New Video. How low can you go?

[danielbriggs]

Just an update:
I've mounted an LN2 'hopper' to contain the liquid and draw heat from the diode running at the end. A T-type thermocouple is bonded inside to monitor the temperature close to the diode. I was planning on getting some cryogenic compatible thermal paste to reduce thermal resistances at the interfaces, however this was $160 a tube, and I'm not fussed on spending that much on a first attempt. An unheated AR window is added to the box.

I've not included any space for correction optics on the cryo-sled, these can be added outside the enclosure. The aim of this is not to tame the beam as well; just to investigate how the peak emission wavelength and power output vary with temperature for the P73 diode installed.

The cryo-sled is fixed on a 1.5" layer of high density polyurethane foam at the base, and surrounded by 3" of the same foam around the sides.

I still need to source a T-type thermocouple interface to LabView and a lens for the diode... then knock up a LabView VI panel to control and monitor the diode controller, spectrometer, temperature, beam profile and laser power.

Only then will I throw some LN2 into the dish

Until next time...











Ooops... a slight miscalculation on the expanded volume.


Much better...!

[planters]

Yes! I love this. I like the lay out and the move to an aluminum diode mount. I think you might want to heat the window. When I went to LN2 and -173 the power began to fluctuate and an orange (yes orange) glow began to surround the central spot. This was corrected for a few seconds with a quick wipe across the window and a haze could be seen accumulating. Or am I not seeing this correctly? Are you going to use a double window?


I am going ahead with a functional build and the optics beyond the collimator will be @ RT. There is a slight and reversible defocus at the lowest temps and this can be corrected with the spatial filter lens when the temperature stabilizes.

You need to make sure the diode remains sealed even when you access the LN2 or the parts are cold.

[danielbriggs]

Will do Eric!

It is a single window; I can always heat it should the need arise.

I sanity checked the thermal response of the system after the foam finished curing this evening; (just a pre-test, no thermocouple readings yet until the USB interface arrives).
I poured boiling water into the hopper and despite not using the pricey Apeizon N grease I was looking at originally, the heat transfer was much quicker than anticipated. The thermal bulk of the system should be able to handle a little diode; but we'll see how it copes with a distant, transient 5W+ thermal load in time.
Should it not be up to the required q-dot, a more involved cylindrical copper design could be made.

Is the change in focus you mentioned due to the change in wavelength? Or thermo-mechanical considerations ?

[planters]

Is the change in focus you mentioned due to the change in wavelength? Or thermo-mechanical considerations ?

I hate to admit it, but I missed that. I just assumed thermo-mechanical, but thinking about it a bit, because a shorter wavelength will require a shorter distance to the diode and lower temperature should cause this then one must dominate otherwise no defocus would occur. This could be checked by inserting a pair of identical, positive, correcting lenses in a one power telescope outside the chamber and determining which is further from the focus.

Based on an estimate of your dimensions and 5W load, I suspect most of your thermal gradient will be in the mount and amount to 3-4 degrees. I don't know how to estimate the interface gradient at the diode and at the diode mount base. If you want to reduce this I would think about a copper diode mount only and indium film @ the base and call it good. Of course, you can always put a low vacuum on the LN2 and gain 16 more degrees.

[planters]

My spectrograph arrived and I am going to set it up to get some readings of the cold diode @ various input currents. I know you want to do the whole thing right as in a single display with the current, power and temperature all neatly visible. But, I am very curious as to how close you can get that thermistor to the -196 goal with your set up. I am building a cryo pbs dual for a short fiber feed into one of my projectors and I am going to jump ahead and fabricate the mount/cold finger. Can I get away with aluminum? Where does the current effect on wavelength (independent of bulk temperature) fit in. In other words "how low can you go"? Can we hit 600nm?

[mixedgas]

Eric, I should mention it helps tremendously to use vacuum, at 1 Torr or less, as your insulation. A simple double or single stage rotary vane pump will get you there.

Steve

[planters]

Yes I know. However, for a lab set up this is fine, but if the cold diodes will be used for display (the ones I am working on) then this becomes significantly more complicated. LN2 is easier than many people might think.

I was a little disappointed by my first test with the spectrograph. At -180 C the output wavelength became much broader and spanned 608-611nm. Also the wavelength shift with current became very small as in 1nm from 300mA to 1,500mA. My cold chamber is admittedly a little funky and with better isolation the diode itself might shift a little further. We'll see.

[FutureDesign]

I was a little disappointed by my first test with the spectrograph. At -180 C the output wavelength became much broader and spanned 608-611nm. Also the wavelength shift with current became very small as in 1nm from 300mA to 1,500mA. My cold chamber is admittedly a little funky and with better isolation the diode itself might shift a little further. We'll see.

That is some really really cool experiments you are doing! Thanks for sharing them with us! I would love to see the -180 C experiment on video. Any chance? Did you try to push the diode harder with LN2 than you did with the dry ice? If yes, did the linear part of the current vs power curve extend?

I was thinking that if the source was fiber coupled then the hole setup would be simpler. The diode could more or less be submersed into LN2 and just the fiber would leave the tank. Hopefully there should not be any condensation within the diode, as it more or less would be same temperature. With such a setup the isolation could be near perfect and the LN2 would last much longer.
The fiber leaving the tank would not be a victim of condensation and the fiber end would be in room temperature. Of cause there would be risks of breaking the fiber.

/Jan

[planters]

FutureDesign,

My current build is a lot more sophisticated, as the first was just a proof of concept. I did push it harder with LN2. Look @ post #39. I am now using an all aluminum set up with substantially shorter and thicker thermal paths and my estimate is 2-3 C temperature drop from the diode flange to the LN2. The entire assembly will be held in a cryostat. Two diodes will then be externally PBS-ed,spatially filtered and fiber fed to a projector. The problem with submersion is that without pumped and flowing LN2, the concentrated heat source will likely lead to bubbles. These can be erosive as well an obvious optical nightmare (unless you are looking at a lumia right out of the box).

I will do a video of this work when I am a little further along. I also am waiting to see some results from Danielbriggs

[andyf97]

This thread is exactly what PhotonLexicon should be about, experimenting, learning and developing ideas in the hobby style. Not the none stop spam full it has become lately with manufacturers/developers using PL as a free market place.

Planters and FutureDesign; you guys are bringing more to PL than I think anyone has for many years so a happy Easter and a big applause to you both.

I say down with commercial adverts on PL.