CE compliance testing

[Nico Coesel]

Stef <stef33d@yahooI-N-V-A-L-I-D.com.invalid> wrote:

In comp.arch.fpga,
Morten Leikvoll <mleikvol@yahoo.nospam> wrote:
"Fredxx" <fredxx@spam.com> wrote in message
news:i4jh01$ukp$1@news.eternal-september.org...
This is for a niche market, though any market has competitors who can play
informer to the authorities.

I would not worry (I used to worry too). The authorities has a lot more
important work to do than chasing low qty products that has a tiny bit more
radiation than the papers allow.

They may not actively chase for such products, but when a complaint is
made (by, for example, that competitor) they will have to follow that up.
And although it may not happen very often, I have seen it happen, twice.

And how do you know you're only 'a tiny bit' over the limits without
actually measuring your device? Take a simple oscillator chip and
connect it to the 'wrong' trace, emissions will go through the roof.

There is no requirement to actually do measurements, you just have to be
under the limits. How you make sure you are, is up to you. If you can
calculate and simulate everything, that may be enough proof you honestly
believe that your device is under the limits. I have not seen that happen,
just measuring is usually simpler.

And even good measurements are not a complete guarantee your device is
always under the limits. These measurements are just scans through the
required frequency range. And if your device produces short burst, they
may be missed in those scans. But at least you have evidence that you
did all you could to make sure your device is under the limits.

The tests are well specified. The measured values are averages, not
peak values. It often helps to use a spread-spectrum oscillator so the
peak is smeared out over a wider band.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

 

[Stef]

In comp.arch.fpga,
Nico Coesel <nico@puntnl.niks> wrote:

Stef <stef33d@yahooI-N-V-A-L-I-D.com.invalid> wrote:

And even good measurements are not a complete guarantee your device is
always under the limits. These measurements are just scans through the
required frequency range. And if your device produces short burst, they
may be missed in those scans. But at least you have evidence that you
did all you could to make sure your device is under the limits.

The tests are well specified. The measured values are averages, not
peak values. It often helps to use a spread-spectrum oscillator so the
peak is smeared out over a wider band.

No, not averages, but quasi-peak.

But that measurement is a scan of the required frequency range, a
measurement receiver can not monitor the entire band simultaniously.
So if your device emits bursts of, lets say, 0.1 seconds every few
seconds, there is a possibility they will be missed in the measurement.
And next time you measure that same device, it may just pop up as an
unpleasant surprise.

The (well known) place where we do most of our measurements actually
does a very quick (under half an hour) peak scan and after that just
measures the ten or so highest peaks with the required quasi-peak
detector. Doing the entire scan with quasi-peak would take hours.


--
Stef (remove caps, dashes and .invalid from e-mail address to reply by mail)

Peace is much more precious than a piece of land... let there be no more wars.
-- Mohammed Anwar Sadat, 1918-1981

 

[Leon]

On 18 Aug, 17:11, "Fredxx" <fre...@spam.com> wrote:

I have a small electronic unit which needs some CE compliance testing.

Cost is a major issue and was wondering if anyone here can offer advice on
an affordable solution.

Where I used to work they hire the testing facility for half a day,
and do their own initial testing. They then fix any problems and the
product usually passes the formal tests without any problems.

 

[Michael Schwingen]

On 2010-08-19, Symon <symon_brewer@hotmail.com> wrote:

I have no idea if this is appropriate to your situation, but you might
avoid a lot of the CE compliance issues by using a wall-wart or somesuch
which is already CE marked. Then all the mains powered stuff is
pre-certified by someone else. Get one with a ferrite thingy near the DC
jack!

For Safety, Yes - EMC, not necessarily. I just had a case where a product
with a wall-wart was above the allowed conducted emissions on the mains,
with the cause being in the product, not the wall-wart. Adding common-mode
filtering fixed that, so a ferrite in the DC cable might have been
sufficient.

cu
Michael