datasheets.com EBN.com EDN.com EETimes.com Embedded.com PlanetAnalog.com TechOnline.com   UBM Tech
UBM Tech

ed ro

's profile
image
President

Ed Rodriguez was founder/CEO of Theta-J Corp(now the Clare division of IXYS Corp). In that role he developed the optical MOSFET solid state relay, now a global defacto standard in the telecommunications industry. He is a veteran of the power semiconductor, power supply and LED lighting technology industries.Early on, before becoming an entrepreneur, he was a product manager with Unitrode (now part of Texas Instruments.) Keynote speaker at 1986 MIT national conference on advanced power electronic technology, he has published dozens of feature articles in the leading technical trade magazines. He also has been noted in Forbes and Wall St Journal. He holds 22 patents with 8 pending, virtually all involving successfully commercialized products. In 2005 , he formed OptoThermal Technologies to focus on thermal management technologies related to high-power LED lighting.


ed ro

's contributions
  • 10.30.2013
  • That 60W-equivalent LED: What you don’t know, and what no one will tell you…
  • Bret--- Efficiency and heat are two different things. A 60 watt incandescnt lamp i svery inefficient and gives off a tremendous amount of heat--put under a small metal plate and it willheat you dinner.Its fi,amnent gets to THOUSANDS OF DEGREES but doe not blow out because it is mad eof Tungsten.The outer glass gets extremely hot but nowhere as hot as the light emitting filament inside. The led equivalent gives off same light but only with about 10 watts but that wattage shows up as heat in the bulb housing so you feel it--but it is still only a fraction of the heat in a 60 watt filament bulb-just that you are protected frm touching the filament by the glass outer shell. Your flashlight doesn't have anywhere near the attage or light output of the led bulb--probably only a watt or two. It does not give off more light--just seems that way.The flashlight hasa reflctor which focuses tghe emitted light into a bright spot or circle and when you look at flashlight it seem "super bright" Ther eisa differnce between light "intensity" as see by a viewer and the total light output. If you took off the reflrcto form a flasjhlight it woud only light up the room a fraction of the led bulb.Simiallry if you put the proper reflector over an l3ed bulb it woud be a dramtically brighter "flashlight" The flash light has much less light but al of it is put in one spot so "seems' brighter.Becasu erhe flasghlight is such lower power, it of course has way less heat, many led flashlights have less than a watt of heat even thougth they can project a very bright "spot"on the wall. No mystery here but there is a little bit of science.You question is common "If LEDs are so darn efficient ,why do they get hot?"--only because what little heat exists is right there for you to touch. A paper match does not give off much heat at all --- a super tiny fraction of the heat of a regular light bulb--but it sure would burn your finger to a crisp.I hope this helps.
  • 11.19.2013
  • High bay LED lighting: Specmanship is alive and well
  • Not sure if you are pulling my chain but here goes. RPI is Rensselaer Polytechnic Institute in Troy NY. It is one of top engineering schools in US but also unique in easily being the leading university in LED lighting technology, especially in its graduate school programs. They have published papers on every imaginable aspect of LED technology--from chip and phosphor physics to luminare design aesthetics and commercial product analysis-- and have a formidable R and D capability in Troy NY Now for the air cooling--don't get me started.You might want to read my other entry on this site some weeks ago on www.edn.com/electronics-blogs/led-insights/4422914/cooling-high -power-leds There is no question that air cooling ( ie active cooling) can dramatically reduce the size and weight of high power LED fixtures. But there is a real hangup among many or most mfrs about fans or other air moving devices. That wall is slowly coming down, brick by brick ,as an increasing number of "enlightened" companies embrace the advantages, I find it amusing that the Internet is gaga over the prospect of a Google-controlled, driverless car but engineers still apprehensive (due to lack of knowledge) about a rock-solid cooling technique which can reduce weight by a factor of 4-5 , size by a factor a 2, knock 25 % off cost, lower shipping cost by a factor of 4 and significantly simplify installation.
  • 11.19.2013
  • Overcome the challenges of driving parallel LED strings
  • You know the old expression: "Don't build me a clock ,just tell me what time it is". This piece on current sharing is WAY more complicated than it needs to be--and does not reflect what folks really typically are doing out there- I realize this is well intentioned effort to sell TI product, and I'm OK with that. The issue involved is mainly associated with higher power LED luminaires--over 25 watts., with strings up t about 50V (nobody wants to go higher than that because of UL SELV complications-----Cree please take note) For COB arrays, which go up to 100 Watts, the issue is "not an issue" unless your luminaire is rated over 100 watts.--Although many high power fixture people, for visual purposes, still want (or need) to use large numbers of paralleled 1-2 watt devices instead of a couple 50 or 100 watt COBs But in any event every engineer with basic analog power management savvy knows he can simply put in a really cheap TO-220 low dropout voltage regulator, operate it in a constant current mode in any string and set its max current with a resistor . By the way any imbalance up to 10-12 % is not evident to the human eye !! and even 15% is not very obvious. No matter what the approach, there will be the tiny bit of power lost in the sense R--no big deal. The lengthy description here is laudable as an intellectual exercise but sort of like chartering a bus to go from one side of Times Square to the other. Hundreds of folks are already addressing this issue in markedly simpler ways. In an increasingly price-competitive (ie nasty) LED luminaire and lamp world, there is little room for design solutions which are "neat" but won't pass the "Do-I think-my-competitor-can-figure-out-how-to-do-it-simpler/ cheaper" test
  • 10.17.2013
  • Cooling high-power LEDs: The four myths about active vs. passive methods
  • Roy-- Some specifics, since the fan and sink in picture is now used in several production LED luminaires, With 60 mm fan at 7V and 1500 RPM, thermal resistance is about 0.75C/watt. With same fan and heat sink but at 12 V, and close to 2500 RPM, thermal resistance drops to about 0.6C/watt, pretty close to being able to keep your 20 watt laser heat generator to a 10 degree rise. Sink has fourteen 1.4 X 2.3 X .030 fins. We're also using a similar but larger sink (5" X 6" base) 37 fins of 1.4" X 5" X .060--seems close to what you described. With a 120 mm fan on top of fins at less than 2,000 RPM (very quiet) and 1.5" distance under fan to allow adequate air intake, thermal resistance is only 0.12C/watt. For your 20 watt laser app, the temp rise would be only couple degrees. So you can see there's some pretty darn good cooling here. You did your thing in 1992. There is always an explanation for these things. You gave no details about your fan or the comparative thermal resistance difference you experienced between lateral and vertical air flow---- no need to beat this to death--- but something obviously amiss for you to get such apparently lousy results. You were using a TE cooler and I assume cooling "it", not the laser heat source directly. So maybe you are saying the whole "system" did not provide adequate heat removal result.. I have a good bit of experience with TE modules and long ago pretty much crossed them off my list as "interesting" but inefficient and rarely useful for any serious heat removal task----for last 50 years pretty much a "solution in search of a problem"--the "Wankel engine" of thermal management.
  • 10.08.2013
  • True or false? High-power LEDs don’t generate IR heat in the forward direction like a filament lamp
  • arioda-- Let me comment that in my application which started all this, I was using phosphor coated sheet made of PET. After my melting episodes I learned from vendor that the phosphor on that sheet has organic binders. Since 95% or all remote phosphor applications to date have only involved multiple, spread-out low power LEDs , the heat concentration just hasn't been there. so the issue was not apparent. But with 25,50 and 100 watt COB's, a different ballgame.For our own prototypes we are now switching to phosphor with inorganic binders and glass as the substrate--so there can be no melting.