Hello,

I have a large 2.2V 30A diode that I need a DC power supply for. From what I have researched, this would be either a constant-current source of some kind, or a dedicated LED driver. I have searched through the Texas Instruments offerings and I think the LM3434 comes closest, but it still only provides 20A continuous.

What I am looking for is suggestions for either a single-controller solution, an alternative method of powering the diode, or maybe a parallel-supply capable LED driver if one is available. I have tried converting a regular buck SMPS to a constant-current supply, but I ended up frying a somewhat expensive board and I would prefer to not go with that option again without more guidance.

Thank you,
Obli

 

If a method of safely coupling them is available, I think using several LM3409's could work.

 

I found this article about parallel LED driver coupling: https://www.maximintegrated.com/en/app-notes/index.mvp/id/4537
Could this be scaled up for my high current requirements?

 

One of these perhaps?

 

One of these perhaps?

Thanks for the response.

Right operating ranges and the price is half-decent too, but does it function as a constant current source?

 

but does it function as a constant current source?

No . It does have current-limiting for self-protection, and it does have a feedback pin, so it should be possible to design an add-on current controller.

 

add-on current controller

That sounds do-able. Do you happen to have a link to something like that? A schematic, or an IC perhaps?

 

Sorry, I don't.
Doesn't the LED manufacturer have any suggested drive arrangements?

 

Sorry, I don't.
Doesn't the LED manufacturer have any suggested drive arrangements?

None that I know of.

 

Would something like this help?
http://www.digikey.com/product-deta...ated/ZXCT1107SA-7/ZXCT1107SA-7DICT-ND/3677922

 

It could do, though you'd need a low-resistance current shunt, a voltage reference and a comparator to go with it and you'd have to find a way to interface the comparator output with the voltage-regulator in the link I posted. It would be simpler and probably cheaper to use an off-the-shelf dedicated constant-current LED driver if you can find a suitably rated one.

 

Would something like this help?

it should work. but your chip (lm3409) does have a high side sensor so the part here wouldn't be as useful there.

it seems to me a far easier route is to go with multiple chips, each rated at lower amps and driving a subset of leds. like 6 leds in serial (<14v), whatever the amperage those leds want.

that way, you are working with much lower current, and the power dissipation on the current sense resistors isn't that much of a concern.

 

How about this for an off-the-shelf solution?
It's very close to my requirements and pretty cheap. How hard would it be to efficiently drop the last ~3 volts? (preferably with some adjustability)

 

Clearly an LDO would not be appropriate for such loads, so it sounds like another SMPS would be needed. Kind-of brings us back to the original problem if that is the case.

 

I can't see anything in the write-up of the item in your post #13 link to suggest it has a constant-current mode (which is what you need for LED driving).

 

I can't see anything in the write-up of the item in your post #13 link to suggest it has a constant-current mode (which is what you need for LED driving).

Thanks for pointing that out.

I'm going to try emailing them about the driver and the possibility of a custom output voltage (they do state that they do customization). I'll post when they give me a response.