CAN timeout values not working as expected

[the_bluester]

Quick query for the development guys (I guess)

I am using a PDM30 and integrating it with a different vendors ECU. What I am trying to do is achieve as much as possible over the CAN bus. I am trying to drive as many PDM outputs as possible with CAN based messaging however I am running into a problem.

An example is, my ECU will spit out the fuel pump demand status as a CAN message which is easily pulled into the PDM. The message resolves to a 1 when the ECU requests the pumps run and a 0 when not and the pump output of the PDM is true when the message is a 1. If I let the ECU time out the pump prime the CAN message to the PDM becomes a 0 and the pumps stop as expected, however if the ECU is switched off while the message is a 1 the received message times out to a 0 as expected (As set up in the CAN RX area) but the output remains on.

I tried setting up a test function which is simply true when "CAN.Fuel pump demand = 1" and it behaves the same way. The CAN message monitoring the PDM changes to a 0 on timeout of the message but the function remains true when it should be false.

Is there something I am missing here? It seems to mean that without a kludged setup to disable outputs if the CAN bus fails I can not use CAN inputs to control important functions if I want to set a fail safe situation based on a timeout value of input messages. The physical outputs of the ECU go to ground when the ECU is off so about the only other way I can think of is to generate an enabling output from the ECU which goes high when true, otherwise if for instance the ECU or the CAN bus fails, everything that is on will stay that way until the battery isolator is switched off.

[adrian]

This is a known issue and will be fixed in the next version of PDM Manager.

This work is scheduled to be completed in the second quarter of 2018.

[the_bluester]

Awesome.

If the car happens to be ready to fire up before it is released I will build the double condition setup to disable things once the ECU is powered off and re write things once it is sorted out.

[the_bluester]

I was going to query a release date on the new PDM firmware again but I note there is a beta version from last month (Which must have been put up just after I last checked)

I assume "CAN Input Message timeout behaviour" in the fixes is what I was looking for with timeout values not being used for conditions?

Is the Beta pretty well stable and useable? If so I will update and then update again when it is in official release.

[the_bluester]

Never mind, gave the beta a go and now it works as expected with the CAN timeout values in logic functions. I will update again when it goes to full release.

I was not expecting a response over the queens birthday weekend in any case.

[adrian]

Yes that should all be fixed now. That particular version should be pretty stable, we haven't had any reports of major issues.

If you have any feedback on the beta please let me know (either on the forum or via our support email). It is always good to get feedback on stability/new features before we promote it to full release.

[the_bluester]

Well, my car is still in the building and testing phase all I can really say at this point is that I did not notice any stability issues with either the PDM itself or in the software and after testing through a number of cycles, the CAN timeout values appear to work properly when used in conditions.

I am using CAN messaging to drive the fuel pump outputs (And others like thermo fan control) rather than hooking up physical ECU outputs to PDM inputs to achieve the same thing. Where before if you killed the ECU while the fuel pump demand was active over CAN it hung up and left the pump output active it now seems to reliably use the timeout value (0 for off in this case) in the condition and stops the pump after 1 second. I have to assume everything else will work as it should.

I might end up hooking up a few ECU to PDM connections yet as my only concern is obviously a CAN bus failure will stop the car as I have it set up now. Maybe just the very basics like fuel pump control and aux power control for the DBW system and let CAN take care of everything else like fans and oil cooling pumps than can just default to "On" states in case of bus failure. I will loose my dash too but at least I could still drive the car back off the track rather than coming back at the wrong end of a rope.

[the_bluester]

Quick question.

Is the "Can input message timeout behavior" referenced as a "Fixed" in the release notes for PDM manager 1.9.1.0012 the full release fix for CAN timeout values not making output conditions change state as they should? Time for me to update from the beta I am using if that is the case.

Aside from that, the change to the "test outputs" function where outputs start with whatever state they were in before would be handy for a few things I am trying to do at the moment. Hard to use my ECU to make sure things work as expected when it looses power as part of the test.