Hi;

I have not heard of "Switching Power Supplies" before, and really don't know how they work. Are they substantially different from transformers, bridges and capacitors?

And the real question is, would they act as a good source for LEDs? Let's say we're looking at one that is 12VDC and 1 or 2 amps.

Last, is it true that they get hot? This equals inefficiency, or even a slight level of danger to home and infrastructure, no?

Thanks, Tom.

 

A switching power supply works by turning a transisor on and off, never running it in it's analog state. A transistor dissipates the least wattage either completely on or completely off. It then uses filter circuitry, an LC network (that also includes a diode to use the coil as a flyback to further improve efficiencies) to get the ripple down to a few millivolts. This ripple is integral to the switcher, it must exist, since the transistor turns off when a peak voltage is reached, and turns on when a miminum voltage is reached. Because their is minimum heat generated as a waste product, switching power supplies are extremely efficient, 80% or better isn't unusual.

If the designer so desires you can filter the resultant voltage even further with a linear regulator to get rid of this ripple.

You can't run an LED without a current limiter, usually a resistor. Other than that a switcher will run just about anything.

More info about LEDs can be found here...

http://forum.allaboutcircuits.com/showthread.php?p=117636#post117636

LEDs can get hot, but compared to light bulbs their pikers.

 

see the referenc

 

Bill and Manba;

Thanks for the replies. I have more reading to do, so I'll work on that link, Manba.

Bill, if a current limiting resistor were used, as you outline in your LED Bible, would this type of supply be good for LEDs in your opinion? Stable, etc.?

And it's odd that my work cohort thought that this type of supply gets particularly hot, I mean, when compared to a wall-wart. Bill, I think you may have assumed that I was talking about LEDs getting hot, but I did mean the power supply.

Maybe it's a case of "hot" being a relative thing, but his observation did concern me. I would never bury a supply permanently inside a wall, but I might mount it in a steel box that hides inside an access panel. I'd prefer to not have to circulate air to cool components such as this. Again, the heat concept is worth being aware of.

How much better is this supply for efficiency than a typical adapter? I am still considering using low voltage to power LED lighting, as opposed to 110VAC.

Thanks again gents, Tom.

 

Any kind of supply is good for LEDs. Are you talking about replacing light bulbs? What wattage of bulbs (they don't translate well for LEDs, but it is a start)?


BTW, it is possible to use AC with minimal parts with LEDs. Have you read the AAC book concerning LEDs?

http://www.allaboutcircuits.com/vol_3/chpt_3/12.html

 

Hello,
SMPS(Switch Mode Power Supplies) are much more efficient than linear supplies. SMPS operate in switch mode to provide the required DC o/p. And the transformer size decreases because of switching.
Heat sinks can dissipate away the heat generated within.

 

Hi Bill and Ritz;

Bill, I am in fact looking at replacing all home lights with LEDs. So the wattage is a variety throughout the home, but I'll start in the Home Theater where I have several 50 watt halogens. Dimmable is critcal to me.

The 2 72LED bulbs I bought thru Ebay were brighter by far than a 50 watt halogen, but definitely the wrong colour of light for me. A very cool blue white, where I'd prefer a warmer slightly yellower white. This is significant from a hobbyist viewpoint, because all of a sudden we're dealing with very different light output. Much lower lumen power for yellowish white. I haven't tried a warm white bulb yet. That'll be soon.

However, I am convinced this can happen regardless. You talked about running AC into LEDs with minimal parts. This is precisely what I was starting to discuss, however my thread was terminated. I was not aware of the concept of isolating AC from the end user with a transformer, and it would have been beneficial to learn about this, but again, it was declared verboten in this forum.

I would ideally like to skip the transformer if all could be made safe. For example, when I do wiring in my household, there is no isolation from line voltage, except for the standard infrastructure, such as insulated wires, conduits and proper boxes. I am interested in knowing if AC can be turned into DC, kept at the same approximate voltage and then simply fed into LEDs thru a high voltage PWM dimmer. That was one approach to my goal. So if you know of a safe way to accomplish this (and I am very comfortable with home wiring and doing it properly) then feel free to elaborate.

Ritz, sorry to keep you waiting ! Just curious how you'd attach a sink to one of the pre-sealed switching supplies. I guess you're talking about making the supply yourself, then adding a heat sink, huh? One thing I'd like to do is not "bury", but rather "contain" all the circuitry within a wall. This would be done inside a very removeable access hatch and any circuitry or power supplies would be sealed safely inside a metal box. So the idea of sinking or blowing cooling air into the box isn't my first choice if I can avoid it. This is why I was curious to know how hot these switching supplies typically get. Also, I think I'd prefer to just buy a power supply, and not make one. I have to pick my battles, and I think I'll rely on a commercially available supply and get creative from there.

Well, any thoughts, feel free. I thank you both for your input this far. I am convinced that lighting my home entirely with LEDs is do-able, and not rocket science, but it will take some tinkering. I'm willing to tinker.

Cheers, Tom.

 

So, did you read the article?

We are always concerned about safety here at ACC, electricity can kill, swiftly and surely. But since it is mentioned in the text book it is an open topic for discussion.

I know what you mean about color, I've run into it with florescent bulbs. There are companies that address these concerns, but their products aren't cheap. That is also in the same article.

One last item, commercial dimmers and LEDs don't mix. The way both work are incompatible with each other.

 

Hi Bill;

Actually I hadn't read the article before you asked, and I have a bit of a confession to make. Which is, that when I'm confronted with a multi-page article about electronics, complete with graphs, stats, diagrams, etc, it can feel a bit daunting to absorb the whole thing. I guess my absorbtion rate is lower than other hobbyists, however, I did go back and focus on the part about AC powering of LEDs. I found it interesting and I got most of it. The only thing that surprised me a bit was that a protective diode isn't used right in the same leg as the LED in series (as opposed to simply added in parallel). I guess this just means that electron flow will simply take the path of least resistance, namely, through the diode in the forward direction. And if this method truly protects a LED, then why not use another LED in place of the parallel protective diode, which is mentioned in the article.

So then, how would a person go about ganging up a bunch of these LEDs, all to be powered with 110VAC, in one GU-10 package? Do it with several pairs of LEDs in parallel, each with it's own capacitor? And wouldn't the cap have an adverse effect on dimming, since it smooths out the pulses, (from a PWM for example) rather than forcing the LEDs to dim in response to the pulses?

My thinking thus far has been to supply a group of LEDs, mounted in one GU-10 bulb, in such a way that the bulb is fed pulses, at 1khz or higher (the threshold for human eyes seeing flickering, according to a Youtube demo that I watched) and have those pulses modulated in pulse width to vary the power and therefore dimness setting. My gut feeling is that any capacitor after the PWM would have an undesirable effect on the pulses and dimming performance. If that is true, and I'm in no position to be sure, then that leads us back to feeding LEDs with less than 110 volts DC thru a PWM, or feeding them some other way with 110VAC pulses.

Am I right about any capacitor mounted near the LEDs having an adverse effect on dimming?

Feel free to steer me away from flawed layman's logic.

Thanks again Bill, Tom.

 

Actually I hadn't read the article before you asked, and I have a bit of a confession to make. Which is, that when I'm confronted with a multi-page article about electronics, complete with graphs, stats, diagrams, etc, it can feel a bit daunting to absorb the whole thing. I guess my absorbtion rate is lower than other hobbyists, however, I did go back and focus on the part about AC powering of LEDs.

As someone who has faced this same experience through 4 years of EE courses... I have to say, don't be discouraged by all of the daunting information. Just keep going back through everything until you really understand it thoroughly. Start with the basics and go over it until it really clicks.

Good luck

 

Hi Sprintf;

Good advice, and thanks. That repetitive method is really how I tackle something like this. I might not read the whole article, but I tend to focus on parts that answer my question of the moment.

I don't feel that I'm starting from scratch, however, as I used to build a ton of electronic circuits. So I don't fear the concept of jumping in. One thing I did learn in Aerospace school was that electronic circuits can be enormously complex, and I always found it (as well as stress analysis) to be the toughest classes.

Overall, it's good that these forums exist.

Cheers, Tom.