Topic: XDAB Channel Limitations
I have a project with fifteen NIONs all XDAB'd together, and have found that the maximum numbers of XDAB channels cannot exceed 486 without the XDAB becoming unstable and passing digital garbage. Sounds like digital audio directly in to an analog amplifier.
The system is fifteen NIONs all XDAB'd, with a single 8i and single 8o installed in the expansion slots of NIONNODE #1. All the other system I/O is via CobraNet, with all channels being used on all NIONs (480 in and 480 out. Do I get an award?). All my CobraNet sources hit the XDAB before they pass to a series of mixers, then passed to the CobraNet outputs (8X1 mixer for each output). Also on the XDAB is a single NION pink noise generator which is distributed to a channel on all the mixers. A signal probe feeds one of the analog outputs, and each output has metering.
If I limit the number of channels fed to the XDAB (doesn't matter if the XDAB size is adjusted, just the number of channels fed) to 486, I am able to pass clean pink noise to all the outputs, including the signal probe. If I exceed that number to 487 channels, then the pink noise is nothing more than clipped digital trash.
What is interesting is the way the signal probe functions when the XDAB exceeds 486. I can clip to the pink noise generator and get clean pink noise. I can clip to the XDAB I/O for that pink noise and get a clean signal. I can clip to the flyoffs following the XDAB and proceeding the mixer and get clean pink noise, and can even connect to the wire feeding the mixer and get clean pink noise. But as soon as I clip to the mixer input wiring node, the signal probe audio is distorted junk, with the metering jumping from -35 to +35 instantaneously (the metering just flashes). I get the same results on the output of the mixer, right through to the CobraNet outputs.
The metering on the outputs of each mixer exhibit the same behavior.
Has anybody else come across this XDAB channel limitation?