CAN-bus library for general CAN-bus controller chip MCP2515

andro · Joined: 12 Mar 2007 Posts: 70 Location: Ljubljana

Hi Per,

It seems to me, You do not use INT pin (from MCP2515). Is that correct?

BR
Andrej

Per Svensson · Joined: 03 Oct 2004 Posts: 185 Location: Gothenburg, Sweden

Hi Andro,

Yes I skipped the INT pin and use register polling instead.
SPI is so fast so I decided not to use interrupt.
But it is of course easy to let DataReady generate an interrupt, if that better suits the application.
Perhaps we should let a compiler directive select between hard INT or soft polling?

/Per

andro · Joined: 12 Mar 2007 Posts: 70 Location: Ljubljana

That would be a major change of the concept. First I would start as it is - without INT / ISR handling.
I already prepared some HW
BR
A

rkumetz · Joined: 23 Jan 2007 Posts: 148 Location: Vermont

I noticed that the values being written to the baud rate registers do not match those shown in the calculation table above the code.
In particular, CNF1 @ 250K and 500K do not match. They may have been swapped.

Per Svensson · Joined: 03 Oct 2004 Posts: 185 Location: Gothenburg, Sweden

Hi rkumetz

Yes there was a typing error in the CASE for 250 and 500 kbit/s
The table was right however. I double-checked it against Kvasers's baudrate calculator.
Here is the correct CASE table:

Select Case Baudrate 'kbit/s
Case 10 : Cnf1 = &H27 : Cnf2 = &HBF : Cnf3 = &H2
Case 20 : Cnf1 = &H13 : Cnf2 = &HBF : Cnf3 = &H2
Case 50 : Cnf1 = &H09 : Cnf2 = &HB5 : Cnf3 = &H1
Case 125 : Cnf1 = &H03 : Cnf2 = &HB5 : Cnf3 = &H1
Case 250 : Cnf1 = &H03 : Cnf2 = &H91 : Cnf3 = &H1 '<- Corrected Feb.18 2020
Case 500 : Cnf1 = &H00 : Cnf2 = &HB5 : Cnf3 = &H1 '<- Corrected Feb.18 2020
Case 800 : Cnf1 = &H00 : Cnf2 = &H9A : Cnf3 = &H1
Case 1000 : Cnf1 = &H00 : Cnf2 = &H91 : Cnf3 = &H1
Case Else : Can_baud = 1 : Exit Function 'Error (ignore new baud)
End Select

How many people have tried this CAN-implementation? Is would be interesting to know.
Just wave hand, or give a short comment here if you have the time...

/Per

rkumetz · Joined: 23 Jan 2007 Posts: 148 Location: Vermont

Thanks for double checking. I looked at the values in the comment table and corrected them but
since I am new to CAN I wanted to make sure that I was not making a mistake somewhere.

This code is super helpful. I started out with the MCP2517 and then found that there are apparently
some odd "quirks" such as an unstable oscillator so I switched to the pin compatible MCP2518.
For my simple application I found that the '2518 was horribly complex compared to the '2515 and
the added functionality did not outweigh the extra time to get it going. There is also virtually little
information out there about the 2518 which makes finding examples tough. I bought an MikroElektronika
"Click" board and hand wired it to the pads for the MCP2518. In an hour I was way ahead
of where I started out.

Thanks again!