Testing of materials for laser diode housings

[Bbe]

This method is good for determining the temp shift. You need to have a super-fast spectrometer to catch the emission frequency at the start, when the chip temperature is known.

Actually, you can make a pretty good estimate of chip temperature from the wavelength shift. At a fixed current, the GaN semiconductor red shifts approximately 0.05nm/C and the AlGaAs red shifts 0.25nm/C

In reality, we will have higher temperatures due to thermal barriers (between parts) and thermal losses on optics. For o-like lenses we can add about 30% of thermal power to the model.

Yet, I was suggesting that the simulation could be compared to real world measurements of the mount with a thermistor attached to the surface. When I operated one of Dave's mounts with approximately the same heat load as your over driven model the top of the brass mount was hot. I had not bothered to measure it at the time, but it was well above body temperature (37.0 C). One of the confounding elements was this laser was using an O-like lens and this has since been replaced by a lens that is provided with the harvested diode and preforms better with a higher power throughput. It is possible that some of the heating should be attributed to the O-like lens.

I meant that the diode is made of silver

You state "laser diode-silver package" What do you mean?

bart,
Heat pipes do not work over short distances due to the fact that not enough conditions for the circulation of the refrigerant inside.

No, I mean a diodemount with heatpipes.
Can you simulate that to investigate if this would be a feasible idea ?

[planters]

You do not need a super fast spectrometer. There is a small red shift to the emission frequency that is not temperature dependent, but rather current dependent. Here, it is necessary to have a nanosecond scale spectrum to separate out this component. But, with steady state operation @ a constant current the wavelength shifts that I gave above are pretty well established in the literature and I have seen these hold for the diodes I have personally tested. The red shift I measured for these diodes was 2nm and this represents a junction temperature rise of approximately 40C I believe this to be reasonable as I calculate below.

Regarding the O-like lenses, I gained about 20% optical power by going to the "stock" lens and this lens itself may have a 10% loss. That is speculation and as far as I know untested, but your estimate of a total absorption for the O-like is plausible. However, this is 30% of the optical output or in other words, 30% less of the total input power will be removed in optical emission.

Looking at your input numbers again I noticed that my operating current and measured V drop were substantially higher. I was running the diode @ 2.5 A and the V drop for these 9mm diodes is higher than the A and M mode diodes. The V drop I measure is 5.0V. Others have measured as high as 5.5V at these higher currents. The optical output is 2.9W and so the heat load that would be applied to the model is 2.5A x 5.0V -2.9 W = 9.6 W. This is a higher thermal load and so I would suspect the gradients for each material would increase proportionally.

In reality, we will have higher temperatures due to thermal barriers (between parts)

I agree. I also agree that heat pipes will probably not be advantageous. Even though they can be made quite small they will have to be assembled and incorporated into the bulk of the block and this will introduce interface losses and the number of holes in the block for the lens, the retention plate and the underside mounting screws would greatly constrain the placement of the heat pipe(s). An alternative (I am not recommending this) is water cooling. I water cool almost everything and the possibility of reducing the distance heat must flow within the block to reach a small channel or even a micro-channel with flowing water greatly improves heat removal. This is why kW class diode stacks routinely use micro-channel cooling within the bulk of their copper mounting. But, the low cost and simplicity of simply shifting a proven design from brass to copper and the substantial improvement in performance is very attractive and this is why I originally discussed this with Dave.

[Bbe]

I agree with you about copper - if the price is comparable with brass, then it will make sense.

This is why kW class diode stacks routinely use micro-channel cooling within the bulk of their copper mounting. But, the low cost and simplicity of simply shifting a proven design from brass to copper and the substantial improvement in performance is very attractive and this is why I originally discussed this with Dave.

I used to think that it is extremely difficult to measure at a "kitchen" laboratory

There is a small red shift to the emission frequency that is not temperature dependent, but rather current dependent. Here, it is necessary to have a nanosecond scale spectrum to separate out this component. But, with steady state operation @ a constant current the wavelength shifts that I gave above are pretty well established in the literature and I have seen these hold for the diodes I have personally tested.

Meanwhile, I checked virtual setup with the new input data:

3-Diode in mode.

Current - 2500mA,
Forward voltage - 5V
Total power - 12500mW
Laser emission power - 2900mW
Heat emission power - 9600mW

3a - Dave's brass housing:


Chip temp - 48.2°С
Diode package temp - 34.3°С

3b -Dave's copper housing:


Chip temp - 42.1°С
Diode package temp - 29.9°С

And a little more for my old housing:



3c - bbe's brass housing:


Chip temp - 46°С
Diode package temp - 32°С

3d - bbe's copper housing:


Chip temp - 40.2°С
Diode package temp - 28.2°С