Yes, there is a CAN bus that speaks to the 1300, which also contains a bank of 210 units. These 210 units output voltages. It doesn't really matter that much, but we then connect the outputs to some controllers (using the voltages to simulate signals that a vehicle would normally see under certain conditions). Essentially we are using it to fool our controllers to think that its receiving input from sensors. Anyhow, the error occurs when we try and send a message across the CAN bus (using CANalyzer and the CAPL scripting language) to tell it to update all of the channels on the 4-210 units we have. If we only send 4 signals, we are okay, but as we increase we run into the overflow problem. I am not sure at what exact amount the FieldPoint unit overflows, but since we are trying to rewrite the voltage outputs at least once a second, it ends up meaning a lot of changes for the units. This is what we want to happen:
1) Start program
2) Write initial value to ALL outputs (24) (this is usually where it fails)
3) As signals come in (they dont all come in at the same time), change the voltages on the FP unit to match them (according to the D to A conversion we have decided on)
It seems that if we sort of skip step two, and never write to all the channels IN MASS, then we are okay. I am just wondering if we are going to run into a problem if we try and run step 3 (where the signals update) at a faster rate. Right now, the signals dont change any faster than once every 2 seconds. But this (because it is simulate input from a joystick and other machine information) is a very slow rate to be simulating real life behavior. Therefore, we want to send signals out maybe 2 to 3 times per second, but we dont know if we even can because we are afraid the fieldpoint unit will just overflow every time.
I am very surprised this isn't a well-known specification. Strange. Thanks for your replies so far Chris_C, I appreciate it greatly.