Settling time on SCXI-1102C modules

[MagmaiKH]

Here's my latest problem in my NI-DAQ saga:
I can't acquire accurate measurements from any of my 1102C modules, with a settling time less than 5us. The documentation on the 1102C indicates that a 3us settling time gives you 0.012% error - which would be more than enough for my application.
What other conditions must be met to acheive the 0.012% error?

The problem is worse when reading channels next to open ones - the open channels float to full-scale, and the channel that actually has a tranducer plugged in is very noisy if it registers any change in voltage at all.

At a 3.125us settling time, the transducer doesn't register, at 4us it's noisy, and at 5us the signal is clean.

Other than providing appro.
termination for all the open channels, is there anything that can be done to improve the settling time?

The 1102 modules are even worse, but since they're thermocouple modules, I can scan them at a slow rate and no one will care.

[Geneva L.]

You can find some good resources at the http://www.ni.com/support pages when you search "settling time and scxi scan". You are correct that you want to ground out the open channels, since floating channels don't have any meaningful data anyway. The KnowledgeBase entries give more description on that. Also, the Developer Zone tutorial shows some tests you can apply to your DAQ board to determine its settling time. Also, I don't know if it will make a significant difference in your situation, but for accuracy in general, you want to make sure your board has been calibrated within the past year. You can look at the http://www.ni.com/calibration pages for more information. If you have an extremely high precision voltage source and DMM, you can even manually calibra
te the 1102 module, as it has an EEPROM.

KnowledgeBase 0HH79DRV: "I See the Same Signal on Several Channels when Scanning, Especially when No Signal is Connected to Those Channels"

KnowledgeBase 28MF7JQO: "Data Acquisition: Troubleshooting Unexpected Voltages or Crosstalk in Analog Input Channels"

NI Developer Zone tutorial "Is Your Data Inaccurate Because of Instrumentation Amplifier Settling Time?"

Regards,
Geneva L.
Applications Engineer
National Instruments