Pseudo LED Load Board

[matt-smithlabs]

I'd like to make a load board to test LED drivers. Ideally this would consist of a string of actual high-power LEDs for demonstration purposes as well as an alternative string comprised of "decoy LEDs" so the part can be characterized without blinding the tester. The equivalent model need not exhibit the exact same behavior as a particular LED, but it would be great to find something that mimics the IV curve of a standard use case.

Note this is a high-power application. A worst-case scenario will drive 1 string of 16 LEDs at I_max=3A, V_fwd_max = 3.5V. I am hoping to avoid deferring to a simple resistor network as this board will be used for a wide range of power requirements, i.e. driving at 100mA using the last example will not yield the V_fwd expected of the LED.

Looking forward to your thoughts!

 

[JimB]

as well as an alternative string comprised of "decoy LEDs" so the part can be characterized without blinding the tester.

Why not just have an opaque cover for the test board?

JimB

 

[Pommie]

Agree with Jim but remember you need adequate cooling.

Mike.

 

[matt-smithlabs]

Yea that was the other option, maybe that is the right way to go. I have a board I have been working with that uses a slab of tinted polycarb on standoffs above the LEDs but it does little to occlude light of that intensity. I can wrap it in something to damp it further but I don't want to compromise the quality of the presentation in front of customers. Could you suggest a suitable material?

What I had in mind was a board that I can hook a current monitor up to on the front end with a dummy load comprised of some TBD network and parallel low-power LEDs driven by current sense -> BJT to indicate current flow through the dummy load string.

See attached

 

[dougy83]

If you want an electronic solution, maybe you can use a small LED to simulate a string of larger LEDs by buffering its current, when it's fed from some fraction of the LED drive current, as per the attached diagram. Adjusting the pot sets the number of LEDs, while changing the ratio of R1/R2 will set the current multiplication factor of the load.

Just covering the LEDs does sound a lot easier though

 

[JimB]

Could you suggest a suitable material?

Metal.
Sheet aluminium is easy to work with.

JimB

 

[Nigel Goodwin]

Metal.
Sheet aluminium is easy to work with.

Cardboard is even easier to work with, as is hardboard or plywood.

 

[JimB]

Cardboard is even easier to work with, as is hardboard or plywood.

True.

But, there is the possibility that this thing could get quite hot, cardboard, hardboard and plywood would not be my choice for such an application.

We also have this little snippet of information:

I can wrap it in something to damp it further but I don't want to compromise the quality of the presentation in front of customers

So, rather than a bit of cardboard and a wet rag to stop it catching fire, use a metal cover or box.
If necessary use a small fan to move the air through the box to keep it cool.

JimB

 

[matt-smithlabs]

I think I will take your collective suggestion to simply forego the dummy LEDs and stick the real thing on the board with an opaque slab. Thanks all for your input.

 

[roy_witkey]

paqu

I think I will take your collective suggestion to simply forego the dummy LEDs and stick the real thing on the board with an opaque slab. Thanks all for your input.

Aluminum based board (LEDs planted on it) would be good choice