CAN Loopback scoped Tx dead

[zamolan]

Posted on November 12, 2014 at 12:04

Hi! 

   I am using a DevKit407 ( stm32f4 discovery + stm32f4dis-BB). 

1.  I managed to set up a perfectly functioning

   CAN1 in Loopback mode

   . I assume I should see the test frame sent on Tx pin (PB9 on stm32f4discovery, or CAN1_TX on BB) with an oscilloscope (sampling rate 2.5 GS/s) .

   But I do not see any movement.

   Internally shorting the Tx/Rx pins will close Tx pin?
2. Is there a problem if I use split termination, but the connected devices will have simple 120 Ohm termination? 

Thank you for your time and knowledge! 

#discovery #can #stm32f4

[Tesla DeLorean]

Posted on November 12, 2014 at 19:27

I don't recall if Loopback mode emits a signal or not, but it doesn't make for a functional bus test, which would require that you have two node, with one receiving and acknowledging the data on the bus.

You could do it with one board using CAN1 and CAN2, and a couple of transceivers or diodes.

https://my.st.com/public/STe2ecommunities/mcu/Lists/STM32Discovery/Flat.aspx?RootFolder=https://my.st.com/public/STe2ecommunities/mcu/Lists/STM32Discovery/STM32F4%20DISCOVERY%20CAN%20bus%20controller%20problem&FolderCTID=0x01200200770978C69A1141439FE559EB459D75800084C20D8867EAD444A5987D47BE638E0F&cu...

[zamolan]

Posted on November 12, 2014 at 20:49

Thank you! I wasn't sure on this setup. 

   Though I have a Vector CANCaseXL at my disposal, if only the acknowledge is needed. Any working network node with necessary speed (High Speed CAN in my case) will automatically acknowledge frames I presume?

[Tesla DeLorean]

Posted on November 12, 2014 at 21:54

I would assume so, I've had STM32Fx devices talking to each other.

The one problem I see most often with the ST examples is that the bus clock and baud rates are wrong. The STM32F4 does not typically have a 60/30 MHz bus configuration.

[zamolan]

Posted on November 13, 2014 at 12:08

Yes. The bad setup at my example was port related (when I struggled to implement this example for loopback, not to scope it).

Though it clearly specified that this example works for STM32F4xx eval boards, the CAN was set up to GPIO G, which isn't even exists. This was found in the Loopback from the library of site provided: STM32F4xx_DSP_StdPeriph_Lib_V1.3.0