DIY lm317 driver for 445nm 1W diode?

[rhd]

Was planning to base off of the basic DDL circuit:
http://www.rog8811.com/LM317 components01.jpg?0.35312663627278806

 

[Blord]

The DDL driver is perfect capable for 445nm. And it is cheaper to build the driver.
The microboost is compact and ready to use.

 

[3zuli]

pros:
- it's DIY
- cheaper
- can go up to 1.5A (microboost only 1A)
- easier heatsinking

cons:
- large voltage dropout (2V) -> hard to use in portable host
- it's linear regulator, wastes more heat and can't boost up the voltage
- you must know how to build it

 

[rhd]

So realistically, because of the voltage dropout the DIY LM317 is only going to work in a portable host that has 2x CR123, not 1x 18650 ?

 

[Kevlar]

Are you building a pointer or labby?

The LM317 are perfect for labbies. (actually, in a labbie I'd use the LM350 because it's rated up to 3A) They are good for pointers too but I think the biggiest drawback is that it can be difficult to find a host to fit the completed driver.

The LM317 has a reference voltage of 1.25V but once you add a resistor, diode, and cap, you are looking at about 2.5V--3.0V as reference. So your host would need to be able to hold enough batteries to supply enough voltage for the diode and the reference voltage which is about 8V. A host that holds 2 Li Ions would most likely be OK. 2 Li Ions charged up is about 8.4V. That doesn't give much run time once you put a load on them. Just my honest opinion. :yh:

 

[rhd]

Oh interesting, ok.

I have a microboost on order for my first pointer, but I've never seen them in the flesh. Is it possible to DIY a driver small enough to fit in a typical 1x18650 host and run off 1 18650?

 

[jib77]

Oh interesting, ok.

I have a microboost on order for my first pointer, but I've never seen them in the flesh. Is it possible to DIY a driver small enough to fit in a typical 1x18650 host and run off 1 18650?

This requires a complex switching driver because you will have to boost the voltage. Switchers use surface mount packages that require advanced tools to solder and a DIY PCB. Layout of the PCB is also important ... see >>this<< thread to get an idea of what a pain in the ass it is.

 

[rhd]

Switchers use surface mount packages that require advanced tools to solder and a DIY PCB.

That stopped my idea dead

I'm comfortable with PCB layout software, and even used to acid etch copper boards using the laser printer DIY approach. But I've never once successfully mounted an SMD component.

Has anyone confirmed an LED-oriented driver board to work for 445nm 1W lasers yet?

 

[rhd]

It looks like 3x AMC7135 has worked in a labby - any reason not to use this in a properly heatsinked pointer?

 

[Blord]

You don't need a pcb to solder the components. Just like this build with smaller LM317LZ, but you need a bigger host for the LM317.

http://laserpointerforums.com/f50/my-635-red-pen-custom-ddl-driver-lots-o-pics-57848.html

 

[BShanahan14rulz]

It looks like 3x AMC7135 has worked in a labby - any reason not to use this in a properly heatsinked pointer?

Because you would need to run 2 Li-Ion cells in series, and that can be dangerous if you don't know what you are doing.

 

[Phantom101]

So you can use a DIY LM317 driver for a 1 watt 445nm laser diode, cause I just got one from ebay and I need a driver. Im new to this:yabbem:

 

[rhd]

Yes - but you need at least two, potentially three li-ion batteries to power an LM317 build with a 445nm diode.

 

[Phantom101]

For the driver if you want it to give off 1w of power you need to switch the two 10ohm resistors out for a 130ohm resistor or a 100ohm and a 30ohm resistor in sequence.

 

[rhd]

For the driver if you want it to give off 1w of power you need to switch the two 10ohm resistors out for a 130ohm resistor or a 100ohm and a 30ohm resistor in sequence.

Nope - If you're looking for 1W, then you're actually going to want about 1.3 ohms, not 130 ohms. If you used 130 ohms, you'd end up with somewhere around 10mA current out.

 

[jbord39]

Be sure to use a power resistor. A regular resistor (smaller than a penny) will burn out. The power needs to be rated for V^2/R = 1.25^2/1.3 ~ 1 W. You could also stick a few non-power resistors in parallel.