New Cooling Technology Uses Air "Bullets" to Shoot Down Overheated LEDs

[jeffosborne]

Here's a fresh article from Scientific American regarding a new way to cool down our LEDs:

http://www.scientificamerican.com/article.cfm?id=led-cooling-bullets

It is a piezo pump pushing the 'air bullets', instead of a massive heatsink or heatsink/fan combo. Evidently Phillips is on board with this method.

Cheers!
Jeff O.

 

[saabluster]

Here's a fresh article from Scientific American regarding a new way to cool down our LEDs:

http://www.scientificamerican.com/article.cfm?id=led-cooling-bullets

It is a piezo pump pushing the 'air bullets', instead of a massive heatsink or heatsink/fan combo. Evidently Phillips is on board with this method.

Cheers!
Jeff O.

I remember reading about this tech quite some time ago. Yet another recycled report.

 

[LEDobsession]

Really cool idea. I really liked their Flash video presentation available on that same article. Be sure to check that out, it really helps you understand how well it works. The idea itself is basically the same as a cooling fan, only it has more velocity in the air being generated. I saw this hoping for a more effective way to cool the light I made but I think this is more suited for CPU/other electronics and non-mobile LEDs (mine would be outdoors on a vehicle). Thanks for sharing!

 

[bshanahan14rulz]

SynJet is old, but interesting, esp. in that there's all this hype and I haven't come across any products to use it. Still, decent idea, turning oscilation to forward motion

 

[evilc66]

What about noise though? I'd assume that it's going to be quiet based on the fact that it is being geared for consumer applications, but there was no mention of it that I could see.

 

[LEDobsession]

Wouldn't imagine it would be, if at all, noisier than a fan it replaces. :shrug:

 

[bshanahan14rulz]

supposed to be silent. Google SynJet

here's a large pic of what I believe is a prototype that students and professors at Georgia Institute of Technology built and tested back in 2003
http://gtresearchnews.gatech.edu/images/synjets39.jpg

As you can see, the main principal is using soundwaves to create a forward flow of air.

 

[blasterman]

Oscillator air pumps have been around awhile. This variation would seem to make the process more solid state.

It's all nifty, but the physical constraint is still based on active CFM air cooling reducing heat sink size. I don't think it matters if the air movement is done with a piezo device like this or a simple fan - air movement is air movement. This device would seem to do away with fan reliability issues.

The reason companies are scrambiling on this is because right now metal prices are low, but won't last. Sooner or later bulky heatsinks are going to get expensive and active cooling will be more profitable.

 

[bshanahan14rulz]

also:
piezo transducers are smaller than fans
they are also quieter than fans. You have a vaporizor, right? (the one for water, for when you have chest congestion) That has a piezo-electric element but it's at such a high frequency that you can't hear it.
I don't know how much power they consume, but I bet it's less than a fan.

 

[blasterman]

Since we're dealing with the straightforward kinetic process of moving air, my intuition is that piezo likely doesn't have an efficiency advantage over a standard radial fan. The physics on this are pretty closed loop and haven't changed in centuries.

Where the Piezo does have an advantage is small scale and directionality because radial / turbine fans have issues in confined spaces.

We have to remember that the microprocessor industry has thrown armies of engineers at this issue for decades, and they all end to come back with the same solution of just strapping a fan to a bladed heat sink with the possible exception of addming Heat pipes to the mix.

 

[bshanahan14rulz]

think of it this way: this technology might be added to a lineup of coolers that a company offers.

company might offer, from standard to high performance, extruded aluminum with fan, heat pipes with aluminum fins and fan, heat pipes with copper fins and fan.

Well, that takes care of the consumers who want high performance, but many consumers also want quiet fans. piezo is quiet, no bearings to wear out and make noise.

I doubt it pushes more air than even the cheapest CPU fan, but it can be as quiet as a passive sink. This is where I think they'd make the most money.

 

[evilc66]

I don't think the intent of this device is to replace fans, but to go into products where a fan would not normally fit, like commercial LED lighting (bulb replacement type lights).

 

[KeithInAsia]

I've worked with their PAR 38 product. If it weren't for the customer's request - I would have not bought into it. It's work marginally and their arguments for the product just don't out weight the fact that as a cooling device, it's an inferior cooling device. On top that -- their retail prices are -- well, goofy. I suppose when you secure millions of dollars of capital venture -- you have to get it back from someone.

I doubt this company will survive in the LED business.

 

[qwertyydude]

I don't think the intent of this device is to replace fans, but to go into products where a fan would not normally fit, like commercial LED lighting (bulb replacement type lights).

I'd have to disagree here since this fan is only 20 mm square and at only 20 db is quieter than a whisper.

http://www.mouser.com/ProductDetail...=sGAEpiMZZMvuXzJ9NcThaqBG/Si4pPVa6eiuU1anOg4=

 

[blasterman]

I agree with Evil and think he's made the right point.

The problem with small, radial fans such as the one liked above is they are very inefficient in small spaces. The center hub for instance becomes larger the smaller the fan is and creates a dead zone where no air flow occurs.

Also, radial fans need to be baffled to prevent air from simply spilling over the sides. Note how desktop computer makers migrated to 120mm fans practically overnight when they saw how much better they worked than multiple 60 and 80mm fans.

When dealing with space confined LED fixtures the engineers are contending with very exacting thermal tolerances. This is why the piezo does a better job solving the problem.

 

[jeffosborne]

Good points blasterman!

Another reason to NOT use a fan might be for power savings. I wonder if anyone has a handle on current consumption for a piezo air-driver. I know piezo transducers, which I have designed into handheld terminals as 'beepers', require only a tiny amount of current.

Also, KeithInAsia, we are fortunate to have some insider views on the topic. Thanks for your input!

Jeff O.

 

[KeithInAsia]

I forget to mention something..... I like to conserve energy when possible and so when an LED system is cold, I like to turn off items like fans, etc. This synjet people tripped up and forget to nicely shut down their device and when you remove power from it - their circuit has some kind of an inductive collapse and you hear a rather large "popping" sound. I had to leave the unit running full time to avoid that noise during operation. I'm sure they can over-come the issue with some programming, but hey -- how can a larger multi-million dollar invested firm over-look something this simple.

Their units are not small either. They are bulky because they are like encased speaker diaphragms. Granted, they can run a tube into a tight space to transfer the air, but over-all, they are larger than standard smaller fans.

I don't really thing the humming of the unit is all that silent. However it is less noisy than a high speed fan. Fans have to be around 4k RPM to minimize sound.