Concept Build need some expert opinions!!!

[dieselmarine]

This is just something I was thinking about trying its basically like a lightsaber I am looking to achieve a thick beam maybe with phr's first to see how it goes. So I want to make an aluminum heat sink just like an aixiz module but maybe around 6x the size. Threaded, but it will have 3-4 diodes press directly into the heat sink depending on which is easier to focus 3 or 4 diodes? I already started on the lense I machined the aluminum lense housing and I'm currenty experimenting with different types of glass lenses like: binocular lenses, magnifying lenses, Also I have to figure out what drivers and power supply to use. I would like to know if this is a waste of my time anyone feel free to reply but I am looking for replys from :

Daguin
Jayrob
larryDFW
Lasersbee
Iskor12
Tech_junkie
Kenom

And if anyone feels left out thats too bad stop being a baby, lol just kidden
go right a head!

PEACE,
-SARGE

 

[brtaman]

Maybe my opinion wont matter...:cryyy: :cryyy: :cryyy:


But the idea as good as it sounds won't work...I had the exact same idea in my first months on the forum, but you will not be able to focus, well in theory you can get pretty close, but that is after quite a bit of research and tons of $$$...really much more trouble than its worth.

Honestly your best bet would be to combine the beams of 2 12X diodes with a PBS cube collimate it and use a beam expander. Would get you the nice thick lightsaber beam you are looking for, with ~1W of 405nm should be damn bright as well...

As far as driving the LD's...really depends on your space constraint and at what current you want to drive the LD's, normally 2xFlexdrives would be perfect, but maybe you want to drive the LD above the flexdrives limit? Then a simple linear driver would be good for the job...

 

[dieselmarine]

Your opinion is valued thanks!

 

[brtaman]

No problem man, happy to help :tinfoil:, honestly I am really interested in how this project will develop.

It is feasible with the PBS cube, will take quite a bit of aligning and some machining and good engineering but will definitely one of the cooler projects attempted on the forum.

 

[charliebruce]

I don't think that this method of joining the beams is able to give a decent-shaped beam without being very complex indeed - the way it would most likely be professionally done is with fiber-coupling then beam-expanding, but you could use a clever combination of mirrors and cubes to join them as well. If you're not worried about efficiency, a set of LiPo batteries wired to give around 20V with an LM317 driver, then putting the diodes in series, would be the cheapest way of driving them with constant current.

In all honesty, I doubt the beam is going to be visible if expanded to more than about a centimeter or two.

 

[Deleted member 8382]

I can't understand the idea. So you are trying to put 4 beams each close to the other in order to get a big beam? Or you're trying to combine 4 beams and then expanding them in order to get the same effect?

Regardless of this, you should consider red diodes rather than 6x. They're more visible, cheaper and are easier to focus. To power all this I would consider to use a single LM317 based driver in series with all the diodes or 4 of them in parallel. All this supplied by 4 18650 batteries

Hope this helps.

Yours,
Albert

 

[Deleted member 8382]

@charliebruce: I see you posted at the same time than me with the same supplying idea so let's ask you the only doubt I had about it. Is a single LM317 going to hold that amount of power? IIRC it can, but I'm not 100% sure

By the way, I just got an idea. What about using a pyramidal mirror?

 

[charliebruce]

To power all this I would consider to use a single LM317 based driver in series with all the diodes or 4 of them in parallel. All this supplied by 4 18650 batteries

To be honest this isn't a good idea, because unless your diodes are identical in every way, the current through each diode won't be equal (one with lowest forward voltage will draw more current, shortening its life), and if one dies and no longer conducts, the rest will soon follow it.

IIRC the LM317 can take up to 1.5A at 37V when properly heatsinked, so yes it should work fine in this situation, provided the input voltage isn't very far away from (total number of diodes * average diode Vf).

 

[Deleted member 8382]

Well, if it can really hold 1.5A it might not be enough for LOCs but yes for GGWs in parallel. About the current differences, if placed in series, wouldn't the current be the same but the voltage drop different in case the efficiency vary from diode to diode? We're talking about using the LM317 as a constant current driver and not as a constant voltage driver right?

About the pyramid idea, I mean something like this:

 

[Laser_Ben]

Well, if it can really hold 1.5A it might not be enough for LOCs but yes for GGWs in parallel. About the current differences, if placed in series, wouldn't the current be the same but the voltage drop different in case the efficiency vary from diode to diode? We're talking about using the LM317 as a constant current driver and not as a constant voltage driver right?

About the pyramid idea, I mean something like this:

You mean Pyramid?

 

[charliebruce]

Diodes in series - current is constant for each diode, but forward voltage of the "string" changes, and so the driver also has to adjust this therefore input voltage must be higher than sum of forward voltages + 2.25V.

Diode in parallel - forward voltage of each diode is equal, but the forward voltage at a current for each diode varies, so the current per-diode varies - it isn't (say) a quarter of the driver's total output. This way, if one diode blows, the voltage will increase, killing the other diodes as well.

Wherever possible (when you can get the high voltage needed) it is much better to use diodes in series, because the current will be constant, and one diode dying doesn't kill all the others.

LM317 could take a 20v input, regulate the current to 100mA, and run this through a string of 3 PHRs (about 16.5V total between them), meaning the energy the chip dissipates is (0.1A * (20-16.5V)=350mW) which should be fine if the chip is heat-sinked.

 

[Deleted member 8382]

@laser_ben: ?
@charliebruce: then why did you say:

To be honest this isn't a good idea, because unless your diodes are identical in every way, the current through each diode won't be equal

xD

 

[charliebruce]

@charliebruce: then why did you say:

xD

I was referring to the 4-in-parallel bit of your quote, sorry, should have been clearer.

 

[Deleted member 8382]

oh, when I said 4 of them I meant 4 drivers with 4 diodes all in parallel to the batteries, I should have been clearer sorry

yeah of course 4 diodes in parallel is a bad idea hehe

 

[FireMyLaser]

Simply put: this will not work.

 

[maxh]

It looks like you're trying to use one lens to collimate multiple laser diodes clustered together. This won't work. The light coming from one diode is fanning out and in order to collimate it, you must position the diode so that the point from which the rays emanate (the tip of the V) is at the focal point of the lens. Of course you can only fit 1 diode there at a time, so the other ones in your cluster couldn't be collimated into a beam.