How to create super bright laser beams

[kecked]

It is easier to heat and hold then to cool. That is why they heat clocks rather than cool to hold them steady.

[mixedgas]

It is easier to heat and hold then to cool. That is why they heat clocks rather than cool to hold them steady.

Kecked is right.

While it is easier to heat then to cool, TECs acting as heat pumps are sort of the worse source of cooling on the Planet. Would one piece of TEC going to a cold plate achieve enough differential? How much temperature differential do you need on the cold end? Small R134 based fridges and chillers are cheap. In the commercial world we used cold glycol or alcohol for these things. We can also control heat flow and regulate temperature better using flow restriction and solenoid valves.



Also if you mount the diode in vacuum, temperature control gets far easier.
But I digress.

Steve

[logsquared]

Looks like the difficulties with the "gradient bar" are making independent TEC look more attractive. But, when one starts to think about the level of control needed for the independent approach it too becomes difficult. If the individual diode spread is ~8deg and we want =-10% then regulation needs to be .8deg, for 1% .08deg. That might not be easy with diodes under modulation. Especially considering each diode will have a different set point and different TEC arraignment under it. The TEC might need the modulation signal injected into the control loop to achieve good regulation? Hard to say.

I think the bar is a good way to start. IMHO its the simplest experiment to get started. Maybe we will find a more brute force approach is needed? Something like a brass bar with massive cooling ability on one end and resistive heating on the other. Another idea might be thermally insulated diode modules where a thin metal strip or wire is used to provide the gradient?

I am really looking forward to the spectrometer readings!

[planters]

log,
I think the way I am using the bar is close to optimal and only a more robust cooling method would significantly improve it.
I think that tempreture steps of closer to 15 C will be necessary based on a need to distinguish the line width which is probably not going to be better than 0.5nm, given a 0.22nm/C relationship.
Although they are feeble, TEC's are solid state and with efficient diode lasers, I believe closed loops with these will prove the best solution. We really need a multichannel, multi-amp controller. With this missing link I think this project will take off. I need to make a PM.

[andy_con]

0.5nm difference between diodes isn't going to work sadly.

[planters]

0.5nm between diodes would only work if the line width was no more than 0.1nm. I have no hope that will be the case, but I will try to measure the line width tonight.
That is why I think we will need a minimum of 3 -4nm between diodes and that will a require 15 deg step.

[andy_con]

I would agree with you there, the bigger the difference the better. Shame I only managed to get 8nm difference from 0-40 degrees

[planters]

Agreed. But, if we can get away with less than 20 degrees per diode and we use direct, multistage TEC's then 4 gaps and 5 diodes are really possible. That is a substantial jump in intensity.

[logsquared]

Shame I only managed to get 8nm difference from 0-40 degrees

That seems to be roughly in-line with the spec sheet. 8/40=.2nm/deg

[andy_con]

Due to my lack of spectrometer it's difficult for me to do testing, but I plan to do some more testing at the next uklem

I have a test rig with two diodes, both tec'ed on it and the optic