How to Control LED Lighting Intensity from LabVIEW Using PWM

[cedricuk]

I’m looking for advice on controlling LED lighting intensity from LabVIEW using PWM.

The goal is to adjust the brightness smoothly while avoiding visible flicker. I’m currently trying to determine the most suitable NI hardware and LabVIEW method for generating the PWM signal.

• Hardware: Not selected yet
• LabVIEW version: Not selected yet
• Driver/version: Not applicable yet
• Error: No error message at this stage
• Troubleshooting: I have not yet implemented the hardware setup.

Would PWM be the recommended approach for this application, and which NI hardware/module would be suitable for generating the required signal?

I can provide a VI and screenshots once I have a working test setup.

[Kyle97330]

You're new here so I don't really have a good basis to judge your experience.

Are you aware that LEDs are primarily current-driven?

NI hardware that can output waveforms of the sort you would want are primarily voltage-based.  The very few that can output current do so at the 0-20 mA level, which is meant for industrial signaling, and would be insufficient for all but the smallest LEDs.  You haven't said what this is for but it feels like it's for a larger application than just a few small LEDs.

Also, most LEDs are not just "on or off" only, and can stand to have their output level managed just by current control instead of needing to modulate the output.  PWM would likely work to dim the output, it's just that unless you're extremely limited in your choice of LEDs and are forced to use ones that can't be dimmed enough just by current reduction, PWM is likely an unnecessary complication.

What application is this all for, anyway?

[cedricuk]

Thanks, that makes sense. I’m looking at LED lighting rather than small indicator LEDs, so the current-control point is useful. I haven’t selected the driver or NI hardware yet, so I’m trying to understand the right setup before choosing the hardware.

[Kyle97330]

NI hardware is great for input.  We're able to cut costs a ton on some things by using NI DAQs instead of oscilloscopes, for instance. 

Their outputs work OK for signaling-level (though they have fewer options, and usually less channels), but you pretty much always need something on the other end between you and the end objective unless your DUT only needs signal-level voltage or current.

You'd probably save a lot of money and time by just skipping directly to using a LED controller that accepts Serial or Ethernet control.  You could still use LabVIEW to control it...

[cedricuk]

Thanks, that’s really helpful. I appreciate the guidance.