals! The PCB tracks
radiated like hell.
4) Implement the various bodges like filtered connectors, enclosure
shielding
| |
| Peter_X@nospam.com 2007-04-03, 9:25 am |
|
"Paul E. Bennett" <peb@amleth.demon.co.uk> wrote:
>It is useless trying to retro-fix a poor design when you have such issues to
>worry about. If your current design does not meet the EU requirements when
>you have tried a range of quick fixes then you be better off taking a bite
>of the bullet and re-design the product.
I take it you are a consultant
| |
| Paul E. Bennett 2007-04-03, 9:25 am |
| Peter_X@nospam.com wrote:
>
> "Paul E. Bennett" <peb@amleth.demon.co.uk> wrote:
>
>
> I take it you are a consultant
I thought that fact was obvious from my signature. Although my consultancy
is not specifically in EMC or the RF behaveour of products, it is an area
that I have to take into consideration in the systems I have to produce.
There are others here, like John Woodgate et. al., who might be able to be
more specific for their usual fee of course.
--
********************************************************************
Paul E. Bennett ....................<email://peb@amleth.demon.co.uk>
Forth based HIDECS Consultancy .....<http://www.amleth.demon.co.uk/>
Mob: +44 (0)7811-639972
Tel: +44 (0)1235-811095
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
********************************************************************
| |
| hrhofmann@att.net 2007-04-04, 9:25 pm |
| On Mar 29, 4:24 am, raki1...@gmail.com wrote:
> Hi everyone,
> we have developed a telematics prototype pcb which has to meet the
> class B requirements of EU.It has many cables(like
> RS232,USB,DB15,Audio jacks) connected to it.
> But certain frequencies emitted by the pcb are very high.
> I have tried options like shielding,enclosing with metallic
> box,ferrite cores on the cables etc.. but in vain.
> Can anyone please suggest the effective ways of reducing these
> radiations????
Are you working with an EMC test house to determine that your
emissions are over the limits?? If you are, the test house will
almost certainly be able to offer advice on what they think will
help. It will cost $$, but your time and delays in getting the
product to market cost money too.
Careful board layout, partitioning of functions and signals, attention
to ground planes are all important factors in design. With 30+ years
in EMC testing and trouble-shooting, I have seen almost everything
that can go wrong be wrong. Band-aid fixes like shielding and
ferrites help somewhat, but good design always wins.
You don't give any information on signal levels vs limits, problem
frequencies, etc, so it is not possible to make any specific
recommendations.
H. R.(Bob) Hofmann
|
|
|
|
|