CANBUS Glitching with 3rd POD

[Pacetech]

When I have 2 Pods on Canbus (ISOPOD-X and ISOPOD V2), it works flawlessly.

Now, I've added a 3rd ISOPOD V2 - just hooking it up and not running any program at all (i.e. not sending or listening on CANBUS), I get glitching on the BUS.

I have the POD-X transmitting every 4 ms and it toggles a LED every 5 transmits (20 ms) and I am looking at it on the scope.

The POD-V2 is receiving and toggles an LED every 5 frames it receives - which is also hooked to my scope.

With just the 2 pods, everything is rock solid. Add the 3rd and I'm seeing the receive waveform jump and sporatic.

Now I set the 3rd POD to receive - just as a monitor and I'm seeing multiple transmissions and missed/out of sequence transmissions. For example, it sends the same frame 1ms later.

Its not constant glitching, but maybe 5-10 times per second.

The POD-X has a terminating resistor on-board (from factory) and I added a terminating resistor to one of the POD-V2 (120 ohm) at R8.

My 3 wires (CANL CANH AND GND) are TWISTED and no longer than 12" each.

I'm running 8 bit ID and 500 khz bus, using IPBus CLK.

Thanks for any help.

[RMDumse]

CAN Bus expects termination on the two ends. There is a place on the Pod's to put terminating resistors, however, having three sets is worse than having no sets, so we generally leave them unpopulated. Besides it is usually easier to install the terminating resistors in the wiring connector than on the board anyway, and then all three (or more) boards can be swaped around, and the terminating resistors remain at the "ends" with their end connectors.

Try putting 120 ohm resistors across CANH and CANL at the two ends, and see if that improves things.

[Pacetech]

Now the ISOPOD-X has built in terminating resistors right? It shows R8 and RN and I measure 115 ohms across CANH and CANL (when not connected to CANBUS and powered down).

And at the other end of the CANBUS wires, I have the POD V2 with a 120ohm resistor installed on R8.

So I have 120ohm terminating resistors at both ends of the BUS already.

[RMDumse]

What about the middle one? it should not have a resistor. Is that the case?

[Pacetech]

Yes, the ISOPOD V2 in the middle has no terminating resistor installed (R8 is empty).

[RMDumse]

For those following this thread, a brief summary of email communications. Looks like the power supplies were overloaded, and spiking the 5V. Since the only thing that uses 5V (except the 3V regulator) on the boards are the CAN drivers, it looks like they might have been dropping out on low spikes.