EMC from AVR Mega32

Is it the micro or the crystal ?

Have you got enough decoupling capacitors  and  are they low esr caps?
On Power supply and all io lines .

Shield the pcb and or use a metal case  or complete mesh cage?

How are your ground planes on the pcb?
And gnd plane under/around the xstal/osc ?

What about trying a different crystal ?
or different capacitors connected to the crystal.

Do you need a 16MHz crystal ?
Can you use a lower frequency like 4 or  8MHz instead of 16MHz
and double/4x the frequency in the rf section ?

What about the ground / earth connection ?
How are you powering the board and how are you grounding it ?

Have you had a read of the  overview of EMC app note from Atmel ?
AVR040: EMC Design Considerations
http://www.atmel.com/dyn/resources/prod_documents/doc1619.pdf




Alex

Different crystal frequency, although it may not  help much, as your
receiver is already desentisized by signal 3 MHz away.

If you are sure the the signal get in at the receiver antenna  (and not via
the powersupply or the ground) A standard 14,.. MHz crystal get the 2nd and
3rd harmonic 7 MHz away

32 MHz trap at the receiver input (serie resonance, parallel to the input or
a parallel resonance in series with the input.

Additional selective filter at receiver  input.

Wim

Hi,

Try running with the on chip RC crystal, or try turning off the full swing
option (Fuse bit P - osc. option).

Have you tried using a different crystal from a different manufacturer, our
try grounding the case of the crystal.


Regards
Ian Dobson

Home of the Atmel based UDP mobile web cam http://www.planet-ian.com
All mails scanned with av-filter.pl (F-Prot / perl)

 >> I'm using an AVR Mega32 in the DIP40 package to interface with 35 MHz
 >> remote control radio transmitters and have found that there is an
 >> enormous radiated emission from the AVR at the second harmonic of the
 >> 16 MHz crystal.  According to my spectrum analyzer the 32 MHz emission
 >> is 40 dB at 25 cm, while the output of the transmiter at 1 m with the
 >> antenna down is only 45 dB.
 >>
 >> This noise is reducing the effective range of the remote control from
 >> nearly 1 km to less than 100 m before the receiver can no longer hear
 >> the transmitter.  I've verified that the AVR alone is causing the
 >> problem; on a breadboard with just the microcontroller, decoupling caps
 >> and crystal the noise levels are identical.  There is no radiated noise
 >> at 72 MHz, so it works fine with the US transmitters.
 >> I've tried a variety of replacement chips, chokes, caps and shields to
 >> no avail.   I even tried a Mega128 in TQFP packaging and found a
 >> similar amount of noise.  Does anyone have any further ideas?
 >>
 > Different crystal frequency, although it may not  help much, as your
 > receiver is already desentisized by signal 3 MHz away.
 > If you are sure the the signal get in at the receiver antenna  (and not
 > via
 > the powersupply or the ground) A standard 14,.. MHz crystal get the 2nd
 > and
 > 3rd harmonic 7 MHz away
 >
 > 32 MHz trap at the receiver input (serie resonance, parallel to the input
 > or
 > a parallel resonance in series with the input.
 > Additional selective filter at receiver  input.
 >
 > Wim
 >