LED Strip as Room Lighting

Originally, a regular chandelier was planned as the main lighting for one of the rooms undergoing major renovation. But recently I came across the super-bright LED strip Ultra 5000 with SMD 5630 LEDs by the Arlight brand. The decision was made quickly, finally and irrevocably — I want this strip as the main light source in the room.

Theoretical Brightness

The manufacturer claims that the Ultra 5000 SMD 5630 strip provides a luminous flux of as much as 1200 lm per meter. For comparison, the luminous flux of a 100-watt incandescent bulb is approximately 1600 lm.

In my case, 15 meters of strip were to be used for a room of 14 sq.m., routed along the entire perimeter. The resulting luminous flux apparently cannot be calculated by simply multiplying lumens by meters.

I started to figure out how to properly calculate the total luminous flux, but after some googling, I realized this problem isn't solved on the fly. One needs to spend some time studying the theory of lighting calculations using the coefficient of utilization method, and find several pieces of data unknown to me:

• the reflectance of walls, ceiling, and other objects;
• the honest luminous flux of a single LED on the strip;
• the characteristics of the profile diffuser in which the strip was planned to be housed;
• the voltage drop along the strip and the dependence of the LED's luminous flux on it.

In the end, I decided to just build it and see what happens. But just in case, I also bought a dimmer for LED strips. What if it turns out to be way too bright :)

Theoretical Spectrum

Bright LEDs are certainly cool. But brightness is one thing, and the light spectrum is another.

If we again compare with an incandescent bulb, it's good because it emits light across a wide range, its spectrum is relatively uniform and partly close to the spectrum of sunlight. Such light is familiar and pleasant to the eye — it doesn't irritate or tire.

The spectrum of white LEDs differs significantly from the spectrum of an incandescent bulb, and not for the better.

White LEDs in this case are phosphor-based — the two humps on the spectral characteristic come from the blue LED emission (blue region of the spectrum) and the phosphor (yellow region of the spectrum).

Deep down I understood that one should be careful with LED lighting and also keep a regular chandelier. Who knows how the eyes will react to such a spectrum. Predicting this in advance is probably not possible either.

For color temperature, I chose the middle of the three options offered by the manufacturer — the so-called Day White, 4000 K. It simply seemed the most pleasant.

Components

Profile

I have a simple stretch fabric ceiling, without any multi-level features or cornices, and since the perimeter lighting strip was planned to be placed under the ceiling in plain view, it needed to look presentable but be as unobtrusive as possible. I wondered which profile to use — straight or corner-mounted? It turned out that the direction the LEDs face on a strip placed near the ceiling made no visual difference in intensity or uniformity of illumination at all. Whether downward along the wall, along the ceiling, or at an angle — all the same. This makes sense since the rated beam angle of these LEDs is 120 degrees, and in reality it turned out to be close to 180. Therefore the mounting angle proved unimportant, and I chose the straight profile as the most compact.

Strip

The Ultra 5000 strip comes from the factory on 5-meter reels, with a cut interval of 10 cm; the store sells it cut to the nearest meter. In my case no cutting was needed — I took three full packages.

The strip width is 12 mm — it doesn't fit into the chosen profile in a completely standard way, but it fits.

Power Supplies

The rated power consumption of 15 meters of strip is 240 W. Calculated from actual measurements — 180 W (I measured on a 3-meter section, the consumed current was 3 A).

But besides power consumption, there's the factor of voltage drop along the strip, which leads to a gradual decrease in brightness toward the end. Power supplies for LED strips (all or not — I don't know, but the ones I bought — yes) allow powering a single common load in parallel. To equalize brightness along long powerful strips, instead of one power supply from one end, you connect two less powerful supplies from both ends of the strip, and in particularly severe cases — also in the middle of the strip. In my case the perimeter is closed, I split it in half and simply took two 130 W power supplies, connecting 7.5 meters of strip to each one separately.

The choice of sealed power supplies was driven by the fact that they have significantly smaller dimensions compared to open-frame ones and don't have fans — that is, they're silent, which is important. Plus, I planned to place them all in a sealed (for fire safety) enclosure located in a hidden spot inside the sliding wardrobe, where heat dissipation would be a problem.

Dimmer

The dimmer that I liked the most in terms of features among those in stock was originally designed for wall mounting.

This dimmer has both mechanical brightness adjustment and remote control via an IR remote. Moreover, besides smooth adjustment, the remote has four preset brightness level buttons (25%, 50%, 75% and 100%), and four more buttons for programming custom levels.

But it requires four wires for connection, which I hadn't laid into the walls. So I decided to install the dimmer in the enclosure along with the power supplies. I'd use the mechanical adjustment to set only the startup brightness level when the strip turns on, and adjust the desired brightness with the remote control.

But for this I'd need to desolder the IR receiver, run it out of the enclosure on a wire, and place it in a convenient, inconspicuous spot. Would it work like that? I tested it — it works.

Looking ahead, I'll note that the dimmer has one significant drawback. The memory of the brightness level set before turning off is handled by the potentiometer. When 12 volts are applied to the dimmer, the strip turns on at the brightness level set by the potentiometer. After that, brightness can be changed both from the remote and with the potentiometer. But regardless of the potentiometer position, in the first instant when the dimmer is powered on, it doesn't immediately start PWM, and raw 12 volts come through. At the moment of switching on, the strip necessarily flashes at maximum brightness for a split second, and then settles to the set level. It's an unpleasant hit to the eyes.

Amplifiers

The dimmer's power rating turned out to be insufficient for my strips. I had to additionally buy amplifiers — one for each power supply.

Enclosure Assembly

Besides the main lighting, I decided to use another 3 meters of the same strip for local lighting above the wardrobe, with a separate wall switch and limit switches in the sliding doors. And I also had decorative backlighting from a simple LED strip and a nightlight on a short section of dim strip with a photorelay. None of this is directly related to this article, but since it required placing three more different power supplies in the enclosure, it's worth mentioning. The photorelay originally has rather large dimensions and an unaesthetic appearance, so I also wanted to hide it in the enclosure to keep it out of sight.

I treated it the same way as the dimmer — desoldered the sensor and ran it outside on a wire, after first verifying that this too would work.

I found a sealed enclosure of suitable dimensions.

Stocked up on brackets and hardware.

And began assembly.

The enclosure is ready.

Profile and Strip Installation

To increase light output, it would be desirable to mount the strip not flush against the ceiling but slightly below, at least 5 cm. In that case the reflection of the strip's light from the ceiling would be better. But I didn't have that option due to some subjective reasons, so I mounted it flush against the ceiling.

On the left, a piece of profile for the local lighting strip above the wardrobe, mounted on the lower edge of the cornice, behind which the decorative backlight strip would be placed.

Cutting the profile with a hacksaw with fine teeth evenly at a 45-degree angle is easy if you use a miter box.

I attached the profile with drywall screws 32 mm long directly into the drywall, without anchors (the drywall is glued to the walls with Perlfix), after pre-drilling holes in the profile at half-meter intervals.

Then I laid the strip into the profile and soldered the power wires. I didn't close the perimeter profile with the diffuser yet (the store hadn't yet delivered part of the diffuser) — I only closed the local lighting profile above the wardrobe.

First Power-On

And so — we turn it on!

Wow! This is amazing!

Of course, I wasn't blinded by the brightness. Bright, yes, but not over the top. And very beautiful!

With the uniformity of lighting visible around the entire room perimeter, the strip provides even illumination everywhere.

I attempted to capture the lighting difference between the strip and a 200 W incandescent bulb that I temporarily have instead of a chandelier. I locked the camera settings under one light source, set the camera to burst mode, and meanwhile switched the light source. Here's what I got.

First, I adjusted for the incandescent lamp light and started the burst — first photo is the lamp, second is the strip. Then the reverse — first the strip, second the lamp.

An interesting effect — under the strip's light, vertical shadows are nearly absent. This is visible, for example, from the shadow of the horizontal pipe and the missing shadow from the vertically positioned regulator handle on it.

From these photos you can see that 15 meters of strip is brighter than a 200 W incandescent bulb. But seemingly not by much. In reality, comparing like this isn't quite correct. The bulb is a point source, while the strip is distributed. Under the lamp's light, the corners of the room are much darker than the center, while under the strip's light routed along the perimeter, it's equally bright everywhere.

I took illumination measurements with a lux meter.

Here are the results in numbers:

As the table shows, illumination under the strip's light varies insignificantly between different measurement points in the room — whether at the floor, at eye level, in corners or in the center — the difference is no more than 2-3 times. This is, of course, a consequence of the even distribution of many point light sources along the room's perimeter. Which is not the case with a bulb hanging in the center of the ceiling — the difference at different points reaches almost 30 times.

When both the lamp and strip are turned on simultaneously, their illumination levels simply add up.

If someone could explain to me in simple terms why lumens don't add up directly (or do they?), while lux values do add up (could this be a consequence of the large spectral difference between the LED and incandescent lamp in the long-wavelength region?), and how with a rated luminous intensity of 1200 lumens per meter, measuring at 1 meter from a 4-meter strip section gives only 530 lux, I would be very grateful.

On a short section, I put the diffuser on the profile and photographed the difference at a short exposure.

I measured illumination at a certain distance: without the diffuser 600 lx, with the diffuser 520 lx at the same distance. It absorbs more than 10%. Unfortunately, there's no diffuser yet for the entire perimeter — can't evaluate the overall reduction in illumination.

Voltage Drop

The voltage along the strip drops significantly.

At the beginning it was 11.5 V, and at the end of a 7.5-meter section it was already 8.5 V. That's 0.4 volts per meter.

The brightness drop doesn't jump out at you, but if you specifically compare, you can see that in one corner the strip is much brighter than in another.

Lux meter measurements at approximately 30 cm in opposite corners of the perimeter gave results of 1600 and 600 lux — a difference of more than 2.5 times. Measurements at other distances gave the same 2.5-3x difference. Therefore the 530 lux value in the table for measurement at 1 meter from the strip is a rough average — I measured at approximately 2 meters from the strip's beginning.

Strip Heating and Enclosure Temperature

The strip heats up, and heats up noticeably.

At the beginning of the strip, the aluminum profile temperature was 55-57 degrees C, but at the end it was already quite cool, around 30 degrees C. With the diffuser installed on the profile, the temperature didn't differ significantly.

Inside the sealed enclosure at maximum load (which in reality probably won't happen), the temperature also didn't rise above 57 degrees C after a 4-hour burn-in. This slightly exceeds the rated operating temperature of the power supplies, but nothing started smelling burnt, everything worked. In normal operating mode, when only the perimeter strip is on, the temperature in the enclosure settled right at the rated value of the power supplies — 45 degrees C. Quite satisfactory.

Summary

I'm quite satisfied with the technical side of the LED lighting system. Convenient and simple installation, quality components, instant light when switched on (which not all LED strip power supplies provide), silent operation, moderate heating, low power consumption with high light output. Only two negatives, but both are solvable in principle — voltage drop along the strip (though the total room illumination is quite sufficient as is), and the dimmer's incorrect behavior at power-on (could be solved by introducing a delay circuit for the strip after powering up the dimmer, but then instead of the flash there would be a delay — I'm not sure which is better).

It's too early to talk about the operational side — that takes time.

First impression — this light is completely different. And you can't say it's better or worse. Just different.

I spent several days assembling furniture under the strip's lighting and felt no discomfort whatsoever. Overall, I like everything so far.

I didn't like the financial side — the entire system cost me approximately 20,000 rubles. The cost of one meter of strip in the profile is approximately 1,000 rubles. Plus power supplies and other equipment. Although it's unknown how long all this will last. LEDs are rated for about 100,000 hours of service life, at which point they lose up to 30% brightness. If the strip is used on average 5 hours per day, it should last about 50 years. We'll see.

Below are several sub-posts not directly related to this article's topic, but closely connected to it.

Controlling Separate Switching of Chandelier and Strip

Background

Not knowing in advance whether we'd like living under LED light, as I already mentioned, it was decided to use both the LED strip and a regular chandelier together.

But always switching both on in parallel is silly. Re-wiring connections in the switch isn't great either. I wanted the ability to separately turn on the strip, the chandelier, or both together as desired. Unfortunately, the LED strip appeared after the wiring was already laid into the walls. And the overhead light now consisted of two different sources that I wanted to control independently.

I'm extremely skeptical of any wireless solutions in this area. Yes, various ready-made systems with controllers and wireless remote switches exist. But I consider that approach overkill, or a last resort when no other options exist. And the need to change batteries in remote switches wasn't appealing. A simple, robust solution was needed.

Originally, three control points were wired — above the sofa and at the room's two doors (it's a through room). From each point, the overhead light could be independently turned on or off. Above the sofa, additional secondary light sources (decorative backlights, sconces) could also be turned on, creating various lighting scenarios.

Solution

After racking my brain over the possibilities of the existing wiring, I came up with a scheme where from all three points you can turn off absolutely all light sources with a single press. That seemed more convenient when you need to quickly kill all the lights. In the main control point (above the sofa), each light source additionally had its own separate switch, including a separate one for the chandelier and a separate one for the strip.

But one pitfall arose that could cause some inconvenience. If at the main control point all light sources are turned off by their individual switches, then the switches at the doors wouldn't be able to turn on any light at all, and you'd first have to navigate through the room in the dark to the sofa.

A solution was needed where it would be possible from any point to definitely turn on at least one light source, regardless of whether it was previously turned off by its individual switch or not. Using the same already-installed wiring, and without resorting to excessive wireless technology.

The solution was found in using a standard electromagnetic switching relay, employed as a memory cell with auto-reset. The lighting control logic now works as follows.

Operating Algorithm

We decide which light source will be the "primary" one — that is, the one that turns on by default under any conditions. And we connect it to the power supply through the relay's NC (normally closed) contacts. To disconnect only this "primary" light source, instead of a simple toggle switch we use an NO (normally open) push-button switch. The button controls the relay activation, which latches itself, simultaneously breaking the power circuit of the "primary" light source. It stays in this state until all lights are completely turned off, after which it automatically resets to its initial state, closing the NC contacts and allowing the "primary" light source to turn on again next time.

This solution has one drawback — to turn the "primary" light source back on after disconnecting it with the button, you need to flip the main switch twice. One could of course build a T-flip-flop from several relays, but the simplest reliably working circuit would require as many as 5 relays, and the flip-flop's starting state would be undefined and could be either one.

It was obvious that modifying the circuit to ensure a defined starting state would require several more relays, ultimately resulting in an unreasonably bulky construction. So I decided to live with this drawback for now.

Final Implementation

The schematic for controlling two light sources from three points with source selection at one point is as follows.

I chose the strip as the "primary" light source for now — if I don't like it, re-wiring later isn't difficult. When light is turned on at any of the three points, the strip always turns on. The chandelier has its own simple toggle switch. The strip can be disconnected with the push-button switch.

Making a push-button from a regular toggle switch is very simple. For such transformations, my favorite series of electrical installation products, the Schneider Electric Unica, is perfect. Switches of this (and perhaps not only this) series have modular construction — one or two narrow modules of various functions can be installed in each post (per frame). The modules themselves are assembled from standardized parts. Using parts from different modules, you can build practically anything. Everything is held by clips, disassembly is easy down to individual parts, and reassembly is just as simple.

This series has a standard item — Blinds switches, catalog number MGU5.207.18ZD. The only product that uses non-latching push-button switches and has the right-sized springs :) From there, it's up to your imagination — many options. I won't go into disassembly details; anyone who dives in will figure it all out — it's straightforward. And since I've already done switch and outlet reconfiguration with Unica more than once, I've built up a parts inventory of partially disassembled modules. From what I had, I assembled a two-switch unit where one switch worked as a toggle and the other as an NO push-button.

The finished control panel at the main lighting control point has: on the left — the master toggle for all light sources simultaneously. Second — the strip disconnect button. Then three switches for various backlights (not shown on the schematic). Last — the chandelier switch.

At the doors — single one-button switches, one simple and one crossover.

The relay is housed in a concealed junction box in the wall.

I use Taiwanese TRY-220VAC-S-4C relays. A group of these relays has been working in my multi-level lighting system in another room for two years already, without complaints. Ideally the relay should be in a socket, but I didn't have one on hand, and they're huge anyway — I soldered it on wires.

Testing of this system was successful — everything works as designed. It's too early to comment on usability — I just finished it recently.

Wardrobe Local Lighting Control

As I mentioned above, I used three meters of Ultra 5000 strip for local lighting of the sliding wardrobe, from a separate power supply and with separate switching. And for this, a somewhat unusual control scheme was needed.

This strip should turn on when the wardrobe doors are opened. Limit switches positioned inside the wardrobe above the doors are used for this. I found roller-type ones — nothing unusual here.

But I thought a situation might arise where the wardrobe needs to be opened but the light isn't needed. And the reverse too — turn on additional lighting without opening the wardrobe. For this I decided to install a separate two-button switch: one button would force the light on even with the wardrobe closed, the other would force it off even with the wardrobe open.

The resulting schematic is as follows.

The wardrobe itself is still in the design phase, so there are no photos. But the lighting already works from the switches. The two-button wall switch is completely standard, unmodified, connected with three-conductor cable. The forced-off button has priority — if it's off, nothing will turn on. The forced-on button will turn on the light only if the first button is also on, regardless of the wardrobe door position. The "automatic" mode of light switching from the limit switches is possible only with the first button at "on" and the second at "off."

LED Decorative Backlight as Main Lighting

Decorative LED strip backlights were planned in advance. One of the objects to be backlit was the modular furniture (a set of cabinets). But the degree of illumination was hard to predict in advance. For this purpose, I bought the dimmest strip I could find.

I planned to place it under the furniture, above the furniture, and inside it too, of course :) And already during installation of individual strip sections and test power-ups, it became obvious that firstly, it looks absolutely stunning, and secondly — it's way too bright :)

Having installed only the lower and partially the interior backlighting, I got a result that mildly exceeded my expectations. Here's how it looks when both the backlight and the powerful perimeter strip are on.

And here's what it looks like if you turn off the overhead lighting and leave only the backlight.

In these photos, the actual light level isn't very clear. In reality, it's quite bright. And that's only a small portion of what's planned :) And already this level of illumination can serve not just for decoration, but as quite comfortable soft light for everyday living. This makes me think that when I finish all the backlights, it's very likely that this light will be used most often and become the "primary" one :)

UPDATE January 27, 2013

I finally received the long-awaited diffusers for the profile. Installing them turned out to be a very labor-intensive task — you need to apply very significant force to snap them into the profile. It's inconvenient doing this up near the ceiling. A couple of meters is one thing, but over 15 meters I even took a couple of breaks since my fingers usually don't press anything heavier than keyboard keys :)

The result is mixed.

It did get darker, noticeably darker. I took measurements again before putting on the diffuser and after — here's what I got:

So it got about 30% darker.

But on the other hand, the light itself changed, and changed for the better. First, it became warmer in color temperature. Even during diffuser installation I noticed a difference in color on the ceiling — where the diffuser was already installed, the light had a more yellow tint. For the eyes, this color became even more pleasant. Second — perhaps as a consequence of the first — color reproduction improved. Previously, everything seemed to have a greenish tint under this light, but with the diffusers that effect disappeared, and object colors became more natural. And third, the strip became more aesthetically pleasing — not contrasting bright dots on a dark background, but more diffused on a lighter background.

Overall, I'm satisfied with the result. The brightness is now comparable to illumination from a single 200-watt incandescent bulb, but the light quality is much better. The maximum level on the dimmer no longer seems excessive, but quite normal. Possibly, when the strip eventually loses some brightness over time, it will make sense to swap the frosted diffuser for a clear one. Or remove it entirely.

UPDATE March 9, 2017

I added a review of four years of operational experience in the article comments.