MultiNav Pro+ BLE - max simultaneous connections vs range, any real limits?

@Charles Hall yeah this is a real trade-off and you're right to think about it carefully.

 

The STM32WB07 at the core of the MultiNav Pro+ does support multiple simultaneous BLE connections, but the radio is fundamentally time-division — it schedules connection events sequentially, so the more peripherals you add, the longer the interval between each individual link's connection events. At extreme range you're already working with weak signal margins, and if the connection interval stretches out too far, you start seeing link instability before you ever hit a clean disconnect. So yes, more connections does effectively trade against reliable long-range performance.

 

The 50km figure is a drone-to-drone spec with both modules elevated and no ground reflections — realistically for a star topology at that scale with 8-12 nodes, I'd expect your effective reliable range to compress meaningfully depending on how you configure connection intervals and what data rate you actually need per link. The GitHub repo has the firmware source so you can dig into the connection scheduling parameters directly — worth tuning connection intervals aggressively and seeing where your specific hardware layout lands before assuming worst case.

 

Honestly the cleanest solution for what you're describing might be a mesh topology rather than a pure star, so the range burden is distributed across hops rather than all sitting on one central radio — the Connect app already supports mesh chat, which means the underlying BLE layer can handle it. Worth a conversation with the team about whether the SDK exposes enough to build that into your application layer.

From what I've tested with the MultiNav Pro+, yeah there's definitely a trade-off - the more concurrent connections you're juggling, the tighter your connection intervals get shared around, and at the edge of range that really does hurt link stability. I wouldn't push past 6-8 nodes if you're trying to hold anything close to max range, otherwise you'll start seeing dropped connections and retries that make the whole thing flaky. Might be worth experimenting with connection interval and supervision timeout settings to squeeze a bit more out of it depending on your data rate needs.

Agreed with Alexandre on the interval thing - also worth looking at whether all 12 nodes actually need to be *simultaneously* connected or if you can get away with a hub-and-spoke approach where some nodes only connect on demand, that'll free up a ton of radio time for the ones at the edge of range that really need the reliability. I've had decent luck doing something similar on a site monitoring build, kept the persistent connections to 5-6 critical nodes and just polled the others when needed.

Jumping in late but just wanted to add - if you're really trying to push range on those edge nodes, dropping your TX power on the closer ones can help a lot since the radio isn't burning cycles on retransmits for nodes that don't need it, leaving cleaner airtime for the far ones. Basically just let the close nodes run a bit "noisier" connection-wise and prioritize the link budget where it counts.

All solid advice above - one thing I'd add is that if you do go the on-demand polling route Joseph mentioned, make sure your supervision timeouts on the persistent connections are tuned conservatively, because I've seen setups where a burst of reconnect activity from the polled nodes causes enough interference to trip the timeouts on the "always-on" links and you end up with a cascade of drops. Took me way too long to debug that one lol.