0
\$\begingroup\$

Thinking this should be obvious, but having trouble identifying a fix: I use multiple channels on an NI PXI DAQ card for RSE voltage measurements. I need to verify the voltage readings returned by the DAQ card with an oscilloscope. Without the DAQ card input running, the Scope shows similar data. If I run the DAQ acquisition at the same time as the Scope, i see periodic blips that happen at the same rate as the card multiplexes between its input channels.

A) Why would I see these blips on the scope? B) What can I do to eliminate them?

Details: The AI connections and Scope probe connections are made at the same location. A higher DC Coupling impedence results in a higher amplitude of the blips. Switching to AC coupling has no effect. The blips are seen even if the circuit is not being driven by a signal.

Thank you!

\$\endgroup\$
  • 1
    \$\begingroup\$ It looks like the front end of the DAQ does not have input analog buffer, so the S/H (sample and hold) unit loads your (finite-impedance) source on every switch of the MUX. \$\endgroup\$ – Ale..chenski Feb 22 '17 at 21:30
1
\$\begingroup\$

Think about how the input multiplexer works...

If it is implemented with solid state analog switches, then these are really FETs. FET gates are capacitors. Switching the FET involves charging this capacitor. This injects a spike into the mux inputs and outputs. Depending on circuit impedances, it will result in more or less voltage.

Most analog switch datasheets include a "charge injection" specification for this very reason.

Now, even if it is implemented with relays, consider the fact that the switch connects your probe point to a circuit inside the DAQ board, and this circuit has capacitance. When switching, current has to be drawn from the probe point to charge stray capacitances to the voltage being probed.

You can't measure something without changing it. However, I would presume the NI engineers did their job properly, and inserted enough delay between the mux switching and the ADC acquisition to let these transient phenomena settle before digitizing the input voltage.

\$\endgroup\$
  • \$\begingroup\$ Forgive my ignorance here, but if the circuit is not being driven by a voltage signal, how could the capacitors charge and cause the blips? The blips show even if the scope is connected directly to the analog input channel, with no other circuit elements (with or without a 221 ohm resistor between the probe signal and ground). \$\endgroup\$ – user140011 Feb 22 '17 at 22:19
  • \$\begingroup\$ I think the multiplexer inside the analog input of your DAQ card generates the blips through charge injection. Every analog mux does this. \$\endgroup\$ – peufeu Feb 22 '17 at 23:34
  • \$\begingroup\$ I'm not absolutely sure this is the reason for what you see, but it's the most likely culprit. Search for "charge injection" in this document if you're interested: analog.com/en/analog-dialogue/articles/… \$\endgroup\$ – peufeu Feb 23 '17 at 9:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.