Literally continuous duty driver help please

[AUTO XX]

Hey guys, I have an idea for a signaling device that would need to be powered 24/7 at at least 500mA.
1.5A might be on the table later on for developing other devices but this is more about longevity.
This driver would be running a laser, either green dpss or 445.
I am leaning slightly toward 445 just because they are rated for continuous duty as well and this device would need to run for at least a year with no maintenance or adjustment.
Is it possible?
I figure at 500mA with a well regulated driver, the 445 diodes should last at least a year. Are there any decent quality green that could boast that, even at 250mW?
The green is preferable because of increased visibility but a higher driven blue would show up as "bright".
Heatsinking is not as problem as this is essentially a lab build and could even have a tec system installed.
ANY input is welcome
:thanks:

-Auto

*edit* would the Flexmodp3 work for something like this?

 

[kiyoukan]

well a simple lm317 will work at that power and you can add a transistor to the adj pin to turn it on and off.
all you would need is a 5v signal. to make that you can use a 555 chip that would blink every x and stay on for x seconds.
What type of signal do you need?

 

[Benm]

I dont see a problem with it either. I'm assuming your powering it from mains, so any linear driver would be fine, as long as you provide proper heatsinking.

I'm not sure the diode will actually last a year, but the driver will not be a problem, these things can operate 24/7 for decades without failing.

 

[kiyoukan]

lets say these lasers at 500ma tec cooled get about 5,000 hours
thats 208 days.
So maybe not a whole year but close. and if its going to be on for one second and off for 2 that will increase the life alot.
At 500ma you are driving it the same as the projector.
With good heatsinking it should last longer that if it was in the projector.
There is less heat for it.

 

[Benm]

Sounds about right. We don't really know what happens after 5000 hours though. It could simply degrade and get dimmer to a point where the projector no longer meets its specifications. It could also fail completely at some point.

Also, we arent sure the blue lasers life is the limiting factor on the lifespan of the projector - it could also be the red led or phosphor that degrades over time.

 

[kiyoukan]

very true.
I just imagine the lasers going before the led but the wheel is an interesting part.
That much laser power on it there no doubt that it could fry given the right amount of time.

 

[AUTO XX]

Very valid points guys
@ kiyoukan This would have to be continuous "ON" for the laser.
Size really isn't a concern but I forgot to add that it would be powered by a solar powered, charge controlled system with a 12V battery.
The flexmodp3 takes on up to 24V but I don't know how it would stand up to the infinite duty cycle.
I'd likely go with a passive cooling system just because machining fins into the aluminum housing aluminum wouldn't even be a problem.
Efficiency would help as it decreases the size of the solar panel needed and so the per unit cost.
Good point about the laser not likely being the component that fails first, I hadn't thought about that. Its usually the moving components that fail.

 

[Bionic-Badger]

You should find out how efficient the Flexmod is compared to the Flexdrive or other drivers if you're going to be powering this from solar panels, especially in the winter when the light levels are far lower. What they don't waste doesn't get turned into heat as well.

How big is your solar panel array going to be if it can drive a TEC + laser continuously especially in lower-light seasons?

 

[AUTO XX]

I can't get Lava to answer emails about a faulty flexdrive let alone specifics on the flexmod....
The size of the array and battery is pretty easy to figure out, this unit might even be seasonal depending on the usefulness so winter months it would be shut off.
I have decided TEC would only be a last resort, with a few fins cut into the heatsink it will be easy enough to get a large enough surface area to keep the entire unit cool.

 

[HIMNL9]

Remember that all the energy that you don't use, is always dissipated in heat, from linear drivers.

Use a sink or source driver (like the flexmod, or any DYI one), and for green modules use a 6V battery pack, or a Li-Ion protected pack, instead a 12V one, so the dissipated energy is much less, you can use a bigger battery, and have a better efficence and longer lasting (and the solar panels can still be used for reload)

 

[Aaron]

If you plan on running it non-stop for a year, my suggestion would be to press the diode in a custom copper module (available from Jayrob). Then, break apart a peltier plate and use the thermocouples from inside it, glued to the module with thermal glue. Then, the hot sides of the thermocouples are directed outward towards a massive labby heatsink.

 

[AUTO XX]

Well, it isn't a handheld.
The heatsink can be literally ANY size. I figure with enough surface area the heat will take care of itself, especially if it is a 405nm only running ~300mA. I revised what I needed and a few fluorescing surfaces strategically placed will have the same effect.
Oddly enough I just got 5 TEC units a few weeks ago (the price was so good I couldn't pass 'em up ^_^ )
I'll experiment with them but this will be a solar powered device so any extra power consumption = a bigger panel = more $ = harder to sell
I decided against 532nm because the dpss lasers are SO heat sensitive.
http://laserpointerforums.com/f42/effect-temperature-dpss-laser-image-heavy-58525.html
A bit more experimenting is in order but I should have a working model after a couple of days in the shop provided the weather improves!

Thanks for the input Aaron
*how would one disassemble a TEC (peltier) unit?

 

[HIMNL9]

how would one disassemble a TEC (peltier) unit?

VERY difficult, keeping it working ..... the junctions cubes are also polarized, but nothing on them indicate the polarity, so you need to take them one by one, and made a custom soldering surface too ..... and they are not insulated, so you just can't stack them on a metal plate .....

 

[AUTO XX]

Sounds like making a heatsink with a machined flat surface would be a wee bit easier.
Thanks for the info HIMNL9, your ability to explain this kind of thing (even with a subtle language barrier) is great, you must know this stuff inside out ^_^

 

[Benm]

I really would not bother with TEC in this case, and certainly not with thermocouples harvested from a TEC module. I think these have a larger chance of failing then any other component in such a setup.

Since its not portable and space isn't much of a consideration, just use a bigass passive heatsink. If i had to construct it, i'd mount the laser in something like this: Alloy Module Holder/Heatsink (12mm) - Detailed item view - OdicForce Lasers Online Shop and then bolt that down onto a sizeable standard heatsink with thermal compound in all junctions.

As for driving it from 12 volts: This can be a bit finicky, the driver will dissipate about as much heat as the diode. You can still do a linear driver, but it will need heatsinking too (not as much as the laser as it can run hotter), and it will lose half of the energy. I don't know if energy efficiency is a concern, but if it is, a switching driver could be the better choice.