On 19 Apr., 04:32, "hrhofm...@[EMAIL PROTECTED]
" <hrhofm...@[EMAIL PROTECTED]
> wrote:
> On Apr 18, 3:09=A0am, default <edmoneyfromlanailstu...@[EMAIL PROTECTED]
>
> wrote:
>
> > I am writting at my diploma thesis with an EMC theme.
>
> > If got a power fan (13V / 6A) witch produces conducted emissions, so I
> > have to reduce them to achieve CISPR25.
> > Up to 40MHz I found an easy solving with LC lowpass filter, but over
> > 50MHz I am out of the CISPR25 limit.
>
> > Does anybody have experiences in EMC design in high frequency bands,
> > doesn't have?
>
> > I think about solving it like the motherboard producers do, they
> > design there signals like little inductors and so one.
> > But I didn't find any guidlines to do this like them.
>
> Are you saying that the lowpass filter is not blocking the high-
> frequency noise components above 45 MHz, but is blocking the mid-range
> frequency components below 45 MHz, and so you need a more effective
> filter against the really high-frequency noise from 45 MHz and
> higher? =A0
yes that's my problem, downsides 45MHz it's okay
Some ferrite beads on the power leads along with shunt
> capacitance between the leads and between each lead on a good solid
> ground should form an effective high-frequency filter. =A0But lead
> lengths are critical at these frequencies, beware of any leads longer
> than 1 inch, these look as much like inductors as real inductors.
i tried two X-C's form plus load to case of the fan and one from gnd
to case, it works fine but the ****blem is that i can't do it.
according to the opnion of my boss, because maybe there are new fans
designed in future that doesn't have a metallic case.
> How many dB are you over the limits, 10, 100???
the worst case level is about 65MHz with 56dB=B5V and the spec is about
30dB=B5V, i am circa 26dB=B5V above the debit level
|
