Durability LED lamp and the luminous flux decline

  • Dec 24, 2019
click fraud protection

I conducted an experiment and compared, it reduces the light output of various LED lamps for forty minutes after switching on.


Before any testing I will warm up LED lamp at least 30 minutes. The fact is that any LED lamp during this time reduces its light output as warm LEDs and electronics.

With the help of self-made "black box" Lamptest-1 I measured the variation of the luminous flux twelve different LED lamps in the first forty minutes of operation. For clarity, I have selected nine of them:

IKEA Ryet 603.979.60 LED1741G6;
OSRAM LEDSCLA60 10W / 827 230VFR E27;
ASD LED-A60-std 7W E27 3000K 630Lm ASD;
Camelion LED ULTRA LED6-A60-FL / 830 / E27 filament;
Lisma SDF-6 A50 filament;
toshiba 00101315012B;
Geniled 01,327;
Gauss 102802115 Graphene filament;
Saffit SBA6010 55004.

Here's what happened.


The first thing that catches your eye - a significantly larger drop in brightness (hereafter I will use that word, not as a technical term, and as domestic concept) in Saffit lamp. After 25 minutes, the luminous flux of the lamp dropped to almost 70% of the initial. Note that this is a very, very bad bulb (10 watts stated, in fact 5 W, CRI stated 80, 72 in fact, the light output is not measured, but I think he is also in the area of ​​half the declared). This decline suggests that most likely the LEDs operate in overload mode. How long this will last is not light.

instagram viewer

The second interesting point - unlike the graphs of three filament lamps. They quickly warm up and stabilize quickly. At Camelion lamp light output has stabilized after only five minutes after the start and then fell by only 1% for 35 minutes.

Interesting graph from the old lamps from OSRAM with the heaviest cooler (lamp weighs as much as 117 grams). It heats up so slowly that even 40 minutes is not enough for a complete stabilization of the luminous flux.

Wavy graph at IKEA Ryet light caused by pulsation (ripple her about 13%).

Possible degree of luminous flux falling can make a prediction of longevity lamps and if 15 min brightness has dropped by more than 20%, the probability of premature lamp failure rather high.

So far, I have no way to test because each lamp, but if you do for Lamptest-1 special firmware, and the devices themselves to produce a few pieces, the process can be fully automated and put on flow.

© 2019 Alex Nadozhin