Re: [Paparazzi-devel] CANbus driver

[Felix Ruess]

Hi guys,

great to hear you are working on CAN support!

Just two quick notes:

* I hope you are are basing your work on the master branch (with libopencm3) and not one of the stable branches...

* Probably not of much use, but we there was the CanServoController a while ago, but we removed it again because it was not ported and used anymore.

 See https://github.com/paparazzi/paparazzi/commit/399016c983b338242bac04d668a3970d5f1f94b1

Cheers, Felix

On Wed, Jan 30, 2013 at 12:52 PM, Karoly Molnar <address@hidden> wrote:

Miguel, what is your planned bus topology? With one ESC in a fixedwing configuration it is not a major issue. However you may know that passive star topology on CAN, which could be an apparent choice for a quadcopter, is not the best choice and very sensitive to proper termination. So in any case you have a better choice by connecting the nodes one after each other on the bus with proper termination on both end.

Just a hint to the setup: remove the LPCExpresso and keep only the LISAM and the monitor node on the network. Then try to transmit a frame with the monitoring tool. If the bus speed is correctly set up and the driver is initialized in LISAM then at least you shall not see error frames on the bus. While in case there is a bus speed mismatch or improper config then you would see lots of un-acknowledged frames (also visible by oscilloscope).

Regards

Karoly

> Date: Wed, 30 Jan 2013 09:49:49 +0000

> From: Miguel Torres <address@hidden>

> To: address@hidden
> Subject: Re: [Paparazzi-devel] CANbus driver

> Message-ID: <address@hidden>
> Content-Type: text/plain; charset=ISO-8859-1; DelSp="Yes";
> format="flowed"

>
> I'm starting to work with CAN bus, so right now I'm just playing with
> bit timing and sending dummy messages to test all nodes. I have 3
> nodes connected: one Lisa/M, one LPCXpresso board (LPC11C24 MCU with
> integrated can controller and transceiver) and one microchip can bus
> analyzer. I will add the ESC32 once I got the can transceiver mounted
> on it.
>
> To start with, i would like to have the existing driver code up and
> running. It looks like a simple driver but as far as it seems it
> should work. Once the basic communications are working I will focus on
> the communications protocol and messages. I'm not sure if it's a good
> idea to continue with the actual driver or if it's better to redesign
> a new one. At least the CAN bus timing settings should be configurable
> by the user.
>
> Regarding the tests, I've got the LPCXpresso and the can bus analyzer
> send and receive messages, but with Lisa/M I haven't managed to
> perform any communication at all. I've tried with different bitrates
> but without luck. I don't have a jtag debugger so I'm only sending
> some telemetry messages and blinking some leds for debugging. I am
> using master branch and fixedwing firmware. Searching for can bus in
> the paparazzi code I only found the "Beth" airframe to work with
> canbus, and since it's a custom code for a bench test, I am wondering
> if the can bus driver conficts with other elements of the fixedwing
> firmare. I have written a small module to test the canbus driver that
> periodically sends a message, and also i have written a receive
> callback to "echo" the received messages back.
>
> Regards
>
> Miguel.

_______________________________________________
Paparazzi-devel mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/paparazzi-devel