I probably wont change what I have done, since it's in place and working but  the next time,,,

What I'm considering is in the attached image.

And this is absolutely right solution!
An alternative way is to remove the on-board internal weak pull-up resistor from the channel circuit: http://wiki.cannylogic.com/index.php?ti … t_diagrams

Are you just trying to improve the efficiency of your logic or is it not working as well as you'd like?  I wouldn't be too concerned about a few extra logic logic steps if the display looks the way you want  it to.  I was impressed.

Having said that, the addition of all of those 1s to identify the gear is something I have spent a lot of time trying to improve - without success.  Surely there is a smarter way to do it than all of those summing blocks!

Reverse has its own CAN message and display, so no need to try and identify it from gear ratios.

Thank you.  I love your gear counter!  It's very clever and I never even thought about doing something like that.  Since reverse is geared a bit lower than first, you might even be able to pick up that, but it's probably easier to use the backup lights.

On my question I guess it all boils down to how the CANNY works.  On the receive side, it must be that the gateway is only the loading Rx buffer when a selector is active and the specified ID shows up.  if you are only looking at one ID, there shouldn't be any problem.   However, if the gateway is acting asynchronously and you have selected several IDs there is some risk that another selected message overwrites the RX buffer before your logic is finished pulling data from it.  On the transmit side, the gateway must not be using the TX buffer at all until you issue a request to send.  I think that makes inserting records pretty easy and safe.

I appreciated  seeing your example.... and knowing that it works, I think I'll be able to do what I need to do pretty easily.

Thanks again.

Are there any examples for how to insert a message when the can gateway is running? My first reaction is just to build the new message in the appropriate CAN TX buffer and enable request to send, but i'm not sure how this interfaces with the gateway.

I have created and sent messages in the usual way, with no special interaction with the gateway.

I would capture the data in the usual way, apply your logic and then update the gateway settings.

I've attached one of my diagrams that does these things (along with a bunch of other stuff).

Are there any examples for how to insert a message when the can gateway is running? My first reaction is just to build the new message in the appropriate CAN TX buffer and enable request to send, but i'm not sure how this interfaces with the gateway.

I would also like to be able to select a message that is coming through the gateway with a specific ID  and drop or modify it based data content,.   It's easy to do based on the ID alone but I'm not nut sure how to  interrogate  the rest of the message.

Your are correct on my connections.  Right now I am only modifying messages from the instrument cluster to the rest of the car.  With the Canny removed all of the bus components function properly so only my insertion of the Canny is suspect.  In my early tests the inversion of the input to Sleep Mode was there but somewhere after that I accidentally deleted it. The reaction of the CAN bus on the car side is consistent with that problem, so I suspect it is the source of my problem.  Thank you in advance!

I also liked your buffer approach on the DIO channel, so it's going into the next version my testing.  I've wired in the speed signal wire to DIO 9 but wiill wait until I have the rest working before adding any logic to deal with it. 

I'm going to go skiing for a few days over the new year so it will be a bit before I can post again., but I feel pretty good about it.

Happy New Year.

Also, I like the use of the buffer to detect the change of state, I have used RED and FED for this purpose in the past, but the buffer method could be used for more than just digital signals.  I will remember this one!

You have set the  "CAN GW Selector 0 Linked CAN IF" = 1, and "CAN GW Selector 1 Linked CAN IF" = 1. This means that messages going in the direction from panel (CAN1) to car (CAN0) will changed. It is correct?

As for CANNY 5.2 duo and FT-CAN: When the CANNY is connected to the FT-CAN network of two or more FT-CAN devices, it is not necessary to switch on the transmitter mode. So, you can safe connect CAN0 interface to car network. When the CANNY is connected to a single CAN-FT device, you must to enable FT-CAN mode of CANNY CAN interface for for proper CAN Sleep Mode functioning. You can enable FT-CAN mode only on the CAN1 interface. So your hardware configuration when the car is connected to the CAN0 and the panel to the CAN1 is correct. CANNY 5.2 duo  are widely and successfully used on VAG with 100k FT-CAN. The situation you described with inadequate car equipment behaviour is typical for improper connection or a faulty on-board CAN network. I am sure that you will be able to resolve this issue. You can count on our help.

For the reverse signal I originally tried using leading edge detection on  channel 8, but since my backup light signal goes open instead of to ground when i'm not  in reverse....the pull-up resistor kept it from working.  Channel 10 doesn't have a resistor so it may work there.   The ADC approach may have been slower but it was very simple to code and it did work, so I don't thinks that's the reason the rest of the bus is being jammed up.

I didn't notice the unit trying to go sleep when i was bench testing it.  It did when I pulled the can input, which was what I expected. but this could well be the problem.

I'll retry and see what happens.Thanks again.