I got a few 30W, 50W and 100W LED floodlights, that have the Ebay-style square LED module (square of Phosphor, in which the LED array is potted). The LEDs in them are 10series x 3/5/10parallel respectively

Drivers are spec'd "30W" or "50W", the 100W floodlight uses 2 identical "50W" drivers with the outputs connected in parallel

The drivers are potted Switchmode constant current power supplies, rated "900mA 27..36V" 30W or "1500mA 27..36V" 50W

I want to convert the floodlight to IR LEDs, which have the same form factor but 14..16V forward voltage. How would that affect the drivers ? Is that acceptable for them ? How far off could the current regulation be ?



In addition, the 100W floodlight is damaged. The LED midule shows cracks in the Phosphor. When switched on, it first works with only some of the arrays in the chip lighting and at uneven brightness, then starts to go out when it heats up. How can i know whether the drivers are still intact ?

What I would imagine happens is the driver/ballast tries to get the voltage low enough, but it can't get it low enough to compensate for the current, so the amps would be higher, unless the driver cannot supply that much current, so it would likely overheat.

As it is a Switch mode power supply, i think it will try to regulate to the correct current anyway. Also, as i clamp the voltage too low, the epected output power will be lower than intended, so i dont see how anything in there can be stressed

What i expect might happen is, that at too low load the power supply cannot stabilize the supply for its internal circuitry, and might start cycling on and off

I wonder if anything else which i fail to foresee may happen. So, if i connect it with the IR LED and it appear to work fine, whether this means that nothing is going to get damaged as result

If it just goes cycling, i can sort this out by adding a resistor or few ordinary rectifier diodes in series with the output to increase the voltage just untill it gets stable again

If the voltage is below minimum, it will be just flashing. The reason is, ther controller uses to be supplied from a separate winding (on the primary side) by a voltage, which follows the secondary output. And if the voltage won't be sufficient to keep the controller above the undervoltage, it will periodically shut down and then restart after a while.

The failure of the 100W unit you are describing looks really like just the LED failure. I don't think the ballast will fail because of that. It features an overvoltage open load protection (monitoring the auxiliary voltage, but that is tied to the output), so that could be the reason for flashing when few of the strings fail and the rest remain overloaded.


For the IR LED's I would go for another ballast, rated for the voltage range corresponding to the LED's you are using. Or you may look into the datasheet of the ballast controller used in the ballast you have, maybe you are lucky and the under/over-voltages are there set by off-chip resistors, so you may modify them so, the voltage range will shift down.

The drivers are like this :

http://www.ebay.com/itm/Waterproof-Aluminum-LED-Driver-Power-Supply-AC-110-265V-to-DC-20-38V-900mA-30W-/271934426275

It is potted with Black Silicon inside Aluminum rail with Plastic end caps. Big work to tear all the Silicon out, and i dont want to do that unless really have to

So i understand that if i try it and it appear to work fine, it is not going to get damaged in any way from long term working in this mode



The 50W White LED is rated If = 1500mA, the IR LED If = 1750mA. As i still power it on the 1500mA driver, it will be a bit underpowered, so better for the LED

The front of the floodlight is a flat Glass panel at ~30mm from the LED surface. Is there risk of the Glass reflecting IR back to the LED to an extent that may overheat it ?

The IR wavelength would be either 850 or 940 nm

So i understand that if i try it and it appear to work fine, it is not going to get damaged in any way from long term working in this mode

Exactly... The only pitfall may be, it will be marginal, so "just working" when cold and so the LED drop higher, but it may start fl;ashing after it warms up, as the LED voltage goes down. Or it could be the undervoltage threshold may go up with the temperature as well.
But in any way I would not expect any harm at all.

The front of the floodlight is a flat Glass panel at ~30mm from the LED surface. Is there risk of the Glass reflecting IR back to the LED to an extent that may overheat it ?

The IR wavelength would be either 850 or 940 nm

I don't know, where the exact border is. Try to partially cover the LED's with that glass and then with a camera check, whether there is any significant difference in brightness. You may power the LED's by way lower current for this test, so you won't need any heatsink and the camera won't be overloaded by too strong "light" intensity.