USB-6255 Settling Time

[mbeneck]

Hi everyone,

I'm using a USB-6255 to measure 16 +- 5V analog signals, and I figured I would actually check the settling time spec given in the spec sheet (1.6us for 1 LSB). I performed the DC settling time test described here, but I'm seeing 1 LSB settling times of >1ms, more than 1000x worse than the datasheet specified. Has anybody observed this way out of spec behavior with NI devices, or have any suggestions for what to check next? Graph of settling results below, ignore the mux line, that's for a different part I'm testing.

Experimental setup:

- 2V connected to differential AI0

- -2V connected to differential AI1

- VI attached

[Henrik_Volkers]

What source impedance did you use?

I'm shure the optiistic spec was created with a sourceimpedance of about 0 Ohm 😉 (well or  the internal resistance of the mux ...)

If you take a look at the figure :settle error versus time for various source impedances  .. keep in mind that both axis are log scale 😄

That's why I always try to avoid multiplexed inputs ...  nothing but trouble

[mbeneck]

Turns out it was actually the source I was using to test settling time - it's not that it has high output impedance, just that it has some internal feedback loop that take 200 us to settle the output. I tried the settling time test again using the USB-6255 AO, and got numbers very similar to the specification 🙂