Just a general question here, it seems that remotes are the only tech that’s not evolving at the speed of everything else.
VESC is growing, becoming more feature-full, battery chem continues to improve in power density, motor offerings are expanding and there are plenty of reliable motor mount systems.
Why does it seem there are only a handful of rock solid remotes out there? The best of which are still PPM.
Is it likely that a frequency hopping DIY remote platform will ever exist with all the telemetry goodies will that we’ve come to enjoy? Something that won’t be hot one year then dead the next?
Not enough demand and too much market competition to justify investing resources in their development. There’s so many remotes available, most people don’t need something fancy, just a simple throttle control.
I was working a bluetooth remote based on the NRF52840 a while back, but decided to work on something else because there simply was no incentive to keep working on the remote.
They are, but I’m not using the Xbee3 on OSRR 1.0, it works great but cost wise is a bit out there. Nrf52840 based radio with upgraded antenna on the 1.0.
BLE 5.0 is pretty reliable, especially compared to past generations. There’s nothing really different between it and other 2.4ghz or 5ghz protocols in terms of connection reliability.
Latency. I won’t argue the connection is more robust than prior specs, but that comes at a direct cost in overhead, which presents as latency. Quite noticeable in a control loop like esk8 throttle.
Do you have numbers? The control loop runs on the ESC, all the remote does is set a target value, anything under 250ms is not too noticeable. If it is indeed more than that, then I think I know where all that damn starting lag is coming from when I play smash bros…
Yep. But- latency is always going to depend upon application environment, hardware involved + BT spec, etc.
I am saying that Bluetooth vs a lower level protocol such as ESB or 802.15.4, has considerably higher overhead simply due to it’s design/extra protocol layers. So much that Bluetooth is not even something most consider for such an application.
Have you ever watched what happens to an output that is piped via BT connection, getting hammered on link budget? As in, what happens in the real world?
A control loop does run on the ESC, however I’m talking about the total HCI loop including the human input from the remote (since we’re encompassing the latency of the connection, it’s clearly external to the ESC).
250ms is a MASSIVE amount of latency for a throttle control. I am averaging 20-50ms response time right now using an implementation over 802.15.4.
I am speaking based upon real-world testing & experience using BT2.0 and beyond + numerous other wireless protocols over the last decade.
