How much 433 MHz emitted by a typical board? Topic is solved

Technical discussions related to any XMOS development kit or reference design. Eg XK-1A, sliceKIT, etc.
Post Reply
User avatar
aclassifier
Respected Member
Posts: 483
Joined: Wed Apr 25, 2012 8:52 pm
Contact:

How much 433 MHz emitted by a typical board?

Post by aclassifier »

For the startKIT there was a chapter on Regulatory compliance:
https://www.xmos.com/download/private/s ... 1.3%29.pdf.
Not that I was so much wiser my it. My fault. Still..

..for the eXplorerKIT there is no such chapter:
https://www.xmos.com/download/private/x ... 1.2%29.pdf
Would any such exist?

What I'm after is how much EMC might these two boards emit in the 433 MHz band? Could they be expected to compete a lot with a Semtech SX1231H?

I am using those and they send at -18 dBm (weakest, strongest is +18 dBm) and receive at about -30 to -40 dBm 1-5 metres away (they would follow to well below -100 dBm). A startKIT communicates with an eXplorerKIT. Like this this but with RFM69 boards instead of WiFi boards plugged in.

I am having issues with the messages, see What is the possibility of CRC16 collisions on 20 bytes of data? (2)


--
Øyvind Teig
Trondheim (Norway)
https://www.teigfam.net/oyvind/home/
View Solution
User avatar
infiniteimprobability
XCore Legend
Posts: 1126
Joined: Thu May 27, 2010 10:08 am
Contact:

Post by infiniteimprobability »

Tough one to answer - completely depends on what the board design is and what signals are being sent out. The edge rates on xmos are pretty fast and with 3v3 swing you could pump a fair bit of EMC out if you wanted. We put series resistors (33R) on any clock out lines for this and SI reasons (reduce reflections). The chip will radiate very little in comparison to any traces.
User avatar
aclassifier
Respected Member
Posts: 483
Joined: Wed Apr 25, 2012 8:52 pm
Contact:

Post by aclassifier »

Thank you!

After much pondering I have come to the conclusion that this most probably is the Birthday problem showing up in causing CRCs being not as good as I thought. My data seem to confirm that the CRC16 has a 50% probablility of an error after 2exp(n/2) = 2exp8, ie we start to see problems after 256 messages. When I added a CRC32 then it is as good as 2exp(32/2) = 2exp16. So now I have 2exp(8+16) which should be fine for the temperatures from my aquarium...

But I may just add some 33R resistors in there to see if I see any changes in how often the radio packets get warbled. If I do I'll be back here.
Off Topic
Please answer even more often! I love that cat |/_
--
Øyvind Teig
Trondheim (Norway)
https://www.teigfam.net/oyvind/home/
User avatar
infiniteimprobability
XCore Legend
Posts: 1126
Joined: Thu May 27, 2010 10:08 am
Contact:

Post by infiniteimprobability »

The rise/fall time of a typical XMOS output is sub 1ns so quite sharp which will give you lots of odd harmonics of your original signal. Depending on PCB layout, you may also get ringing at a frequency dependent on your PCB design.

Series resistors help this and generally it's good to go as slow as you can tolerate for EMI reasons.
Post Reply