On the subject of EMI, I just assisted a client in improving his EMI
susceptability. He had setup a homemade sheilding wall, which was handy, and
had a 5W handheld radio. We found that at some mysterious angles and
distances, the temperature readings went nuts, then the processor packed up and
qhit scanning the LED displays. When we replaced the temp inputs with solid
voltages, we got the same thing. We then noticed that the power cord position
was important. Bingo! we had our man. When we filtered the supply going into
the power supply, almost all of the problems disappeared. We now have to get
within 4 inches with the 5 watts to make an error, and cannot get the processor
to pack up. In the end analysis, the 60Hz transformer offered no impedance or
attenuation to RF. It simply walked through the supply and crashed the works.
We even found that when the scope was attached, we didn't see anything fishy.
Well, of course not. How much deflection will you get on an old 40MHz scope
with 150MHz signal? The DC component looks an awful lot like the 5 volt line.
I was personally suprised that we got as far as we did with tin foil and a
radio shack radio. Even the filters that we used to check the theory were
radio shack.
Chris Eddy
Pioneer Microsystems, Inc.
pedraki wrote:
{Quote hidden}> Does anybody have some information (or whrere to find it) about
> electromagnetic inmunity of various microcontrolers in the market. What
> microcontrolers are using in critical applications like aerospatial (What
> are the microcontrolers used by the NASA ?, for example), dangerous
> control, etc.
>
> It is obvius that the systmes must be shielded again interferences but i
> think that when the shield fail it is best to use some microcontrolers than
> others.
>
> Thanks.
