The above picture shows the Project Andromeda ground station circuitry. You can also see the DNT900 from RFM. I believe there has been some interest on this radio in the DIY Drones community and once I finish the construction of the autopilot I will be able to give a recount of my experiences with it. So far, however the radio seems to be extremely capable (especially compared to the Aerocomm AC4790 radios that we previously used).
You can find more information related to the ground station and the DNT900 in the full article available at: http://www.projectandromeda.com.au/blog.
Comments
My autopilot uses variable package sizes to send different things downstream at different rates. For example, voltage, GPS location and other low update rate values are sent at 5HZ while aircraft attitude and altitude information is sent at about 20Hz. Waypoints, flight control commands and other information is also sent upstream from the ground station.
I think for a one way application the AC4790 will do fine, but the hop duration is fixed at 50ms. That means that you can add 50ms of delay straight away. If your packet misses the end of the hop, add another 50. If you read the blog, you can see I've set the hop duration of the DNT900 to 4.7ms. In that duration the radio can reliably send 64 bytes. That is more than a 10 fold improvement on the AC4790.
There's also a lot more in the datasheet if you have the time to browse through it.
looking forward to see more details , meanwhile can you tell us what type of comparison and rusults did you get with AC4790 and why did you change to RFM 900? I am using AC4790 but haven't pushed it to limits yet. Thanks for sharing your experience.