Delete Delete Delete

[The_Almighty_Bacon_Lord]

Delete Delete Delete

[Marco]

You almost always drive a LED with semi fixed current, either with a constant current driver or through a current limiting resistor. Trying to use the LED's own resistance is not recommended (COBs do it any way of course between strings, exception, not the rule). You have a 3x3 COB LED module I assume since you are trying to feed it 10V? Single chip LEDs (ie. 3V range) more common for this kind of thing, most of the ready made flash light modules seem designed for it.

I think this will work.

The XL6005 can work.

[Marco]

Here's a cheaper one. Interesting circuit by the way, the LM2577 doesn't have a CC mode and since the feedback voltage threshold is too high to directly use with a shunt there's an extra opamp on there to amplify the shunt voltage (and to combine it with the constant voltage mode). If you really wanted to make something yourself this technique can extend the number of boost converters you can use. Also it's a SEPIC converter, but that's less relevant.

Here's someone making a high power boost converter on perf board (constant voltage, but it only differs from constant current because it feedbacks the voltage from a divider on the output rather than a shunt).

[Marco]

constant voltage, but it only differs from constant current because it feedbacks the voltage from a divider on the output rather than a shunt

Look at the datasheet of the XL6005 at the typical application circuit, notice the how there is a resistor below the LED and the voltage across it goes back to FB. That's a shunt, a resistor to measure a current by converting it to a voltage. The XL6005 compares that voltage to a threshold voltage (0.22V) and if it's too low it starts pulsing the internal MOSFET, which raises the output voltage, which raises the current through the diode and shunt, which causes the the XL6005 to use shorter or no pulses.

Now look at the datasheet of a constant voltage boost converter, lets say the XL6009. Instead of the shunt there is now a resistor divider across the output, which leads to FB.

The major difference between XL6005 and XL6009 is the threshold voltage, because the XL6005 is designed to develop that voltage across a shunt which sees a large current it requires a low threshold voltage, 0.22V. The XL6009 uses 1.25V because it gets feedback from the large resistor divider.

Regarding the video, i'm pretty sure that doesn't deliver constant current because the voltage changes, and if it does deliver constant current, how would I make it so it always delivers a max of 1a, no matter the voltage?

It was more to give you a general idea on how to construct a circuit  around XL6005 if you want. These would be the changes you would need to make relative to the circuit he made :

- no need for the compensation network, the XL6005 has it build in.
- no need for the feedback divider, instead you need a shunt as explained above of 0.22 Ohm (0.22V/1A) ... making it adjustable would be slightly awkward, potentiometers at such low resistance get expensive.
- inductor is a little too small, needs to be around 22 uH (and needs to able to stand a DC current of around 1.5A)

PS. the supplies I linked earlier might actually have a problem ... the ebay page warns the constant current limiting stops working below Vin = 7V, I assume it's because of the opamp/comparator they used.

[The Soulman]

Not the most efficient solution but a cheap one, a lm317 and a power resistor.
It could need cooling but google around on lm317 and you will find plenty of examples and circuits.