Laser diode array

[Oxymorphone]

Hi, gyus.

Just curious. There are laser diode arrays being sold on ebay, which give away huge amount of power, like 40W here

ebay dot com/40W-High-Power-808nm-CW-Laser-diode-Array_W0QQitemZ130268174579QQcmdZViewItemQQptZBI_Electronic_Components?hash=item130268174579&_trksid=p3286.c0.m14&_trkparms=72%3A1205%7C66%3A2%7C65%3A12%7C39%3A1%7C240%3A1318

for a ridiculous price.

What`s the catch here? I mean can you put to use the hole amount of power? Does it give out a single 40W beam?

 

[mike666]

Couldn't get the link to work. Came up with 'page not shown'

 

[ElektroFreak]

Diode Bars/Arrays are capable of a phenomenal amount of power. They are made up of a stack of individual high-power diodes which work together to provide the necessary output. Depending on how they are used, they can put out a single beam, but the only way I know of to do this is to use fiber-coupling. Fiber-coupling uses a special coupler to absorb all of the output from the diode array and funnel it into a fiber. The fiber output will be a single round beam of 40W in this case.

Most of the time, these arrays are used to pump Nd:YAG rods. This is done by placing one or more diode arrays around the rod, allowing the rod to absorb the pump energy through its sides and output a round beam from its ends (this also requires mirrors and optics).

If you were to purchase one, I'd recommend using fiber-coupling to give you a round beam, which could then be used to pump crystals to give you a DPSS setup capable of several watts of colored visible output.. [smiley=evil.gif]

 

[digital_blue]

You can pipe them down through a fibre optic "cable" but you won't be able to harness all 40 watts. Also, be aware that these things need extreme cooling(watercooling or something) if you are going to run them at full power.

 

[ElektroFreak]

I should add also that the idea of building a multi-watt dpss setup is best left to people who know what they're doing. It's definitely not outside the realm of the hobbyist, but is certainly not a beginner project.

 

[billg519]

If you get one of these bars, make sure to find out if it has a fac lens or not. You'll need one, fast axis divergence is crazy on these bars. Some ebay sellers offer a fac as a $50 extra. To test and experiment with the bar, different sizes of glass rod can be used to focus the fast axis. If you want to focus for burning / cutting, you have to be able to focus both axis (two different divergences) to the same point. This can be challenging as fast axis divergence can be 4 times the slow axis divergence. Get the bar onto a watercooled or TEC heatsink for any serious testing. Powersupply must be able to deliver massive current. I have made a power supply for testing as follows. 120 vac variac powers a microwave oven transformer which has had the high voltage secondary removed and replaced with a low voltage winding made of a few turns of booster cable. This transformer feeds a full wave bridge rectifier rated 150a, a 6v 50,000 uf capacitor with a 5w 470 ohm resistor in parallel smoothes the rectified DC, which goes through a 100a DC ammeter, to the diode bar. Voltage across the bar is also monitored by a DMM. Turn up the variac, watch the current, bar will hit threshhold and lase, turn up power as required, keeping an eye on current and its also a good idea to monitor bar temperature.

 

[Cyparagon]

Diode bars don't give off a heck of a lot of heat. They're around 50% efficient so that's only 40W of heat. Air cooling is entirely in the realm of possibility. Many CPUs generate 100W or more of heat.

 

[ElektroFreak]

Diode bars don't give off a heck of a lot of heat. They're around 50% efficient so that's only 40W of heat. Air cooling is entirely in the realm of possibility. Many CPUs generate 100W or more of heat.

I think you might be surprised at how much heat a 40W bar can generate. Keep in mind that laser diodes are electrically pumped. This process is not totally efficient, with a significant portion of the energy being lost as heat. Laser diodes are also extremely temperature sensitive, and generally like to be kept at room temperature.. The temperature that some CPUs run at all day long would kill a laser diode in a matter of moments.

Laser diodes are the most efficient light-generating devices that exist. Many times their efficiency can exceed 60-70%. I'm not sure about the exact figures for these bars, but let's say hypothetically that in order to get 40W out you need 60W in. That means that 20W is immediately lost as heat and this is combined with the heat directly generated by the lasing process. If the bar isn't mounted on a LARGE heatsink, this amount of heat dissipation will cause it to burn up quite quickly. Air-cooling requires a bulky heatsink, so if you want to pack the maximum amount of power into the least amount of space, liquid-cooling is the answer. In some DPSS setups using multiple bars to pump a rod of Nd:YAG, liquid-cooling is the only practical way to keep the whole assembly as cool as it needs to be to remain stable.

So yes, air-cooling is possible, but for maximum lifetime and stability, liquid-cooling is the only way.

 

[RA_pierce]

[quote author=Cyparagon link=1226955235/0#6 date=1227024192]Diode bars don't give off a heck of a lot of heat. They're around 50% efficient so that's only 40W of heat. Air cooling is entirely in the realm of possibility. Many CPUs generate 100W or more of heat.

I think you might be surprised at how much heat a 40W bar can generate. Keep in mind that laser diodes are electrically pumped. This process is not totally efficient, with a significant portion of the energy being lost as heat. Laser diodes are also extremely temperature sensitive, and generally like to be kept at room temperature.. The temperature that some CPUs run at all day long would kill a laser diode in a matter of moments.

Laser diodes are the most efficient light-generating devices that exist. Many times their efficiency can exceed 60-70%. I'm not sure about the exact figures for these bars, but let's say hypothetically that in order to get 40W out you need 60W in. That means that 20W is immediately lost as heat and this is combined with the heat directly generated by the lasing process. If the bar isn't mounted on a LARGE heatsink, this amount of heat dissipation will cause it to burn up quite quickly. Air-cooling requires a bulky heatsink, so if you want to pack the maximum amount of power into the least amount of space, liquid-cooling is the answer. In some DPSS setups using multiple bars to pump a rod of Nd:YAG, liquid-cooling is the only practical way to keep the whole assembly as cool as it needs to be to remain stable.

So yes, air-cooling is possible, but for maximum lifetime and stability, liquid-cooling is the only way.[/quote]

I'm enjoying this thread.

ElectroFreak:
Do you work with lasers or is it just a hobby?

What of TEC?

 

[ElektroFreak]

I earn a living as an electronics engineer. Unfortunately I don't get to use lasers much at work (don't I wish I did), so that aspect of electronics is just a beloved hobby for me.

My interest in physics in general has led me to do a LOT of research into light and lasers since lasers embody a great many physics principles into their design and development.

I don't have as much experience with gas lasers as I do with solid-state lasers, which I have been studying/building for about 7 years now.


As far as TEC cooling a diode bar: TEC cooling is possible and sometimes used, but usually requires a heatsink attached to the TEC to carry away the heat it absorbs. This puts us right back in the realm of bulky heatsinks, etc.


While the size of the laser doesn't really matter most of the time, especially in a lab environment, it's just more practical to use liquid cooling since it not only reduces the size of the laser head itself, but it does a far better job of cooling overall. Obviously this leads to greater stablility and higher output from a smaller package. This is why nearly every laser module out there that contains very high power diode bars uses liquid cooling.

For some examples of very high power DPSS modules, and information about them, check out CEO laser (part of Northrup-Grumman):
http://www.st.northropgrumman.com/ceolaser/

 

[ElektroFreak]

Also, I should add that everything I have said about the practicality of liquid-cooling applies to commercially available high-power diode systems. As far as designing and building one for hobby purposes, it might be more sensible to use either TEC or active air cooling since it is simpler to design, and will work. Liquid-cooling has been used by hobbyists before, so it is possible to do. Practicality is another matter, however.

For CW systems, using TEC/air cooling might cause the system to require a duty cycle instead of continuous usage. For pulsed systems, air-cooling should be sufficient for continuous pulsing depending on pulse rate and power output. Quasi-CW systems most likely will have the same requirements as CW.

 

[billg519]

I have a 60w bar mounted on a copper plate that in turn sits on 2 x 96w TEC modules that sit on an aluminum heatsink with fan. A thermistor on the diode bar senses temperature and controls the TEC modules. This hobby setup has worked well for me so far. For short term testing, I put a 40w bar on an aluminum heatsink. I first tried a 2 3/4 x 8 x 3/4 finned heatsink, this only allowed full power for a few seconds at a time. Switching to a 3 1/2 x 10 x 2 1/2 heatsink helps, but fan cooling is still needed for longer run times. For a bench setup, my watercooled is the best, with 100% duty no problem.

 

[Cyparagon]

So yes, air-cooling is possible, but for maximum lifetime and stability, liquid-cooling is the only way.

Yes this is arguable, but the same can be said of all high-power laser diodes.

TEC cooling is possible and sometimes used, but usually requires a heatsink

Not just usually, but at all times. A TEC without a heat sink is like a car without a radiator.

TEC cooling ... puts us right back in the realm of bulky heatsinks, etc.

A large heat sink is still likely to be smaller than the entirety of a water cooling system.

using TEC/air cooling might cause the system to require a duty cycle instead of continuous usage.

An average hobbyist does not require anything more.

 

[ElektroFreak]

I use lasers primarily for building stage scanners using an AOM (acousto-optic modulator). If I were to use a 60W diode bar as a pump source in a home-made DPSS system (a wicked awesome 10+watt green one), I'd most likely require at least 5 hours of continuous operation, probably more like 8 in order to incorporate it into a scanner. Unlike standard TTL/analog modulation, an AOM allows me to modulate the beam without ever changing the power to the laser head. The laser remains powered up continuously while the AOM modulates the beam. Using this method prevents shortening of diode life from constant power changes. Where I'm going with this is that anything less than 100% duty would be unacceptable to this average hobbyist.

Also, a liquid-cooled module would be ideal in this case since I could keep the liquid pump and the power supply together outside of the unit, and mount the laser head in the scanner itself. Even more ideal would be to mount the liquid pump, power supply and head together in one location, and fiber couple the beam into the scanner. this way the scanner itself would be very manageable in size, and I could keep the PSU/coolant system in the booth with me.

 

[ElektroFreak]

I should add also, that AOMs are made for modulating gas lasers like Helium-Neon and Ion lasers. My decision to incorporate them into my scanners stems from my opinion that the modulation circuitry included with most Chinese lab-style modules is not clean enough to protect the diode from gradual damage. I believe in over-engineering as well (so long as it is practical), since it helps longevity.