High current step down CC drivers

[clint357]

Does anyone know of any constant current drivers which accept a higher input voltage than the output? These would be a huge step in the right direction for automotive applications.

 

[LEDobsession]

I believe these are constant current.

Correct me if I'm wrong.

 

[clint357]

Input voltage only goes up to 4.6v though. I'm really looking for something to power the SST-90.

 

[Illum]

EDITED
Good observation spencer

 

[spencer]

Having posted this in the automotive section, I believe he wants something that can take 12V give or take.

 

[TorchBoy]

What do you consider "high current"?

You could use several AMC7135 linear regulators in series and they would cope with the variation in input voltage quite nicely. I don't think a single AMC7135 with dropping (LE)diodes would have enough flexibility.

A single buck driver able to handle 10-15 V input would be a more elegant solution for 1 to 3 LEDs in series. There are some nice ones available that even have built in rectifiers, which is nice if making drop-in replacement bulbs.

Should this thread be merged with your other? They seem to be addressing the same issue.

 

[TorchBoy]

What do you consider "high current"?

I'm really looking for something to power the SST-90.

Hm...

If you want to run it at a full 9 amps you might have to settle for three drivers in parallel.

 

[clint357]

I'm trying to get away from the 3 driver setup, but it might be my only option right now. I would like to get one of these setups while I still have my Bridgelux to make a good comparison between both of those and a p7. BTW, how exactly do you wire the 3 drivers in parallel?

 

[TorchBoy]

With many (most?) buck drivers you just connect the outputs in parallel. A few are funny though (because of how the output is rectified or something :shrug so adding more drivers doesn't give as much extra current as you'd expect. It's something to check when you find a driver you like the sound of.

 

[clint357]

I may get 4 drivers from Kaidomain to run in parallel. 2.8A x 4 = 11.2A. I know that max it 10A, but I'd be surprised if it's actually that high....actually, I happen to have 4 or 5 of those at home right now. It just sucks to have to wait to get the SS-90 since they are always sold out.

 

[Icarus]

Take a look at de buck driver from Der Wichtel.
Output current is 2.8A but I suppose they can be wired in parallel.

HipCC from Taskled can also be used but with input voltage = 12V you need to wire 2 or 3 LED's in series to get good efficiency.

 

[bshanahan14rulz]

just put 3 in series and make a simple driver and call it a day. a really bright day.

 

[clint357]

Can you test the output with a multimeter? I want to see if I actually need 3 or 4.

 

[TorchBoy]

To test how much current the driver will spit out at any particular voltage (such as the operating voltage of your LED) is easiest to do with the LED itself. However...

If you assume the Vf of the LED will be around 3.9 V then at 10 A that's 39 W. You want a resistor to drop 3.9 V across it while dissipating 39 W. From Ohm's Law you know R = V/I so your resistor needs to be 3.9/10 = 0.39 ohms. You might be able to make one yourself from toaster wire.

:tinfoil:

 

[spencer]

Here.

 

[clint357]

Here.

Looks like you just about have to build the thing yourself. It would be nice if they just put a trim pot on it and put all the other stuff inside to begin with.

 

[spencer]

It would be nice if if was ready to go but thats not what they do. Anyway its a good project to make one. That would be a quality driver for cheap.

 

[clint357]

It would be nice if if was ready to go but thats not what they do. Anyway its a good project to make one. That would be a quality driver for cheap.

Yeah, but I would want the option to be able to make a lot of them and I don't think it would be worth all the time....plus, that's a little over my head. I bet the guy at taskled.com could make some.