It’s easy to create a constant current driver with a FET, opamp and voltage reference. And I’m sure that there are also linear constant current sinks in that range. Or use a constant current buck/boost regulator.
PWM does not limit the current, it just toggles it on/off. In turbo mode when FET is open there is no current limitation at all.
I am not sure such circuit will be more compact unless there is a single component that includes all this functionality.
I am after something like 8*713 in a single package ie. a 3A current limiter than can be PWMd
You want size of 1 resistor instead of 8, but same power ?;)) OK, maybe QX7136 ?
The device does not provide the power, it just controls it. For example FET that controls a lot more power is smaller
The electronics inside is small percent. Most of the size is package and the terminals are also huge wast of space since it is hard to solder 2 very close components.
From engineering point of view it is possible to make a 3A current limiter way smaller than 8 7135s but I am not sure it has a market.
So you want to buy or you want to make ? There is a plenty options already made
I want to buy but so far haven't find any ~3A current limiter. If you know one please share :)
I am out of my depth, but I have read FET can utilize two channels, one linear, one with PWM.
Also, according to maukka’s review here on the Noctigon KR4 only with E21A emitters,
…but I am unclear on whether this constant current/noPWM implementation in the Noctigon KR4 w/E21A is due to this “new type” of FET, or due to the noPWM version of Andúril it uses.
What diameter for “3A current limiter” you need and for how many volt LED?
How easy for an electronic noob ? It would be nice to have an open source led4power LD like driver, with community firmware compatible MCU.
Hardest part would be to design the PCB and the choice of components. The KR4 uses a classic FET current sink controlled via an opamp.
Just because the die is small (it isn’t always small, depends on the power) doesn’t mean the rest is just for the pins. It has to sink all the heat away. That’s why power components have large thermal contact pads or several pins in parallel.
FET and PWM are different things. A FET is like a valve. You can close and open it fully and you will get no or all flow. You can open and close it fast to control the total throughput - this is PWM.
Or you open the valve only partly to restrict the flow. Just look at what comes out and open or close it a little bit more. This is a constant current regulator as used in the KR4. The downside is that the FET heats up.
AMC7135 do almost the same, but in a single package. They have a FET, a voltage reference and a resistor to measure the current. And they heat up because of the same reason.
The valve with water is a nice analogy. Opening and closing the valve very quickly causes the water hammer effect that might destroy your pipes, valves and fittings. Using PWM with a FET is similar, the high power switching causes electromagnetic noise.
Not sure what I wrote to suggest otherwise. As far as I know, FET drivers utilize PWM in nearly every light I have heard of with a FET driver, the Noctigon KR4 w/ E21A being the only exception I know of. PWM is not for its own sake, it is some other gain that is desired. Many don’t mind PWM, but it kind of sucks. My comment was to suggest to OP into looking into FET without using PWM.
Thanks for the explanation and analogy.
The Convoy’s SST40 drivers use a FET as a component in the linear driver, not sure exactly what components they use.
To my understanding, FET as current regulator requires some kind of control loop, so by default a complected circuit with many components.
I am after the most simple and compact solution (does not have to be the most elegant) A component similar to AMC7135 but with higher current rating (single cell voltage)
What about using a solution that the laser community came up with. The LM317 constant current circuit is easily found, the laser community has used it for years.
If you need 3 amps then LM350 should work.
I like the QX7138 but it’s not single component. It requires an external FET and sense resistor. Still much more compact than 8*7135.
I’m not saying the 7136 is bad. It’s been attempted here a bit over the years. I think there’s been some struggle with it’s “soft start” feature as can be read about here: [retired] [WIP] 20mm single sided & 17mm double-sided ?-amp linear driver - surprisingly good!
What I’d really like to see is us come up with a good mid-power buck-boost similar to the Lume1 but at a lower cost and perhaps a bit more simplicity. Pipe dream, perhaps, but a guy can dream…
Yeah, its not direct drive, i limited to 3Amps.