Hi to all,
I changed the RFM12Pi on my raspberry version B to the new RFM69Pi.
I thought that that change would have improved performance even though my emontx v3 still use the RFM12.
Actually it seems that the RFM69Pi misses about 7-8 % of samples. The receiver is about 4 metres from emontx that sends frames every 10 seconds and there is a thin wall in between.
With minicom I have the following output:
Current configuration:
O i15 g210 @ 433 MHz q1
OK 10 0 1 0 0 191 0 4 0 122 89 189 6 173 1 118 138 (-52)
OK 10 246 0 0 0 182 0 4 0 123 89 189 6 163 1 119 138 (-48)
OK 20 26 7 (-52)
OK 10 226 0 0 0 162 0 4 0 125 89 189 6 145 1 120 138 (-53)
OK 10 223 0 0 0 159 0 4 0 126 89 189 6 142 1 121 138 (-49)
OK 10 240 0 0 0 176 0 4 0 123 89 189 6 158 1 123 138 (-47)
OK 10 248 0 0 0 184 0 4 0 105 89 189 6 165 1 124 138 (-47)
OK 10 236 0 0 0 172 0 4 0 109 89 189 6 155 1 125 138 (-48)
OK 20 26 7 (-52)
OK 10 20 1 0 0 162 0 55 0 101 89 189 6 198 1 127 138 (-48)
OK 10 30 1 0 0 166 0 59 0 107 89 189 6 207 1 128 138 (-48)
OK 10 39 1 0 0 176 0 60 0 122 89 189 6 217 1 129 138 (-47)
OK 10 49 1 0 0 185 0 60 0 120 89 189 6 227 1 130 138 (-47)
The 16th column is a counter, as you can see, in this example, there are two missing, 122 and 126.
Where are they gone ?
P.S.
With RFM12Pi I had less than 1% of missed frames in the same conditions.
[There is no such thing as a RFW69Pi - title changed. Moderator - RW]
Re: Lost samples passing from RFM12Pi to RFM69Pi
I didn't solve that problem but, maybe, I've found its origin.
I put quiet= false and I discovered that the frequency 433 is extremely noisy at my house, there are lots of unknown samples discardeded by the receiver. The RFM69Pi has a higher sensitivity and receives more noise than RFM12Pi, when the unwanted messages overlaps the good one, the good one becomes corrupted.
Now I have to investigate where that noise comes from.
Re: Lost samples passing from RFM12Pi to RFM69Pi
Clearly, removing the noise / interfering transmissions at source is the best solution.
Alternatively, the RFM69 has some options to control the sensitivity and the AGC action, which might be helpful. Take a careful look at the data sheet.