@LevitonDave - Yep, a new product line might do it. But there is still the issue of the cost of that new product having the expensive LED. Oh, and let's not forget we need to get the marketing and sales folks to find a market to buy our new design. Just can't go making something without having someone who wants/needs to buy it, huh?
@Rerepete - Using the same component in other products is a great idea. Unfortunately, the other analyzers in the product line use a different technology. Those lines don't use "the bulb." That's one of the reasons we couldn't justify going the route of the LED solution. But, with all that is going on in the LED industry right now, conditions may make it feasible to make the custom LED at a lower cost.
@LevitonDave - I completely agree concerning reputation and reliability, but unless the issue is of a much higher degree than this bulb, it doesn't get much attention at the management level, and I have little "pull."
@Rerepete – Thanks for the ideas. We looked at quite a few alternatives, but it's tough to beat a cheap bulb. Ya know what I mean? Try convincing managers and a CEO. All it takes in one "sit down face-to-face" with the boss gang to make one realize that you'd better have all the right answers and all your ducks in a row or things are just not going to fly. These folks are all about $$$, and a cheap bulb is tough to beat. Even if I do the number crunching with good MTBF numbers, almost any idea gets shot down.
"Does applying a PWM/Dimming have any significant impact on the life of an LED?"
Not noticeably. I don't think anyone has really studied the subject. I seems to recall reps from Lumileds said no effect from high frequency. You might argue that dimming means lower current and thus leads to longer life.
@Rerepete - I realize that. NRE (Non-Recoverable Expense) is a reality of all developmental work at all levels of design/development. But only the company doing the design/development can specify what those minimums are, and all the customer (my employer in this case) can do is consider the response.
@Rerepete - I am not an LED component designer, so I can't say too much authoritatively. All I can do is repeat what the experts told me. They stated that they could develop an LED for my specified wavelength, but to do so, we'd have to order a minimum quantity. That quantity was well beyond our needs, so the effort died at that point.
@flared0ne - Don't forget that the micro has its own ability to do its internal calibration to make up for variance between one bulb and another and also due to aging of a bulb, thermal drift, ambient drift, etc.
@Carol, I have a question that may apply to tomorrows session. I see references to white LED changing wavelength (colour) with varying forward current. Are special dimming techniques required because of average current during PWM or is it more based on peak current?
Isn't the wavelength of a mercury bulb dependent upon the elemental characteristics of Mercury (plus whatever phosphors you add to manipulate the output spectrum)?? I would imagine you could approximate portions of the emitted spectrum, but LEDs don't fit a "gut feel" for a good alternative somehow. And (of course) I could be wrong.
@Carol - At one time I worked on a project involving gas chromatography. It used a mercury light bulb on one end of a tubular sample chamber and optical detector on the other end. I don't remember the wave length of the mercury bulb, but it was chosen for its wavelength as it pertained to the gas being detected. The bulb was an MTBF issue, so I wanted to change the design to an appropriately chosen LED. Unfortunately, the only way to get it was to have them custom made, and we couldn't meet the minimum quantity demands, so the effort was dropped.
I can't help but wonder if all this new activity in the LED industry can bring this project back to life? What do you think, Carol?
I mentioned this yesterday -- do LEDs exist which have a single package containing the primary emitter and a secondary detector which can monitor emitter output and control an integral on-die feedback loop, maintaining a desired output level?
I am intnerested in greenhouse applications. I have an organic farm under construction in northern VT. I know I will need to have a lot of light to get the crops started early indoors. We will be off-grid, so energy consumption is an issue. I am an electrical/electronics engineer specializing in control systems and this is a new direction for me. I have not yet designed with LEDs, but have lots of LED lighting in my life from COT sources.
I'm having my students build underwater robots with cameras. The camera work great in the swimming pools, but nothing in the murky ponds. So I want them to build a few LED arrays so that we can increase our underwater vision.
5. How to Enter: To participate, you must register by clicking on the Digi-Key Continuing Education Center (on DesignNews.com) registration link and accurately completing the registration form, and attend the course. Mailing and email address must be correct in order to receive a Starbucks E-Gift or Card.
6. Limitations on Entry: The system will only allow you to register once.
WRT Slide 10 yesterday' slide deck, and the connection to 120VAC, I note that bridge rectifiers can be very inefficient and even at 70mA or so, the theoretical peak voltage of 168VDC rapidly drops to 145VDC or so even under this moderate load. At 100mA it is closer to 135VDC.
BTW, since the diodes have a turn-on knee, a bleed resistor is highly recommended across the DC capacitor.
After reading back a bit, while there's nothing much else going on...
Regarding IR and UV LEDs -- if you want to get really technical (trying to approach diminishing returns on the whole "predictable results" curve) you will probably also want to search for a nomograph on the propagation/attenuation versus frequency of radiated energy passing through the atmosphere. Those propagation curves are seriously non-linear, involving a wide range of opacity/transparency across the electromagnetic spectrum. Personally, I remember being a bit surprised (mostly because it should have been obvious) when someone pointed out that a lightning strike emits a huge amount of xrays -- generally without much effect (other than causing most of the ozone smell after a strike) because the atmosphere absorbs xrays so effectively.
So your IR and UV propagation distance CAN be calculated with some precision.
@Ran: You give me too much credit. I had no idea that you too had logged in early today. It was a happy coincidence. And far from being on top of things, I was simply lazy. I left yesterday's chat page up when I left work and noticed the new postings when I fired it back up today. See you in class in an hour!
Chen, you can google most of the information you're asking for. Yes, you can buy IR LEDs. You can also buy UV LEDs. Most data sheets will also detail the information you're looking for in a graph relating current and luminous intensity.
@Mr. E - You're very welcome for the MTTF/MTBT post and the Buck/Boost posts. I hope that info was of some value. But I'm also surprised at your knowing I posted it. I had to miss the airing yesterday, so I caught up last night. I posted, but I didn't know if anybody would ever benefit from those posts. I'm impressed that you went back and revisited the postings. You're really on top of things, aren't you?
@Mr. E - Wow! I'm totally taken by surprise to see your post. I am NEVER in hear this early. I don't even plan to stay here at this time. I'm just very busy and didn't want to wait until the last minute to set up for the show, so I wanted to download the .pps in advance to save a minute or so. But what surprises me most is that you even knew I was here. How'd you do that?
The streaming audio player will appear on this web page when the show starts at 2pm eastern today. Note however that some companies block live audio streams. If when the show starts you don't hear any audio, try refreshing your browser.
Are they robots or androids? We're not exactly sure. Each talking, gesturing Geminoid looks exactly like a real individual, starting with their creator, professor Hiroshi Ishiguro of Osaka University in Japan.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.