fast thermocouple monitoring

[biff]

I am trying to put together a system to monitor 25 or so small thermocouples at rates up to 75Hz using LabView to look at flow instabilities in plastics processing. From reading around it seems that the best way to do this is via the SCXI 1102 coupled with a suitable terminal block and DAQ board (I already have a PCI-6035E so I'll probably use this). Can someone please tell me if I am correct in my specs and of any shortfalls/problems in utilising this system. Many thanks in advance.

[Brent_R]

What is the actual response time of your thermocouple in this system? The 1102 has a fixed lowpass filter at 2Hz. For most thermocouples this is fast enough and is why it is recommended over the 1102B/C in the catalog, however if you system actually has a higher bandwidth, then you could be losing useful information. If your thermocouples respond significantly faster than 2Hz, I would recommend the 1102B, which has a cutoff of 200Hz. You will need to sample faster to prevent aliasing (maybe 1kHz per channel), but this should not be a problem with the 6035E. You can then filter the data in software down to whatever bandwidth is appropriate for your system.

Regards,

Brent R.
Applications Engineer
National Instruments

[PdB]

Interested experiment (I used to develop for a plastics R&D lab).
Your limiting factor will be the bandwith of your transducer (thermal heat-up time of the sensors metal and the protected coating/housing).
Be aware of electrical noise in your low-voltage measurement signal since extruders/injection molding machines have high-power electric heaters, pumps and motors that are not designed to make the work of the T&M guy any easier;-).

Patrick de Boevere (info@serenity.nl)
Serenity systems design

[biff]

My thermocouples are 'hand made' and consist of 0.3mm wires electro-welded in a mesh orientation. I would reckon we should get at least 20Hz, if not more, out of them. I guess the 1102B would be the one to go for in that case and sample fast and either perform hard or software filtering to prevent aliasing. Thanks for your help.

[biff]

Already discovered the delights of working in high noise extruder/injection moulding environment. Thermocouples not standard but small enough to give 'fast' (say at least 50Hz) response. Will try the 1102B with a TC-2095 terminal block. Fingers crossed this should be OK.