How to create super bright laser beams

[planters]

Once I have the diodes running,I will be able to provide a lot of data. I'll measure the line width, the centerline for each diode, the beam dimensions and divergence and the change in beam dimensions after the beam passes through the prisms. This last will be the result of a finite line width and a practical rather than a theoretical estimation of what will be the limit on the diode count. I will also measure the losses and determine if polarization rotation will be necessary.

[kecked]

one of the advantages of heating rather than cooling is the modulation affect will be smaller due to the diode already being warm. TEC/resistor heater or what ever will be heating much more than the modulation result I would think so it would fall back down quicker.

[logsquared]

one of the advantages of heating rather than cooling is the modulation affect will be smaller due to the diode already being warm. TEC/resistor heater or what ever will be heating much more than the modulation result I would think so it would fall back down quicker.

The injection current shift seems to be instantaneous. Not sure if it is linear with diode temp? As soon as Planters gets his rig running we should have some answers.

[planters]

one of the advantages of heating rather than cooling is the modulation affect will be smaller due to the diode already being warm. TEC/resistor heater or what ever will be heating much more than the modulation result I would think so it would fall back down quicker.

This is probably untrue. The diodes will gain the same number of degrees C with the same number of J/sec added, no matter what the set temperature. In addition, the life of these diodes will probably be lower when heated, given that they are asked to run at the same distance, temperature wise from 25C. I'm not sure of this, but running at 75C vs -25C is probably not good. An interesting issue is that although heating is certainly easier, the maintenance of a set temperature when a variable source of heating is added, requires that either the diode mount is massive or there is the ability to add or remove heat rapidly. A bipolar heat pump like a TEC with a high enough capacity to reach operating temperature quickly even with a diode mount that has some reasonable mass (not huge) to provide high frequency thermal damping is an attractive approach.

On the other hand, one issue that I will check is the linearity of the wavelength shift, with temperature. This might be significant. To my memory, when deep cooling the red modules in my projectors (I have 5 of these) the initial gains in power at the same drive current are larger. When cooling a couple of P73 diodes to -190 C, I was excited with the initial blue shift, but as the cooling proceeded the rate of the shift slowed. This might mean that the optimal temperature range for a string of diodes will be less cold than I anticipated.

[planters]

The injection current shift seems to be instantaneous. Not sure if it is linear with diode temp? As soon as Planters gets his rig running we should have some answers.

It is. Someone (krazer?) a year back when commenting on one of my videos, linked to a paper that showed the effect is on the nanosecond time scale. This shift is unavoidable, but its significance is proportional to its size relative to the temperature step between diodes. This and the dispersion "smear" of the individual beams means that the best results are going to be with the largest steps between diodes that can be achieved. This in turn means that to accommodate as many diodes as possible, the largest temperature range for the string should be the goal.

[planters]

Well,well... The Oclaro 170mW diodes are NOT the way to go! Firstly, I am not as enamored as I thought I might be by the single mode beam from the Oclaro. It is unquestionably better than the P73 diode and you might get away without spatial filtering, but it is not nearly better by enough to justify 1/5 the power for more money. In any case, I'm very surprised by the line width. At 250mA the line width is 2.5nm. The P 73 is only 1.5nm!

The relationship of line width to diode line spacing works out such that if the line width exceeds the line spacing then the prism pair will actually worsen the beam because the smearing due to dispersion will be larger than your ability to overlap beams. When these values are equal then the net effect of the prisms is zero. If the line width is say 1/4 the line spacing then you would be able to overlap 4 beams, but each beam will smeared to 1.25 its original width.

I only tested a single diode because I am STILL waiting on a delivery of Lasorbs from Pangolin. But, I doubt that this will change much when I test the other diodes. I am not actually disappointed, only surprised. I now expect that I will actually build this module with P73 diodes. At 1.5nm the string will probably need a 5-6nm spacing and it may then be limited to only 4 diodes, but at 1.2W each that's 5W in a single beam and 10W PBSed.

[edison]

At 1.5nm the string will probably need a 5-6nm spacing and it may then be limited to only 4 diodes, but at 1.2W each that's 5W in a single beam and 10W PBSed.

1.2W each? That would need some serious cooling if you ask me. Hard to justify to a customer when selling a module.......
So we are back to combining 2 quads. If the p73 diodes are that great why don,t we see the bigger manufactures like Kvant and Artos using them? The p73,s are too much hassle to get a graphics quallity beam but they are fine for beamshows and that is pretty much it. You see that chinese companies that are in the low-end beamshow market use them but in the highend market the divergence has to be lower then 1mrd and that is very hard to get. But you get an A+ for effort from me because i haven,t seen anybody here that are doing so much testing as you have

[andy_con]

because i haven,t seen anybody here that are doing so much testing as you have

other than me you mean?

[planters]

1.2W each? That would need some serious cooling if you ask me. Hard to justify to a customer when selling a module.......
So we are back to combining 2 quads. If the p73 diodes are that great why don,t we see the bigger manufactures like Kvant and Artos using them? The p73,s are too much hassle to get a graphics quallity beam but they are fine for beamshows and that is pretty much it. You see that chinese companies that are in the low-end beamshow market use them but in the highend market the divergence has to be lower then 1mrd and that is very hard to get. But you get an A+ for effort from me because i haven,t seen anybody here that are doing so much testing as you have

The power I am accustomed to getting is this high because I deep cool all my red modules. This is why the multiplexing is a natural (for me) extension to these modules. I use the downstream optics to focus, expand and filter these beams and when striking a 5mm aperture scanner the divergence from my modules is much less than 1mrad. The down side is the optics require space and adjustment and now, if I add thermal beam stacking, the prisms will be added as well. The up side is a very powerful beam. When I run a module at -25C the efficiency increases by approximately 1% for each degree of temperature reduction, so at the same current that an un-cooled P73 needs to generate 850mW, I get 1.2W. Because the wavelength blue shifts down to approx 630nm the red is visibly brighter as well even though it appears just as red. The reds out of my projectors are warm to the touch and they are very, very bright.

I think the major manufactures and I include you as well are faced with a dilemma here. If you are going to sell a module, you need to ship a turn key solution to producing a given color laser beam...period. The optics that are necessary for systems like mine require even more space than the module and they add the possibility, no the probability, that they will have to be adjusted. For those of us that are comfortable with this and certainly for those that built their own projectors, that is not a serious problem. In that case, you lay out the components on the optical deck (and the heat exchanger below the optical deck ) allowing for the necessary support optics and beam path.

So, when you say "we are back to combining two quads" this is incorrect. We never moved away from the plan to combine multiple diodes of some number, whether it was single mode or multimode doesn't negate the potential advantage of thermal stacking. The problem with the low power of the Oclaro's and the wide line width is that only a few of these will be stack-able and they are, well, low power.

Andy,
You indeed have played an important role here. I have been thinking about this project for a couple of years, but I did not pursue it until your initial experiments and the images you posted demonstrated its application. The fact that you were restricted from showing the combining optic is of little significance. It worked! Although I am doing the lion's share of the research at this point, this too is of little matter. If this technique is to become an accepted method for producing very intense red laser beams then others will have to step up where they have skills such as with producing a multichannel TEC controller.

[logsquared]

Planters; any progress?

I got a picaxe 18m starter kit in the mail today for another project Im working on. These things seem pretty sweet. Thank you to kecked for mentioning them in an earlier post. I might try some temp regulation experiments with it after I learn how to operate it.