Your comments

Deleting the External Function reference to convert them to standard function groups causes the errors to go away.

I'm still getting some strange channel movements on my display after the latest update of 5.04.  Anyone else experiencing this?

Not supporting X*2 modules means we will not be able to use this software for years.

We experienced a similar problem with a Microsoft Surface (4k, I think).  Clicking on a channel would cause the other channels to jump to a new location on the screen.  Overriding the high DPI scaling with the "System (Enhanced)" setting fixed this problem.

One method I have used is to set the integrand to the value of the channel multiplied by a negative number.  This will drive the integral to 0 relatively quickly.

I upgraded to 5.04 and now I have errors in some external functions with "Invalid identifier".  This appears to be related to references of channels with the scope set to "Public".  Was this intentional?

I should have been more specific.  The bug only occurs if the timer exists in an external function.  I edited the external function, saved it, updated the external function in the Design application, saved it, and started the simulation from Design.  I can send you a simple .idax file that shows the bug.

Any update on this?  I would vote for it a thousand times if I could.

So this works, but doesn't work at the same time.  Yes, I can calculate the 29-bit identifier and have a generic frame out.  However, I can't attach that GFOUT to a J1939 device, so I have to create a new generic bus with a generic device.  Now the problem is that I can't have a generic device and a J1939 device share a bus; even if the generic bus is set to 29-bit ID with the same baud rate as the J1939 bus.  So I either have to convert everything else on the bus to generic or go back to my original problem of not being able to change the PGN.  Either way is painful.  I think generic and J1939 should be allowed on a same bus if the communication is the same.