LED Troffers rondomly don't fire up

[McGillNIC]

I'm a lighting rep (with contractor experience) working with a contractor to troubleshoot a problem. We have two 277V circuits with Philips (48W) LED troffers (43 on one circuit and 47 on another circuit). These are controlled by Wattstopper ceiling mount sensors. The issue is that when the lights come on, randomly one or two lights won't fire up. If you switch the fixtures off and back on, these fixtures come on and a different one or two don't come on. The contractor thought it was an issue with the drivers (doesn't sound like it) however, we replaced 8 drivers and had the factory test them and they were good.

The first thought is that it is a problem with in-rush current since LEDs have a high demand when they are first switched on.

Another thought, the transformer feeding the 277V panel may be sized appropriate for typical lighting loads but can't handle the initial surge from the LEDs.

Any thoughts or experience with these problems or troubleshooting what the problem is?

Thanks for the help!

 

[PetrosA]

It could be an inrush problem. It would be nice if manufacturers built in a time delay dip switch on fixtures typically installed in large numbers so that you could have stepped activation instead of huge inrush. I don't know how the inrush affects the starting exactly, but I've seen similar issues with LED retrofit lamps in large numbers.

 

[McGillNIC]

It could be an inrush problem. It would be nice if manufacturers built in a time delay dip switch on fixtures typically installed in large numbers so that you could have stepped activation instead of huge inrush. I don't know how the inrush affects the starting exactly, but I've seen similar issues with LED retrofit lamps in large numbers.

Yeah, I agree. I found some inrush current limiters, but I've never seen them used.

 

[GoldDigger]

I'm a lighting rep (with contractor experience) working with a contractor to troubleshoot a problem. We have two 277V circuits with Philips (48W) LED troffers (43 on one circuit and 47 on another circuit). These are controlled by Wattstopper ceiling mount sensors. The issue is that when the lights come on, randomly one or two lights won't fire up. If you switch the fixtures off and back on, these fixtures come on and a different one or two don't come on. The contractor thought it was an issue with the drivers (doesn't sound like it) however, we replaced 8 drivers and had the factory test them and they were good.

The first thought is that it is a problem with in-rush current since LEDs have a high demand when they are first switched on.

Another thought, the transformer feeding the 277V panel may be sized appropriate for typical lighting loads but can't handle the initial surge from the LEDs.

Any thoughts or experience with these problems or troubleshooting what the problem is?

Thanks for the help!

It is much more likely to be an inrush or noise voltage on the supply lines if any of the circuitry in the WattStoppers or the LED drivers themselves uses a microprocessor rather than discrete analog and digital circuitry. Some of those systems do not reset or boot properly when the supply voltage is noisy during the startup process. RF filters on each driver may make a difference.
Simply low voltage will not cause a problem since the circuitry, if current starved, should simply energize the LED normally once the voltage rises high enough and is stable.
And the LED drivers will normally be seen as a large capacitive load when their internal DC power supply first starts up, which is over a much shorter time frame than the inrush and starting current for a motor.

If the WattStoppers are two terminal devices which use the load as the return path for their own internal power, it is much more likely that they have a problem too and may be affecting the LED drivers.

A quick (?) diagnostic test will be to disconnect half or so of the fixtures and see what happens, then increase the number until you see a problem. Splitting them up into a larger number of smaller circuits with a time delay controller function on each may allow you to do what the LED drivers themselves do not do.

 

[McGillNIC]

It is much more likely to be an inrush or noise voltage on the supply lines if any of the circuitry in the WattStoppers or the LED drivers themselves uses a microprocessor rather than discrete analog and digital circuitry. Some of those systems do not reset or boot properly when the supply voltage is noisy during the startup process. RF filters on each driver may make a difference.
Simply low voltage will not cause a problem since the circuitry, if current starved, should simply energize the LED normally once the voltage rises high enough and is stable.
And the LED drivers will normally be seen as a large capacitive load when their internal DC power supply first starts up, which is over a much shorter time frame than the inrush and starting current for a motor.

If the WattStoppers are two terminal devices which use the load as the return path for their own internal power, it is much more likely that they have a problem too and may be affecting the LED drivers.

A quick (?) diagnostic test will be to disconnect half or so of the fixtures and see what happens, then increase the number until you see a problem. Splitting them up into a larger number of smaller circuits with a time delay controller function on each may allow you to do what the LED drivers themselves do not do.

Thanks so much for the input. We will try splitting them up and check the control devices.

 

[cadpoint]

...Another thought, the transformer feeding the 277V panel may be sized appropriate for typical lighting loads but can't handle the initial surge from the LEDs.

I think you need to stop and see what the exact specifications are for both your lights and your drivers (transformers).

Here a link for Philips LED Drivers, note that three of them are at 48W.

Here
Philips has plenty of equivalent 48 Watt lights = + - 8, 9.5 W to 48W LED's.

 

[Ragin Cajun]

Thanks so much for the input. We will try splitting them up and check the control devices.

Can you update us with what you have found and is there is a solution?

RC