Ice is Nice....

[CDBEAM]

Well….just a little TE Test Bed project I have been working on. Let’s call this the "Ice is Nice"….or… Poor Man’s TE Cooler !! See the pics !!
I am using a 40mm x 40mm 48W TE coupled to a copper radiator and two standard 4”computer fans to move the heat away from the TE. (I have a 40mm x 40mm 192W TE plate that I will try next)

The Test Bed shows an Aubers Controller on it….I may use this….but for now…..with the limited delta T the 48W TE can do….the TE is just ON….nothing else.
The LD is a G71. Optics are 6X cylindericals. LD Mount…my custom design…and very adjustable and very easy to adjust !! ( LDAM V7b )….but…I must note that the original, original inspiration for this design was posted by Logsquared…..somewhere….in the Grail Thread !!

The goal is to start to document the effects a slight amount of cooling has to both stabilize the G71/P73 and to possibly increase the slope efficiency. No great body of data yet !! The LD stabilizes at about 800mW @ 18C @ 800mA…just works out to be about 1mA/1mW… @ 18C. This test was run over 3hr....no modulation.
In comparison….when running for over two hours...NO TE…temp stabilizes at about 25C….with an output of 620mW….still with a 800mA current in. So…even in this somewhat simplistic set up…ICE ( Cooling) is nice....and I think all modules using the G71 or P73 LD will benifit from a TE application.

I will experiment with the larger TE unit next….and a P73 LD. The end goal is to establish a prototype unit for the PL Spec Projector project. For now…the G71/P73 still rules the skies….IMHO….and I would predict….that with this set up….maybe with the larger TE….I can get the G71 LD closer to 0c….and see what the efficiency will be ??? Then I will experiment with a P73…

Anyway….no Dry Ice, No Liquid Nitrogen….NOT that these type experiment are not VERY valuable….just that I will not mess with the “ Fickle Cold Finger” haha ….Just wanted to show some simple experiments… with the most simple basic set up…. a TE/Radiator/Fan sandwich !!!.....Things that MAY make it into the PL Spec projector ??

CDBeam

[polishedball]

Nice looking CDBEAM... I have been very happy with the ovenind controller I just tested out in my latest build. Kinda pricey at $55 but not bad.

[planters]

CD,
From the pictures I can't tell how the TEC is positioned vs the diode holder. If the entire Al plate is being cooled then watch out for condensation once the temp drops below 10C. I have found the weak link in this simple sandwich approach (it currently runs in all three of my projectors) is the hot side/fan. As the TEC differential is increased with current most of this is negated by the increase in the hot side temperature. So, I would go overkill on the radiator/fan area and flow. I have found the improvement in power output performance for both of these diodes @ any given current is pretty close to a linear 1%/degree C down to at least -30C. Good luck.

[polishedball]

Do you bring the hot side right to a heat sink and fan?

CD,
From the pictures I can't tell how the TEC is positioned vs the diode holder. If the entire Al plate is being cooled then watch out for condensation once the temp drops below 10C. I have found the weak link in this simple sandwich approach (it currently runs in all three of my projectors) is the hot side/fan. As the TEC differential is increased with current most of this is negated by the increase in the hot side temperature. So, I would go overkill on the radiator/fan area and flow. I have found the improvement in power output performance for both of these diodes @ any given current is pretty close to a linear 1%/degree C down to at least -30C. Good luck.

[CDBEAM]

Thanx PB/PL !! Yes...The stackage is: LD in mount/Ally plate/TE Unit/ Large CU radiator/Dual fans....and Yes...I agree !!! Overkill on the fans/radiator !!.The goal here is to MOVE that heat away from the hot side....the better this is done...the greater the achieved Delta T and the more efficient cooling is delivered.

I suppose an arrangment could be designed where the TE unit is in direct contact with the mount ??? As in to inlay the TE within a machined pocket in the Ally base plate....but then....there is a thin cross section of Aluminum between the TE and the CU radiator....mmmm....which " Master to serve " ????? OR...One could "float" the TE on a ID ledge within the base plate....and have the radiator a smaller footprint than the ID of the ledge ???...just some thoughts....and there is always the issue of optical positioning and stabillity.

I think this Test Bed does have a good design to pull heat away from the TE hot side....but....as this is my first attempt...future iterations could increase heat transfer...there is always room for improvement. I am sure this problem has been examined...and solved...MANY times over...by " others "....Any " Other "....feel free to jump in....and light the way !!!

Next question.....Do we need a controller for the TE...at all ???? Since there seems to be no downside to cooling down to 0 c...no harm to the LD....no condensation issues (Eric... I seem to remember that you noted that down to 0 c...the AL baseplate acted as a " Getter ")....why do we need to control ?? Just have the TE run balls out ??? In a projector real life setting....there will be a range of red LD modulation....but...I would think that most show content will have an average amount of red. So...why control the TE....just let it " Free Range " haha

Thanx for the advice !! I will be experimenting !!

Beam out_______

PB...just saw your question...Yes...the hot side of the TE is Artic silvered right next to the CU radiator....which is directly next to the fans.....oh...and I supppose one could double stack the TE plates...I know that is done....And...for the record...YES TE is systemically inefficient....BUT....it is simple...and neat....and...at the end of the day...we are not trying to land a man on the moon here....just trying to keep the LD happy....increase PO stabillity....get alittle more PO....and perhaps....increase the MTTF alittle !!

[polishedball]

What amp tec are you using? also even if pocketed under the diode would you still risk alignment issues as it heated or cooled. The material i would think would heave slightly?

[planters]

CD,

Firstly, don't bother with any controllers. This is completely unnecessary. Pick the optimal drive voltage for a given TEC based on the efficiency curve that comes with the unit. Custom Thermoelectric as one example provides this curve for each of their standard models. Then just drive the TEC @ that Voltage and let it act as a pure resistance load (as it is)

For this simple an installation I would forgo the complexity of a screwed down, squeezed sandwich with thermal grease and rather get some Arctic Silver (as one example) two part epoxy and as long as the components are no more than 25mm in lateral dimension, just glue the mount to the TEC and the TEC to the heat sink. This will be very stable both mechanically and thermally, provide even better heat flow, is more compact and easier to do. The diode needs to be sealed from the atmosphere at lower temperatures and if sealed the coldest part of the diode mount acts as a getter for the RESIDUAL water in the sealed space, but cannot prevent a never ending source of room humidity from ultimately condensing on the lens.

[Stiffler]

Just chiming in to get notifications, carry on

[steve-o]

Great work and topic CDB! I've been wanting to TEC a pair of P73's, after seeing planters experiments with the dry-ice. It's starting to seem fairly uncomplicated if a controller-less peltier 'chiller' is possible . Less space required, less stuff to tweak, break or go wrong.

[polishedball]

Great work and topic CDB! I've been wanting to TEC a pair of P73's, after seeing planters experiments with the dry-ice. It's starting to seem fairly uncomplicated if a controller-less peltier 'chiller' is possible . Less space required, less stuff to tweak, break or go wrong.

not to hijack mr. cdbeam's work. Just not sure if this ever landed here when pl was down. Quad P73 doing 3.3W cooled to 18C