Any Digi SX experts here?
I want to setup a radio network on the farm to distribute RTCM RTK correction data from my base station (100' tower) to the tractors/rovers. I have one SX PRO (XBP9X-DMUS-001) for the base transmitter and I'm planning to use SX non-pro (XB9X-DMUS-001) for the rovers as the data only needs to go from base (xmit only) to rovers (recv only). Based on my simple range test, I might want to setup repeaters at additional locations to get better coverage, most additional locations are already connected via PtP IP network radios so the repeaters wouldn't have to repeat the primary base's RF data but they should be within range to do so which would simplify the repeater installations (no additional ethernet-serial bridges needed).
Is this a Mesh network that I'm describing or a PtMP with repeaters? There might be some rare cases when it would be beneficial for a rover radio to mesh/repeat the data to another rover in a blind spot.
I'm also not sure how important it is to block the rovers from accidentally sending data to the RF network and causing RTCM stream corruption on the other rovers. I have started by setting my rovers' DL (Destination Address Low) to 0 so that all other devices ignore it but depending on how you recommend I setup my Digi RF network, rovers may need to transmit if they're in a mesh or repeater mode? Or does the DL parameter only apply to data coming in via serial and not other behind the scenes RF magic?
I tried to extend my range by lowering the BR (RF Data Rate) from 110 kbps to 10 kbps and setting MT (Broadcast Multi-Transmits) from 3 to 0 but I think that's too slow to handle the RTCM data stream (about 500 bytes/s) as I can see that not all RTCM packet types get sent. I also tried to lower my base stations RTCM output to less then 1hz but that only affects 4 of the 8 RTCM packets, the rest still go out at 1hz. It might still work if I could implement hardware flow control because I suspect my base station outputs the RTCM in bursts (at 1hz) which causes buffer overflow in the SX PRO but I've increased BR back to 110 kbps for now. Currently I'm feeding my main SX PRO's serial input via a USR-TCP232-T2 (ethernet-serial) device on the top of my main base station tower. The TCP232-T2 does have a decent buffer (4 KB) on the network receiver so I've wondered if I could just slow down the serial baud to better match the lower RF Data Rate setting, I'm not sure how to match those rates.
Another question regarding the remotely located nature of my base station SX PRO, I have not figured out how to make changes to my SX PRO's configuration remotely without having to climb the tower with my laptop and an XBee USB adapter. Is there a possibility of using an ethernet-serial server/adapter to make config changes with XCTU or would it be better to setup the RF network so that I can use a rover SX radio to remotely config the base SX PRO?
When I mentioned my simple range test, this is what I did. My base station SX PRO is set to output full 30 dbm, it has a 16' LMR195 cable to a high gain (4' long) 900 mhz omni 110' above ground. I used a magnet mount 11 dbi antenna on the roof of my truck, with the SX plugged into the XBee header on an arduSimple simpleRTK2B F9P gps receiver while I drove around watching the GPS data with Ublox's u-Center software on my laptop. Watching the GGA NMEA sentance I could see how old the differential correction data (RTCM stream) was. Out in the open, I received a decent rate of corrections (at BR 110 kbps) at a distance of 2.5 - 3.5 miles. At 1.5 miles trees 1000' away were already an issue. Compared to our old AutoFarm system (20 yrs old?) which uses Freewave FGR09 radios (also 1W output), they achieve good operation at 5 miles in the open and 3.3 miles still works when on the dark side of a treed yard.
So I've spent a enough time on this already to know I need help from someone with better knowledge of how the SX radios work. If you'd rather communicate verbally instead of typing, please let me know. Doesn't look like I can attach pictures here, otherwise I'd post some Google Earth screenshots of our locations, but most of our land is all within a 5 mile dia circle with multi yards all spread out.