Fwd: RE: [En-Nut-Discussion] SRAM speed

[Harald Kipp]

>Date: Mon, 16 Sep 2002 11:03:15 +0200
>To: <ajeremy at dotaussie.com.au>
>From: Harald Kipp <harald.kipp at egnite.de>
>Subject: RE: [En-Nut-Discussion] SRAM speed
>
>Alastair,
>
>It doesn't help really to get a CE mark for the
>Ethernut Board. As soon as you combine it with
>add-on boards, put it in a box, using longer
>cables etc.etc., you have to do a new test.
>
>How to design for this requirements? Well...
>mhhh... never counted the number of books and
>articles I read about. Nevertheless, I still
>never managed to leave the test field with a
>CE mark on the first go. But things have
>dramatically improved after we got our own
>test equipment for pre-compilance tests.
>
>Here's how we do it now:
>
>1. Building a first prototype fast, adding
>capabilities for cutting specific traces.
>2. Checking 3 to 5 prototypes with near field
>probes to find problem areas.
>3. Using more time to design a second prototype.
>4. Again checking with our equipment and, if
>it looks good, going to the test chamber.
>
>Steps 3 and 4 may be repeated, of course. From
>my short time experience I can say, that problem
>areas are typically not where you expect them to
>be. For example the crystal capacitors on the
>Ehternut board were not properly placed, as I
>found out later. But we never had any problem
>with these. Really interesting, that the whole
>Ethernet Area has a very low radiation. But the
>data bus is very noisy.
>
>Harald