Help driving many laser diodes.

[Helmholtz Watson]

Hi, I am currently planning to build a tabletop cigarette lighter in the style of the National Ignitions Facility. I intend using a largish (~25-35) number of laser diodes affixed round the inside surface of an acrylic sphere focused at a central point. There would be an insertion tube which stopped just short of the focal point.

Just a few questions though.
1) Can multiple laser diodes be driven in parallel off a single driver board?
2) Does shining one laser into the aperture of another laser do any damage? (I intend mounting the lasers in pairs so when there isn't a cig in the way there won't be loads of lasers escaping the sphere)
3) Can anybody think of anything to improve this scheme?

Any and all input is welcome, the design spec. of this project is that it should be cheap, safe and capable of lighting a cigarette.
H

 

[PlasmaCube]

1) Can multiple laser diodes be driven in parallel off a single driver board?
2) Does shining one laser into the aperture of another laser do any damage? (I intend mounting the lasers in pairs so when there isn't a cig in the way there won't be loads of lasers escaping the sphere)
3) Can anybody think of anything to improve this scheme?

Well, first of all, yes, you could theoretically drive multiple lasers using a single driver board. However, you must remember that current is shared in parallel circuits. Therefore, if you wanted to drive, say, 5 red diodes each at 300mA, that would require a driver that was outputting around 1.5A. Therefore, most of the common drivers that are used for hobby lasers (FlexDrive, rkcstr, standard DDL) will not suffice.

For your other question, yes, shining a laser into the aperture of another can and will kill your lasers. Diode lasers have been known to kill themselves just by having something obstructing the aperture (like a dust cap) and the reflected light being enough to destroy the sensitive diode. Perhaps instead of using a sphere, which would warrant mounting the lasers up in pairs, you could use a dome. That way you could focus the lasers slightly below the lowest diode and close to the bottom surface (which could be some dark-colored material). This way, the light would be emitted from above the focal point only and there would be virtually no danger of lasers flying around everywhere.

Also, you mentioned that you wanted this project to be cheap. Well, I'm not sure what diodes you were going to use, but your best bang for the buck as far as burning power goes would most likely be either the red open can LPC-815 or a blu-ray PHR-803T. Both are around $10, give or take. However, if you were going to use 25 diodes, that'd be $250 in itself. Besides, if you're just lighting cigarettes, you really don't need that much. I mean, correct me if I'm wrong, but having just 5 LPC-815s running at, say, 350mA would probably give over 600mW, which would be plenty...

 

[Helmholtz Watson]

Thanks for the reply. I didn't really phrase the original question very well. I have a reasonable amount of electronics knowledge (building robots and synthesizers) but no real laser diode specific knowledge. I was just worried that there was some factor involved I knew nothing about that would prohibit mounting the diodes in parallel.

A dome will absolutely not do
This is the National Ignition Facility we're talking about here!
I'm not trying just to light a ciggie with a laser, I want a mockup of the NIF on my coffee table.

I plan on using many low power diodes focused on the same point. Each individual laser should be around 5-20mw, so each individual laser is pretty safe but leads to a high enough combined power to burn.


In case people are in the dark about the NIF:
Wikipedia- NIF
Pic of outside of laser chamber:

Inside laser chamber:

 

[charliebruce]

The best way would be to hook up diodes in "strings" of maybe 3 or 4 in series, then regulate the current through each chain. If you do it in parallel then as the temperature or forward voltage varies, then the current may increase in other diodes. Worse still, if one goes dead then you may blow all of the other diodes too! The reason for having multiple strings is to keep the forward voltage relatively low. It would make more sense to pair diode up to beam-blocker. One other consideration is getting smoke on the delicate optics.

Have you seen Rog's or Kipkay's lighter? They both use a single PHR-803T violet diode if I recall correctly, however if you want more power then a GGW or SF-AW210 would definitely provide a bit more power, though they can be temperamental.

 

[Helmholtz Watson]

Yeah that's more what I was worried about, that cheapo LDs might vary enough that they would drift relative to each other and cause spikes in other parts of the chain/ladder.

Kipkay's lighter in kinda cool, if stupidly dangerous but it is not the aim of this. Just to restate , I am more than capable of lighting a cigarette using a single high powered laser.
This project is about making a working model of the NIF, not just having a technosmoke.

This means the following aspects of the design are not up for debate:
1) Must use many lasers, at least 10 preferably more like 30
2) Must have lasers mounted around the surface of a sphere pointing to a central focus point
3) Must have a tube leading to central focus point for "sample" (ciggie) insertion

In case people are not familiar with the NIF:

(v. rough explanation) All of those little holes in the walls are where the lasers come in, and they all hit the end of that spike where a tiny amount of hydrogen will be zapped with a picosecond pulse of 500 terawatts.

 

[Flaminpyro]

You should probly try smoking something else as it's been proven cigaretts cause cancer. O wait a min you must allready be smoking something else if you think you need one of these. :crackup:

 

[bobhaha]

hey, just got an idea... tell me if you like it... how about having only the top part of the sphere mounted with the lasers, then have them all pointed at a focal point down near the bottom... like this (sorry for crappy image)

So imagine that, in 360 degrees, where all lasers are pointed down to the centre point at the bottom of the sphere and you can insert the cig from the bottom, so the top of it is in the focal point or you can do like you wanted to do before, and slide it in side ways... hope that gives you a better idea.

Hope that helped! -Adrian

 

[PlasmaCube]

This project is about making a working model of the NIF, not just having a technosmoke.

Heh, well, that's pretty admirable. Although, in that case, you'd need to drastically alter your design to accommodate various aspects of the NIF, such as the Master Oscillator (ytterbium doped optical fiber laser), 48 Preamplifier Modules, as well as numerous other optically pumped amplifiers, as well as a massive capacitor bank to store the energy for the flash lamps...

Being serious now, if you're just going to use low-power diodes, then all I can say is that you're going to have a great time trying to align them all... But anyway, that's an interesting project.

 

[mfo]

Don't connect all of the diodes in parallel to the driver. If one diode kicks the bucket, the rest are going to get surged with that extra juice. If enough diodes should die...

 

[FireMyLaser]

Generally if a diode "dies" it just stops lasing, it still draws the same current and voltage. Still a string of diode in series like posted earlier is the best way. If you want to power them in parallel you will need to hook up a low ohm resistor in series with each diode to compensate for the slightly different voltages that each diode has. Otherwise the current will choose the easiest path (the diode with the lowest voltage drop) and it will pull most of the current and die shortly.

Perhaps combination of the two is the best way...

 

[mfo]

Generally if a diode "dies" it just stops lasing, it still draws the same current and voltage. Still a string of diode in series like posted earlier is the best way. If you want to power them in parallel you will need to hook up a low ohm resistor in series with each diode to compensate for the slightly different voltages that each diode has. Otherwise the current will choose the easiest path (the diode with the lowest voltage drop) and it will pull most of the current and die shortly.

I thought it would eventually break down entirely and stop lasing all together?

 

[FireMyLaser]

Na, it is just an LED pretty much.

 

[mfo]

Na, it is just an LED pretty much.

Well I've seen a burnt PHR eventually fade into oblivion. Plus you yourself said "Generally" in your earlier post. Either way, I say better to play it safe than sorry. Would be a horrible way to lose multiple diodes, wouldn't you say?

 

[FireMyLaser]

Yes but he is using low power red diodes. They don't cost much and are much more consistent than violet diodes.

 

[Helmholtz Watson]

Thanks for the input!
Bobhaha- I think that would make aligning the lasers harder as they would require mounting at odd angles relative to the surface of the sphere, plus I lose an essential design feature of the NIF
Plasmacube- I had thought of doing the whole thing with a single high powered laser feeding 48 fibre optic lines into the surface of the sphere, which would be more faithful but harder to achieve. The allignment is definitely going to be a bit of a pain, but hopefully not too bad, just a case of drilling holes all over the sphere and mounting the modules flush to the surface so they all naturally point to the centre.

I think I'm going to go with 48 5mw laser modules with separate driver boards for each, there's probably too much to go wrong with them wired in parallel. I might do them in groups of 4 to save on parts.

It also occurs to me that the whole thing would be a lot cooler if I could have a big rotary switch that turned on the lasers one at a time, ramping up the power until ignition.

edit: Could a better mathematician than I tell me how to work out how to evenly space 96 points (laser entry and backstop) over the surface of a sphere?

 

[bobhaha]

Thanks for the input!
Bobhaha- I think that would make aligning the lasers harder as they would require mounting at odd angles relative to the surface of the sphere, plus I lose an essential design feature of the NIF
Plasmacube- I had thought of doing the whole thing with a single high powered laser feeding 48 fibre optic lines into the surface of the sphere, which would be more faithful but harder to achieve. The alignment is definitely going to be a bit of a pain, but hopefully not too bad, just a case of drilling holes all over the sphere and mounting the modules flush to the surface so they all naturally point to the centre.

I think I'm going to go with 48 5mw laser modules with separate driver boards for each, there's probably too much to go wrong with them wired in parallel. I might do them in groups of 4 to save on parts.

It also occurs to me that the whole thing would be a lot cooler if I could have a big rotary switch that turned on the lasers one at a time, ramping up the power until ignition.

Yeah I know... But that will, as far as i know / can tell, will give you the best results... BTW this pic

gave me the idea.. maybe you could use something similar to what Jayrob uses for his 7 color build... its a mounting heat sink which is also adjustable so aligning would be far easier.. here's a pic

See you could mount them flush with the sphere then use the screws to adjust the alignment... now i dont know how big this thing is gonna be.. but assuming you are going to have 48 diodes on it... I'm thinking its gonna be quite big, so these will all fit... It also gives the same idea of whats shown on the first image.

Here's a top view of it!

Notice the three screws... adjusting them will align the laser to where you will want them!