Mystery: LED fixture randomly goes out
I have an LED strip fixture that randomly, mysteriously stops working. I’m trying to figure out why, so I can remedy it.
The hardware: 20 feet of PureEdge SS5C LED strip (4.4 watts/ft, total 88 watts), 96 watt PureEdge 24V electronic power supply, Lutron RRD-6NA dimmer switch (which PureEdge states is compatible). Power supply is installed in an open cove, right next to the beginning of the light strip.
The symptoms: the light randomly goes off, even though there’s still power to the switch. The circuit breaker does not trip, nothing else on the circuit is affected, and no other LED fixtures or anything else in the house is affected. Turning the switch off and on turns the light back on, and it remains on. Sometimes this happens a few times a day, other times it it happens once a day every day for a few weeks, and other times weeks go by without happening at all. It doesn’t seem to be related to temperature inside or outside nor how long the fixture has been on.
Troubleshooting steps so far: PureEdge states the 96 watt power supply is sufficient for this light strip. They replaced the transformer. My electrician replaced the dimmer switch and checked all connections. The symptom seems to probably happen less often since the replacements, but it still happens.
Questions:
1) Could PureEdge be wrong about whether the transformer is correctly sized for the strip? It seems unlikely, because (a) they’re the manufacturer so they should know; and (b) the problem is intermittent. But a different manufacturer of low voltage landscape lighting recommends sizing their power supplies to about 50% more than the total wattage of lights you intend to drive. If this is the problem, a higher capacity power supply is the fix.
2) Could the fixture be shutting off due to voltage spikes in the electrical supply, even though other fixtures/appliances seem unaffected? I’m in a semi-rural area with aerial power lines, lots of trees, and pretty frequent electrical outages. Also, there was a stretch of several weeks during which the symptom occurred almost uniformly on sunny days, a little after noon. Coincidentally or not, that’s when my solar panels typically finish recharging my home batteries and instead start sending power to the grid. That’s what made me think the cause could be voltage spikes. But for the last few weeks we’ve had plenty of similar conditions—sunny days when the power flow makes that switch around noon—without the symptom happening at all.
3) If the cause is voltage spikes, could a whole-house surge suppressor fix the issue?
I’d be grateful for any advice, thoughts, input… I’m stumped.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Random problems are by far the most difficult problems to find and fix. I’ve had many commercial projects tracking down screwy electrical problems and it gets expensive and takes a lot of time.
I doubt you have an undersized power supply if you’re using the one the manufacturer of the light recommends. Lutron makes good dimmers too, so it’s probably not the problem. You’re saying that the problem doesn’t seem to be related to temperature (which, along with loose connections, would have been my first thought), but that the problem does seem to line up with things your solar system is doing. That gives me some ideas.
A little background on AC dimmers: they are what are known as “phase controlled dimmers”, which means they work by chopping up the AC sinewave using a thyristor, usually a triac. What these dimmers do is to adjust the precise time at which they turn on during each half cycle of the sinewave which results in them controlling the amount of power (not voltage) that reaches the load by varying the exact amount of time the unit is “on” during each half cycle.
The power supply has to work with the chopped up sinewave, which isnt as easy as running on a “normal” sinewave. This is where the compatibility stuff comes in, but there is something else that might be going on...
Your solar system inverter will cause small notches in the waveform when it switches in and out. These little notches MIGHT be causing some small glitches with the dimmer switch causing it to loose synchronization with the AC sinewave. This can resulted in missed half cycles where it doesn’t turn on at the correct place in the waveform. I have actually seen this when testing before, but it’s usually no noticed so it isn’t considered a problem.
I think what might be happening is that while the light, dimmer, and solar system independently are all working fine, the INTERACTION between them MIGHT be the cause of your problem. The simplest way to test this is to isolate the components and check them separately. In your case, the first thing I’d do is replace the dimmer switch with a regular on/off switch, and then watch the light over a few days or weeks (long enough to be SURE the problem would have occurred if it was going to) to see if the problem goes away. If there is no problem with the dimmer out of the circuit, then the dimmer is getting confused by something on the AC line and is causing the problem, or at least triggering the problem. I don’t think the dimmer is bad, it’s just the combination of events here that is resulting in undesired operation.
You could try a whole-house surge protector, which I would recommend you install regardless since they are good to have. I like the little Ditek units. I don’t think a protector like this is likely to solve the problem with your light though. It may be possible to use a line filter with the solar system inverter to solve the problem, but this would take some design work.
Bill
Wow, that is super informative! Thank you. It gives me a solid possible explanation and a course of action. And I’ll look at the Ditek surge protectors. Quick follow-up question: would this explanation imply that occurrence of the symptom could depend on the dimmer setting? I ask because for a while I had myself convinced it did—but now I’m not sure.
Yes, dimmer settings can play a part here. Certain dimmer settings, usually towards the lower (dimmer) end, are more critical and thus less stable and more likely to show problems. I see that in some chandeliers here that use LED bulbs and a Lutron dimmer. The problem is especially noticeable when running on generator power (I’m in a rural area too, and also have frequent power issues). The generator is less frequency stable compared to grid power, which causes the specific operating point of the dimmer to drift around a little. The result is the chandeliers flicker a little (and randomly) when on generator power.
Bill
I have a circuit of 5 LED lights in my ceiling that do not work with the gas generator - everything else does.
At least theoretically, I consider a 0-10V dimmer a better solution than ELV.
Phase dimming (which your dimmer is) is pretty rough on most electronics down stream. Even though they are designed for it, it doesn't mean it will work well.
This generally means that most will have a hard time delivering rated power on a dimmer. My guess is that the strip is too close to the max of the power supply and it occasionally trips it. You can try disconnecting 5 to 6 feet of leds temporaly and see if this solves the issue.
I doubt it is the solar inverter as most grid tie units are don't do any switching, they are always connected to the grid and transition from sinking to sourcing power without any glitches.
If you do have bad power, it could cause the inverter to go into backup mode and take over from the grid, which would definitely cause some glitching. You can check the log from the inverter to see if this is happening.
>"I doubt it is the solar inverter as most grid tie units are don't do any switching, they are always connected to the grid and transition from sinking to sourcing power without any glitches."
The inverters don't do "switching" with a relay or actual switch, but when they transition between operating modes, there will be small notches in the waveform due to step load changes. You can see this on an oscilloscope. Normally the disruption is pretty small, and causes no issues, but sometimes there can be small glitches in operations of sensitive equipment. I agree is shouldn't be an issue here, but seems most likely, unless the problem is only with the light dimmed very low, 10-20% or so, in which case EVERYTHING on the line is more likely to cause glitchy problems.
Bill
And to confirm: the problem occurs with the dimmer at 100%. I’ve never seen it happen with the dimmer at 10-20%, but that could be because I use low dimming settings much less often than high ones.
These comments are educational: I’ve been Googling 0-10V dimming to understand it. My existing phase dimmer is RF controllable (part of Lutron’s RA2 Select ecosystem). It looks like Lutron has a 0-10V dimming module in its RA2 Select line (LMJ-5T-DV-B) but It seems its not as simple as swapping the existing dimmer for that one, so I still have some reading to do. Thanks for the Dimmers 101 class, folks!
I believe you would replace your existing dimmer with a plain switch (always left on) and install the RF module near your power supply. Different but not hard.
0-10v dimming is usually much more reliable. Phase controlled dimmers work by chopping up the power supply, which works fine with incandescent light bulbs. With anything that has a ballast, which includes LED lights, phase control dimming is chopping up the power the light needs to operate, so the light needs to operate with a variable power supply AND extract the light level information from the chopped waveform.
With 0-10v dimming, the ballast runs on regular power and receives light level information from the 0-10v control line. This is much more reliable and is more common in high end commercial setups. You can get 0-10v interfaces for lutron’s smart systems if you need to integrate with an existing lighting control system.
Bill
That dimmer has the capability of being controlled by some fancy remote system. Do you specifically want that capability? If not, I'd try replacing the dimmer with something less "smart", perhaps after confirming that it works with a simple switch.
You might also try Lutron's 800 number to see if they have advice for you.
Yep, remote control via Lutron’s RA2 Select system is a must in this application. And dimming is also a must. But agreed with you and Bill that temporarily replacing the dimmer with a simple switch is a good idea to isolate the cause of the issue. And also that going to Lutron’s support resources is as well.