I saw one of these on Big Clive's YT channel. I could have used one recently if I had known. Inside is just a bunch of resistor voltage dividers, of course. So it can't achieve the labeled current on the front, because of different LED Vf. Still neat, though.

 

it can't achieve the labeled current on the front, because of different LED Vf.

There are far more types of LED's than just the small 5mm types. Some achieve pretty high currents. It depends on the voltage supply and the LED type.

So what's your question?

 

I'm thinking about making one of those, but with a current source so I can get specific currents. I have some blue 3mm LEDs that are still bright at 1mA. I had to use my bench PS to sort through my green LEDs for similar brightness.

 

I had to use my bench PS to sort through my green LEDs for similar brightness.

How are you determining brightness?

 

My two eyeballs. It was a front display panel; didn't need accuracy, just pleasing consistency.

 

A cheap LED appears to be bright when its cover focusses the light into a narrow beam.
Never feed a voltage to an LED, it will burn it out. You must limit the current with a series resistor or current-limiting circuit.

 

You can easily build such a tester. Consider and LM317L regulator as a current source as described in the datasheet. For use, I just grab a resistor (usually 1k) and connect it and a power supply or 9V battery and then eyeball the LED, mostly for color.

Years ago I had to go to Shenzhen, the LED capital of the world and wanted to find some LEDs of specific wavelengths. With me I carried a 9V transistor radio battery (ancient term, I know) with a 1K resistor soldered to one terminal and the other battery terminal and the other end of the 1k resistor connected to a dual banana plug. When a vendor offered an LED of the wavelength I requested it only took a couple of seconds to confirm (or not!) the approximate wavelength. It made the whole process a lot less stressful.

 

Consider an LM317L regulator as a current source as described in the datasheet.

Good idea. I checked the datasheet and it puts a limit on the output current as: 5mA < Io < 500mA. I'd like to get down to 0.5mA, or at least as low as 1mA. Will I run into a limit with a simple transistor current source? Or two transistors set up as a current mirror? Are some transistors better than others for this? Maybe I should just experiment and find out. I plan to run this from a 9V or several AA's.

 

A cheap LED appears to be bright when its cover focusses the light into a narrow beam.

Is there a way to objectively assess this without expensive equipment? The 3mm blue LED's that I have are quite bright even from a side angle, but I can surely compare the light cone to one of the seemingly dimmer green LED's.

 

Modern 3.3V green LEDs are much brighter than old 2V green LEDs, because they have the same chemistry and voltage as blue and white LEDs. I simply shine the LED on a nearby wall to see its amount of focusing.

 

Since you are concerned primarily with relative brightness an LDR in a dark case with a small hole to stick the led in seems promising.

 

Since you are concerned primarily with relative brightness an LDR in a dark case with a small hole to stick the led in seems promising.

It depends how you define brightness. If you clip the same area (cone angle) from all the LEDs you test, and you're ok with that, fine. It's kind of like shining an LED at a wall and looking at the brightness at one square centimeter in the middle of the illuminated field.

However, if you want to define brightness by how bright the room will be when you light the room oath those LEDs, you need to account for the area that the LED illuminates instead of just looking at the square cm in the middle of the field.

a very small 45mW LED with a 3° half power angle can be much brighter in that center 1 cm area than a 1-watt LED with a 165° Half-power angle. But the 1watt LED will light up the room much more.

 

Good idea. I checked the datasheet and it puts a limit on the output current as: 5mA < Io < 500mA. I'd like to get down to 0.5mA, or at least as low as 1mA. Will I run into a limit with a simple transistor current source? Or two transistors set up as a current mirror? Are some transistors better than others for this? Maybe I should just experiment and find out. I plan to run this from a 9V or several AA's.

You should have no problem getting down to a fraction of a milliamp with a simple transistor current source.

 

Hello,

As @DickCappels said, a simple transistor current source can go down very low:
http://www.ecircuitcenter.com/Circuits_Audio_Amp/BJT Current_Source/BJT_Current_Source.htm

Bertus