NI DMM Waveform Default Settling Times ?

[zou]

https://www.ni.com/docs/en-US/bundle/ni-dmm/page/waveform-acquisition-defaults_2.html

What does it mean ?

[Messenger2 · Accepted solution]

The 1st table refers to default configurations, but looks like each configuration can be adjusted by looking up the device manual. 

The 2nd table refers to Waveform Default Settling Times, which means that when acquiring a signal, 1ms, 2ms, 5ms, etc. must pass before the output accurately reflects the change in input.

Say a signal is at 5V. It's less than 10V, so we'll use 1ms settling time. When 5V drops to another value, e.g. 2V, the output will accurately show 2V after 1ms. Before 1ms has passed, the output may show 2V, or randomly it may show 1.893V, or it may show 2.4V - not a reliable value - until 1ms has passed and the multimeter can adjust to the change and show 2V. 

If my explanation is wrong, please feel free to correct me 🙂

Regards,
Messenger2

[Messenger2]

Dear zou,

If my reply helped, can you please mark this post as solved? Thank you.

[zou]

Thanks for reply.

We are trying to measure the voltage drop/rise time using DMM.

The waveform settling time makes our measurement accuracy to ~1ms for signal < 10V.

And why the waveform t0 is always 0?

[Messenger2]

By t0 do you mean the first sample (t = 0)?

Can you attach screenshots of your measurement, please? Or relevant code/etc.

Here's an interesting link that might be helpful in the future: Digital Multimeter (DMM) Measurement Fundamentals - NI. I don't want to recommend something without understanding your situation fully, but you could look at the various configuration options of DMMs, such as Null Offset or Auto Zero. 

e.g. see Configuring Auto Zero - NI

[zou]

t0 is a waveform component.

I'm using NI DMM examples.  niDMM Read Waveform returns a waveform with t0=0.

[Messenger2]

I looked into one such example (NI-DMM Waveform Demo.vi). It has a subVI, acquire & update.vi. It does not interact with t0 component, only Y, the 1d dbl component. (Please see green circles in screenshot)

If t0 is not interacted with, then it's probably because it comes from a constant 0? Nothing changes the 0, so it stays as 0.

If you need to interact with the t0 component, at some point you need to provide timestamps, like this:

[zou]

I'm trying to do a hardware triggered measurement, use current time as t0 doesn't work for me.

Other devices such as NI Scope, or DAQmx can return a valid t0.

[Messenger2]

You could try creating a timestamp that aligns with your trigger. For example, you could set the DMM Trigger as its clock and call the Get Current Time function alongside your trigger. That way, each "block of samples" will have its own t0 timestamp and when that "block" is added to a Waveform Chart, you can accurately see each block's position on the chart.   

Alternatively, you can read about how DAQmx functions handle timestamps here. 

Can I ask for more details about your project? Maybe you don't need to specify a t0 at the time of acquisiton - instead, if the data is continuous, a single t0 timestamp after the acquisition starts may be enough. 

In short, we need to understand where in your code/acquisition do you need to set a timestamp. What are you trying to accomplish?

[zou]

It seems that NI DMM is not suitable for my test.

End of discussion.