Servers by jDrones

KDECAN - a new CAN protocol, first in ArduPilot

(Corrado Steri) #21

Ah ok, thanks.

Will support in arducopter start with 3.7 or maybe backported to 3.6 final that i understand is going live in the next days?


(Francisco Ferreira) #22

3.6.0 certainly won’t have it. Backporting to a later 3.6 release isn’t impossible, but requires quite a few changes.

3.7 will certainly have it.

(KDEDirect) #23

Thank you for all of the questions and to Francisco for implementing KDECAN into ArduPilot.

KDE Direct has spent thousands of hours developing our ESC line, fine tuning them to work with our motors and releasing new technology for our customers. The main reason behind the license agreement is to protect our work and prevent duplication which may cause system incompatibility and failure.

As an industry leader, the features we want in our KDE components need to be easily understood, reliable and easy for our team to modify as needed. Creating KDECAN from the ground up allowed our team to implement the features that integrate directly with our equipment. By developing and maintaining this at KDE Direct, we can provide continual updates as our ESC technology develops in the future.

For more information on KDECAN please visit our blog, Resource Center or Support Ticket System – customer service is a top priority.

(OlliW42) #24

many thx for this additional insight

I do see reasons why one would not use UAVCAN, which include but also go beyond what was indicated here, and I don’t see that v1 would change that much. I guess in this sense I do understand the decision to go with KDECAN.

However, in some alignment with some voices in the above, I find the differentiation highly unfortunate. To be specific, one CAN bus is completely occupied for “just” the ESCs, excluding other devices such as GPSes, magnetometers, barometers, power sources, gimbals, etc. on the same bus. It is one of the great strengths of CAN that it allows multiple types of nodes on one bus, which appears to have been broken here.

Unless KDE plans to introduce their own set of such units, and to develop further their protocol (which currently is more of a set of messages than a “real” protocol) towards supporting such devices, in which case it all would make much more sense.

I would find it highly interesting to hear your, KDE’s, vision as regards non-ESC devices on a KDECAN bus.

(Rick) #25


I am using KDE on my industrial drones since 3 years, they have the best support team before and after sales.

I am sure they will continue with the same in the CAN subject.

Great work KDE.