Using the cheapest light meter or laboratory work on the effectiveness of LEDs

Since I have long been interested in LED lighting, I needed a light meter.
For some time I used the light sensor in the phone and, in principle, it worked well, but a light meter would be better. And so I finally bought it and started measuring everything.
In this review, a simple alteration of furniture fixtures from a 20-watt halogen light bulb to an LED and a small laboratory work - the dependence of brightness on power, current, temperature, etc.

If you are tired of the endless series of advertising reviews p.18 and discounts (including mine) - read on
LED lamp needed in the closet. In the visor of the cabinet were two flat 20-watt halogen lamps that were used as a night lamp. Since a 20-watt night light is eating too much energy, it was decided to convert it to LED.

For me, the easiest way was to redo the lamp by replacing the halogen with a 10-watt - 9 V, 900 mA LED. I already had one.
I disassembled the lamp, pulled out the halogen and the base from it.

I started to test the LED. I thought that it would work normally without overheating at a current of 300 mA, this is only a third of its power. But it was not there. I connected the LED to the "3 VT - 300 mA" driver and after 20 minutes I burned a finger about it. His saliva hissed. The mess. I reduced the current in the driver by changing the current-setting resistor, the current decreased to 185 mA. The LED still warmed up, but not so much.

I decided that the lamp would still coolLED, glued the LED to the aluminum reflector in the luminaire and decided to check the LED modes - what will be the brightness increase with increasing current from 185 to 300 mA and how the brightness will decrease with increasing temperature. That is, if you increase the current on the LED twice, will there be more light? Or maybe less?

Since I can not fully appreciate the lightflow from the LED, I just put the reflector of the lamp on the matte sphere of the light meter, turned it on by a factor of 100 and estimated the number of lux with such a constant position of the LED. At the same time, the light meter shows a value directly proportional to the number of lumens that the LED emits. We assume that the light meter shows the light flux from the LED in conditional parrots. If the readings on the light meter were twice as large, then the LED began to shine twice as bright. I hope the principle is clear.
I also attached a thermocouple to the rear surface of the reflector.
The M14 nut in the photo simply serves as a clamp.
So, the first tested LED with a driver for a current of 185mA.
Cold start - the light meter shows 1339 lux.

After some time, the LED with a reflector warmed up, the temperature stabilized by 56 degrees. The light meter shows 1261.

Total during warming the LED lost 1339 - 1261 = 78 which is 5.8%.
Not so much.
The maximum that managed to catch is 1261
The current and voltage parameters can be seen in the photo.
Total it is 8.78 V X 0.188 A = 1.65 watts.

That is, at a current of 185mA, the LED consumes 1.65 watts and gives out light to 1261 parrots
So, the second one is a LED with a 300mA driver.
Cold start - the light meter shows 1949 lux.

After some time, the LED with a reflector warmed up, the temperature stabilized at 81 degrees. Light meter shows 1687

Total during warming the LED lost 1949 - 1687 = 262 which is 13.4%.
Already a drop in brightness is much larger.
The maximum that managed to catch is 1687
The current and voltage parameters can be seen in the photo.
Total is 9 V X 0.294 A = 2.65 watts.

That is, at a current of 300 mA, the LED consumes 2.65 watts and gives out light to 1687 parrots
Now let's compare how much the LED was able to give out light with one driver and with another.

The first current of 188 mA and light on 1261 parrots, power 1.65 watts
The second current is 294 mA and light on 1687 parrots, power 2.65 watts
The current increased by 1.56 times, the brightness increase by 1.33 times.
Count with power
Power increased 1.6 times, more light became 1.33 times.
With a driver of 188 mA, one watt on the LED gave 764 parrots
With a driver of 294 mA, one watt on the LED produced 636 parrots.
For those who are still interested in the light meter, just see

The result - with increasing current, the LED's efficiency drops. The temperature rises and the life of the LED decreases.
If you want more light in the lamp - you canEither change the LEDs more often (they will burn out or degrade), or organize better cooling or supply a better LED with large crystal sizes. Or put a 100-watt LED instead of a 50-watt LED - both the luminous flux and the service life will increase.
And a bonus - by the percentage of brightness drop during heating, you can indirectly judge the degree of heating of the LED.
The fall
- 3-5% - good operating mode, warm LED.
- 5-15% - tolerant, hot LEDs
- 15% -20% - such lamps for several monthsdegrade smoothly and imperceptibly, the light goes into the violet part, it seems they have not burned out and are eating the same way, but it seemed to be hard to see in this room. And then you screw in a spare lamp from the same batch and it shines several times brighter. It seems to work and you can’t hand over the warranty. But not bright.
- more than 20% - will not work for a long time.
To assess the luminous flux drop of the lamp, it is quite possible to use the LuxMeter application - there is a maximum, minimum and a schedule.

Well, of course, the cabinet, of course, is halogen on the left, an LED with a 1.65 watt driver on the right.

readings of the light meter close to the glass on a factor of 100 X 428 for the LED and 840 for the halogen.
For a long time like that there was no cabinet on the muska.