lyleboys@yahoo.com wrote:

> BUT, I couldn't find that specified in the CAN spec downloaded from
> Bosch.

And if you look at that for a while longer, you may eventually find
out why that is: the CAN spec doesn't mention message filtering at
all.  Filtering is not part of CAN --- it's just so obviously a useful
thing to have if a CAN controller offers more than one receive
mailbox, that basically all of them have it.

There are also the so-called "basic" CAN controllers that don't have
any filtering.  They just have a single receive mailbox (optionally a
FIFO hiding behind it) that receives everything passing over the bus
except what it sent itself.  It's then up to the CPU to figure out
what to do with all that data.

-- 
Hans-Bernhard Broeker (broeker@physik.rwth-aachen.de)
Even if all the snow were burnt, ashes would remain.

It is mentioned in the Bosch can2spec.pdf, part B, section 4.

It talks of filtering based on the whole identifier, not the whole
message, so could a controller abort the message early, based on
filtering, and not generate any response?

lyleboys@yahoo.com wrote:
> That is my experience as well, that the filtering happens after the
> message receive and ack.
> 
> BUT, I couldn't find that specified in the CAN spec downloaded from
> Bosch.
> 
> Rather than chip specific info, I was wondering if the _the spec_
> dictates whether filtering affects generation of the ACK.
> 
> The hardware seems to do as you say, generate the ACK independent of
> filtering, but is that only chip by chip?  Or is there a specified
> behaviour?
> 
> thanks
> 
I don't recall the filtering even being mentioned by the Bosch spec -- 
AFAIK it's just a really cool feature of the first CAN controller chip 
that everyone else adopted.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Posting from Google?  See http://cfaj.freeshell.org/google/

Un bel giorno lyleboys@yahoo.com digitò:

> Is the filter for CAN messages before or after the receiver / ack
> hardware?

As far as I know, the ACK simply means that the message has been received
without CRC errors. The process of masking/mailboxing belongs to the
"higher" levels: a CAN node receives all of the messages even if it doesn't
put any of them into a mailbox for user processing.

You can get the specs directly from Bosch site:

http://www.semiconductors.bosch.de/pdf/can2spec.pdf

At page 15 it is written:

"ACK SLOT
All stations having received the matching CRC SEQUENCE report this within
the ACK SLOT by superscribing the ’recessive’ bit of the TRANSMITTER by a
’dominant’ bit."

-- 
asd

That is my experience as well, that the filtering happens after the
message receive and ack.

BUT, I couldn't find that specified in the CAN spec downloaded from
Bosch.

Rather than chip specific info, I was wondering if the _the spec_
dictates whether filtering affects generation of the ACK.

The hardware seems to do as you say, generate the ACK independent of
filtering, but is that only chip by chip?  Or is there a specified
behaviour?

thanks

lyleboys@yahoo.com wrote:
> Is the filter for CAN messages before or after the receiver / ack
> hardware?
> 
> Consider only two nodes on a bus, the first node has ALL messages
> filtered out, the second node transmits a message that will be filtered
> by the first node.  Will the first node still generate the ACK on the
> bus, assuming no errors in the packet?
> 
> Does the ACK occur before or after the filter is applied?
> 
> Please provide sources for your information / opinion.
> 
> thanks
> 
That is, of course, dependent on the design of the CAN controller chip 
that you're using, and you haven't told us what it is.

On the chips that I have looked into or used the ACK happens in hardware 
whether or not the message is filtered -- the only thing that filtering 
controls is whether the message is saved and the processor interrupted. 
  My only direct experience with this is the TMS320F2812 DSP -- but all 
the data sheets that I have read seem to indicate the same thing.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Posting from Google?  See http://cfaj.freeshell.org/google/

Is the filter for CAN messages before or after the receiver / ack
hardware?

Consider only two nodes on a bus, the first node has ALL messages
filtered out, the second node transmits a message that will be filtered
by the first node.  Will the first node still generate the ACK on the
bus, assuming no errors in the packet?

Does the ACK occur before or after the filter is applied?

Please provide sources for your information / opinion.

thanks