LIN bus rate

[sinurb]

Hello,

I have tried your project on Arduino uno board, and it works perfectly fine. I have two problems:

1. When my LIN master sends a frame header (without data), and I print out the frame ID, I always get some garbage data on sio before the ID (which is detected correctly). What could be the reason for this?

2. When I send LIN messages fast (from an scheduler thread), I get SYNC error. What is the fastest that the code can receive new frame header after sending data (or better how can one measure this?)

Cheers, S

[zapta]

Hi brunisgirl,

Do you have access to an osciloscope or a logic analyzer? It will allow you to look at the signals.

Even a cheap 10-20$ one from amazon should be fine for linbus (which is very slow).

And some of the, such as the saleae compatible, even come with a Linbus decoder.

As for your project, I don't know enough details to comment. For example, do you use a serial interface to generate the linbus master signals? For example, how do you generate the sync period before each frame? Because it's not a standard serial byte, I think it requires such trickery with the serial interface and/or i/o port to generate.

Hope it helps,

Z.

On Wed, Dec 27, 2017 at 1:07 PM, brunisgirl notifications@github.com wrote:

Hello,

I have tried your project on Arduino uno board, and it works perfectly fine. I have two problems:

1.

When my LIN master sends a frame header (without data), and I print out the frame ID, I always get some garbage data on sio before the ID (which is detected correctly). What could be the reason for this? 2.

When I send LIN messages fast (from an scheduler thread), I get SYNC error. What is the fastest that the code can receive new frame header after sending data (or better how can one measure this?)

Cheers, S

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/zapta/linbus/issues/24, or mute the thread https://github.com/notifications/unsubscribe-auth/ACFWQYKWoNHM8CglMOnCqRUEPSMCJYvMks5tErGBgaJpZM4RNz9X .

[sinurb]

Hi Z,

I do have an oscilloscope with LIN bus. But since my master is waiting for the response to send the new header, and the arduino goes to SYNC error, I can not send another frame meanwhile. However, the first header is received correctly and a response is generated from arduino (I have added a LIN send function to main loop after receive of a frame). I added the function just before the line: // Print frame to serial port. lin.send(frame.num_bytes());

void Lin::send(uint8_t nBytes)//(uint8_t addr, const uint8_t* message, uint8_t nBytes,uint8_t proto) { byte message[] = {0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77, 0x88}; // LEDs ON //uint8_t addrbyte = (addr&0x3f) | addrParity(addr); uint8_t cksum = dataChecksum(message,8,0x32); serialBreak(); // Generate the low signal that exceeds 1 char. serial.write(0x55); // Sync byte serial.write(0x32); // ID byte serial.write(message, 8); // data bytes serial.write(cksum); // checksum }

this function comes from: https://github.com/gandrewstone/LIN I have modified it a bit just for testing. and the oscilloscope detects it as a LIN message.

For the master I use a Hercules board from TI, and with LIN transceiver in master mode. Then generating the sync or break is handled by the SoC on the board itself (i.e. I just send a header, the board adds the sync and break itself, which are detected by oscilloscope as well)

Best, S

[sinurb]

Hi,

I have figured the problem. It was the collision due to writing using the "sio" to the serial line as well. Now, I would like to ask, how would you implement a slave response sending data? Does it make sense to add a new state after receiving the header where I send the response?

Thanks