This is indeed an issue with many electronic systems. There are good CAE tools that would allow one to model these situations and to plan proper mitigation. The software is expensive, though, and working with suppliers, like Littlefuse, can help avoid some of those costs.
When I was designing LED lighting fixtures for a major ceiling fan supplier, I would immediately draw out a system block diagram including the power supply. Within this subcircuit block, I included all interconnects, transient suppression devices, and electromechanical devices. A detailed product specification will be written describing the functionality of the subcircuit blocks and their power and electrical requirements (Undervoltage, Overcurrent, etc) of the system block diagram. This engineering approach prevented significant damage to the prototype during testing because of the stringent transient electrical specifications spelled out in the design document.
notarboca, Thank you for the kind response. After seeing blue smoke let out of LED prototypes, there's a point where you start to think logical and document the exact function(s) required for the system/product under development.
LEDs are typically always the INDICATORS of failure, but not always the source of failure. Improper circuit protection against abnormal conditions can just as easily kill a drive component as it would an LED, but the LED will almost always turn off due to any failure. (message to Quality guys...if it doesn't light up, it's not necessarily the LED's fault!)
The failure that I see mostly is when the LED system design does not properly manage the heat generated within the LED and/or drive electronics. Good, worst case analysis should be done to ensure that the LED will always operate in a 'warm and fuzzy' region thus ensuring long and reliable life. (note: I don't have to design against lighting strikes! That sounds tough!)
General rule of thumb; Lumen maintenance failures result from overdriving the LED (high current density), but catastrophic LED failure results from exceeding the Junction Temperature rating.
Very good points! Your absolutely correct regarding thermal manaagement concerns with LEDs and their driver circuits. Lumen maintenance failures caused by overdriving LEDs was the biggest challenge I faced when developing my lighting fixture prototypes at Hunter Fan. Although the LED continued emitting light, the opto-device was overstress by driving the component extremely hard with max current (Ifd). The naked eye couldn't see it put measuring its luminous flux showed the LED being degraded over time. In addition to protecting the overall product's system circuits from high voltage transients, thermal management must be high on the design priority list as well.
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.