-1

We have a DAQ board from Measurement Computing which has analog and digital outputs and inputs. It is written as a warning on it that one shouldn't plug an output channel back to the board as an input. Why is that so? But as far as I know we can do this with an Arduino board (?)

Here is the device: http://www.mccdaq.com/PDFs/manuals/USB-1616HS-BNC.pdf

user16307
  • 12,049
  • 52
  • 184
  • 320
  • 1
    Impossible to answer without some kind of documentation on the board... – Matt Young Nov 25 '13 at 13:22
  • As Matt said you need to supply more info. Most likely reason I could think of is the outputs have a larger voltage range than the inputs but there might be other reasons. – PeterJ Nov 25 '13 at 13:24
  • yes u re rigth here is the device: http://www.mccdaq.com/PDFs/manuals/USB-1616HS-BNC.pdf – user16307 Nov 25 '13 at 13:31
  • btw some engineer wrote this warning on a paper. im not sure why he did that. – user16307 Nov 25 '13 at 13:33
  • I suspect, but don't know, that you might be tying the AI ground to the board ground in ways that are less than ideal, and overriding some carefully thought out design decisions at the board level. – Scott Seidman Nov 25 '13 at 13:43
  • what if one plugs an output to an output of the same device? maybe he typed it wrong? – user16307 Nov 25 '13 at 13:48
  • What is a "analog digital output"? – Olin Lathrop Nov 25 '13 at 13:59
  • i forgot the comma or and. now analog and digital just edited – user16307 Nov 25 '13 at 14:00
  • I followed the link but found a large PDF document. No, I'm not going to read the whole thing to find the warning you are talking about. Without a specific reference, there is no information there and this question needs to be closed because we can't tell what is really being asked. – Olin Lathrop Nov 25 '13 at 14:02
  • no the warning is not in the pdf. its just that one engineer wrote a note on a paper so everybody now afraid to connect this way. we are all scared cause we dont know . the engineer is not workin anymore. we dont wanna blow up this expensive device. i wanted to confirm here from eligible people. – user16307 Nov 25 '13 at 14:03
  • If I said they could definitely be interconnected, would you believe me or would you, as a final check, contact the supplier of the apparatus? – Andy aka Nov 25 '13 at 14:15
  • i dont know i thought it might be a ground loop issue but i cannot figure the picture now how is the inner connections – user16307 Nov 25 '13 at 14:22
  • If the signal conditioning circuitry or grounding is thoroughly crummy, you might create an oscillating loop that way :) – rackandboneman May 30 '17 at 10:33

2 Answers2

3

While your specific case might have some reason not to do this, in general it is fine to connect a DAC to an ADC on the same device. Indeed, I do this all the time to test the DAC, anti-image filter, anti-aliasing filter, and ADC all at once. I call it a "loopback test".

In your case, I would guess that the reason for the warning is that the DAC output range is ±10 volts, but the ADC input range can be configured to be smaller, as small as ± 0.1 V.

One should also check the source impedance of outputs and the load impedance of inputs, to make sure that the input won't draw excessive current, etc.

Can you tell us where in the manual the warning occurs?

nibot
  • 1,690
  • 1
  • 14
  • 23
0

In scenarios where a device may output a signal which depends on the value of an input signal, looping the output back to the input may have undesirable effects. I can't think of any measurement apparatus where I've seen such issues, but such problems could occur if the board has a mode to make the output signal mirror the input signal and such a mode is enabled by default until the board is given some other specific signal to output instead. If the output of such a board was e.g. connected both to a motor controller and the board's analog input, and if the board's analog gain could (depending upon temperature or other factors) vary from 0.999 to 1.001, it would be possible that powering up the system would 99 times out of 100 give a "zero" command signal to the motor controller (when the system gain was less than 1.000) but without warning give a full-scale positive or negative command instead.

supercat
  • 46,736
  • 3
  • 87
  • 148