Collect 120 thermocouple inputs at the same time

[KVogeler]

I am creating a system that needs to collect ~1 sample/minute from 120 thermocouples spread out across a 130' x 115' area. Would it be better to route all of the wires to a single DAQ module about 30' away from one of the sides or should I use multiple smaller DAQ systems daisy-chained together to a single computer. I need to retain as high of an accuracy as possible. For the single DAQ system we would use a PXI system with 4 PXIe-4353 modules. For multiple systems, we would use a multiple CompactDAQ with NI 9214 modules. 

Thanks all!

[nanocyte]

If I had to guess, I'd say digitizing the signal as soon as possible avoiding long wire runs would be more accurate. It might be better to try and estimate how accurate you needed to be. Thermocouples are more known for their broad range rather than accuracy. 

[CFER_STS]

Do you anticipate any common mode voltages?  I am doing something similar, using five NI 9214 modules in a NI 9144 EtherCAT expansion chassis connected to a cRIO-9030 (FPGA mode).  I am using multiple pair thermocouple cable (type K, 24 AWG, 16 pairs + COM) which pairs nicely with the NI 9214 module (one cable per module), and provides a relatively cheap way to noise shield the thermocouple pairs.  These multi-pair cables terminate in junction boxes with mini-plugs, and then short thermocouple wires (5-8 feet) go from there to the test specimen.  One advantage of central signal conditioning is that the cold junctions are all in approximately the same location.