New Video. How low can you go?

[danielbriggs]

Eric,
You're video prompted me to have a tinker this evening and measure the actual wavelength shift of a red laser diode. (So, thank you!)
I didn't bother explaining the physics behind it all in the video, just a quick, accurate emperical look-see.

I would be really interested in doing some LN2 tests of my own too; I've just ordered some bits to make a similar insulated test platform, and some low-temp thermocouples.

So not meant to be a direct comparison (and nothing we didn't know already), but:




I'll probably do something similar when the LN2 rig is setup.

Cheers,
Dan

[White-Light]

I knew it...

Anyway, I did not want to repeat the whole thing. But, for the next production I think I will send the sound guys out for ice cream.

Nah. Just buy them some dubstep.

(N)ice video btw.

[macona]

LN2 is surprisingly cheap. I bought 50 liters this last summer and it was about $80 if I remember right. Of course you need a dewar and those are not so cheap.

[planters]

So not meant to be a direct comparison (and nothing we didn't know already), but:

This is not at all competitive, but rather like "collaboration at a distance". What makes this type or investigation interesting is that it is really applied research and the application to projection and display is right down the road.

There are a number of interesting questions that need to be answered in order to decide if this is a practical technique. In my -5C modules which have been running without a hitch in several projectors for several months, there is no instability or focus shift, just a little more power. But, at these far lower temps the effects of differential expansion with the Al base plate, brass mount and glass lens will probably become significant. However, the effects may be additive or compensatory. What about the junction? These are semiconductors and there conductivity may change markedly @ very low temps. Will this be useful with the single mode diodes, the green diodes and the 445nm diodes? Imagine the new 4W 445s run very hard/cold...spooky. Will the pursuit of new wavelengths be an alternative to high current and high power output? These may be somewhat exclusive of each other.

Also, the cold chamber can undoubtedly be improved and a more classic vertical dewar and vertical "cold finger" up to the diode plate might be superior if LN2 is useful.

Keep it coming.

[mixedgas]

If you need more collection of data points:

http://www.dataq.com/data-acquisitio...arter-kits.htm

WinDaq is not bad.

Otherwise, Measurement Computing is your friend.

Steve

[dsli_jon]

..as-usual, you get the vote for Top PL-Phenom...

...Will the pursuit of new wavelengths be an alternative to high current and high power output?

Personally, I'm most-keen to that-end... the 488 DI is just.. *shiver*.. too-bad it's boku-coin, still.. also, Yellow... Wonderful to-see DI-green close to the 'Argon-pallete lines' maturing-rapidly..
All we need is a nice 460-470-ish and something 'tween 550 and 600nm, and we open up a whole host of 'quazi Ion-replacement sources', with fast-modulation, and richer-palette vs just 'RGB'... Really keen to see multi-line solutions 'nearer to earth' than $15K / color OPSLs... and it's not just cost - Not every pj / application needs or wants that much power... ie: 5 Watts of Yellow?? Crikey, 200-500mW of Y on a dome / flat-screen is *way* more than enough... yes, 'there's yellow dpss', but.. modulation-congruity over color, I say, and who want's to deal with color-boards for a 5-6+ line color-engine, to reconcile an odd dpss, if it can be all-done with DI..

Imagine the new 4W 445s...

..Woah, wha'd I miss? ..You talking *spec'd* 4W?? Or the 9mm's, with a lightning-rod up the keister?

j

[planters]

Quote Originally Posted by planters View Post
Imagine the new 4W 445s...
..Woah, wha'd I miss? ..You talking *spec'd* 4W?? Or the 9mm's, with a lightning-rod up the keister?

Yep. spec'd , but not fully released as in the green diodes 6 mo. ago.

Dan,

Nice video. I just love this stuff. I believe that the wavelength shift with current can be explained with temperature. The junction itself is not being measured and stabilized. Rather, it is the temperature of the bulk of the 5.6mm diode and its immediate surroundings. If all the heat that is being radiated out of the diode is being generated at the junction then there must to be some gradient in temperature between the junction and the TEC. This gradient has to increase as the current increases otherwise the heat flux would not change and that's impossible. That is why I believe there is a trade off with power and blue shift. The more power you want to produce the less you will be able to modify the wavelength. Tonight, I will check the voltage drop. The shorter wavelength suggests that the electronic transitions have moved apart and require a larger drop, but the resistance of the leads, solder and the semiconductor should decrease at a lower temperature. Net, Net?

[planters]

Well it just gets better. The voltage drop @ 739mA input does indeed increase from 2.334V @ threshold and RT to 2.498 V @ full cold (-78C). This is an increase of 6.8%. So, the heat deposition due to this fact will increase by this amount.

However, I checked the DELIVERED current at different modulation voltage settings. They were substantially LOWER than I reported. I had CALCULATED by setting the gain pot on the flexmod so that at 2.5V the current was 970mA. All the other currents were calculated from this. Oops! The delivered current is as follows: 1V=417mA, 1.5V=570mA, 2V=739mA, 2.5V=970mA, 3V=1.110mA, 3.5V=1.282mA, 4V=1.1446mA, 4.5V=1.573mA, 4.8V=1.700mA, 5.0=1.742mA.
. 1V=391mW, 1.5V=603mW, 2V=843mW, 2.5V=1.064mW, 3V=1.236mW, 3.5V=1.44mW, 4V=1.618mW, 4.5V=1.798mW, 4.8V=1.898mW, 5.0V=1.929mW

So the Flexmod is far from linear. The output does not roll off. Rather the output is actually accelerating.

[-bart-]

Far from linear?
I bet you didn't take the bias-component into account.
Output = (input x gain) + bias
Deviding output / input does not give you gain. At least not a factor you can use to extrapolate.

Based on your data I think you have a bias current of 86mA and a gain of 331mA per volt.

(Risking sounding as a smartass, not intended)

[danielbriggs]

Looks ok to me...