I find that there is a point at which you need to switch from doing academic research to carrying out some practical experimentation. I reached that stage on Sunday evening while investigating whether LED “strip” lighting would be a good replacement for the traditional fluorescent work surface lighting.
I realised that I didn’t have any feel for how bright any of these tapes actually were, or if their characteristics would make them good or bad sources of light for work surfaces. I’d also had suggestions from contacts on Twitter that it was a lot harder to dim these strips than I was expecting.
But it transpires that in my study I happen to have a run of wall cupboards over my desk which closely mirrors a kitchen layout. The desk is much deeper, and the cupboards are mounted higher than they would be in a kitchen, but the principle is the same. If anything, my study would be a more challenging environment because of the higher mounting point and larger area to illuminate.
So I ordered 2 meters of moderately high output “dimmable” LED strip; this is built with 60 cool-white SMD 5050 LEDs per meter of tape, operating at 12v and drawing a little under 15w a meter. It’s all encapsulated in a silicone coating, and backed with a 3M self-adhesive coating. I added a 33w “TRIAC dimmable” LED driver that someone had reviewed as working successfully for them, and a Varilight V-Pro low power dimmer switch.
The advantage of that dimmer switch is that it can run with a minimum load of only 10w, unlike normal dimmer switches that usually require a minimum load of 40w or more. It’s also a “smart” dimmer switch, where the mode (leading/trailing edge dimming) and minimum brightness point can be “programmed” into the switch.
I’ve just set it all up “loose” on my desk, and it works quite well. The LED strip is very bright; more than sufficient to light a work surface under a kitchen cabinet. In fact, for my immediate “test” application in my study, it would be too bright without the dimmer.
The dimmer works very well in trailing edge mode. It’s completely silent at minimum and maximum brightness points, with only a very slight buzz from the LED driver at the mid-point. Minimum brightness is (subjectively) about 25% of the maximum, and control of the light level within those extremes is very smooth. Perhaps the only issue is that turning the LEDs on with the dimmer at anything other than full brightness seems to take a second or two for the dimmer to fire everything up. Noticeable, but not necessarily a problem.
Out of interest, I also tried the dimmer in leading edge mode; it wasn’t a good experience. The LEDs did not dim very much, and the LED driver produced a much more noticeable buzzing noise. Trailing edge is definitely the way to go, at least for this set up.
So, the summary is that the LED strip tape is fine for my intended use. I will almost certainly need to be able to dim it, and I now know that is possible without resorting to expensive professional remote-controlled low-voltage dimming, even though it’s perhaps not as ultimately good. From what I can see of the problems with this type of solution (noise, failure to start the LEDs) are probably all related to the LED drivers, so finding ones that are known to work well is going to be the key to success.
Update: Having had this system properly wired into my study for the last few days, a small issue has arisen; the adhesive backing tape (which I think is probably double-sided adhesive tape that is pre-applied as part of the manufacturing process) is not proving man enough for the job. About 2 days after I initially applied the strip to the bottom of the overhead cupboards, it it started to peel off. Pressing it back into place makes it stick again for a while, but it’s definitely not a good long-term solution.
At the moment I’m looking for a better fixing system, but the fact that the tape is encapsulated in silicone is not helpful, as getting anything to adhere to it is problematic. My first thoughts are around using a silicone-based adhesive/sealant to stick it up … but it will need supporting in place while everything cures, which could be interesting. More research required.