Switch to English Language Passer en langue française Omschakelen naar Nederlandse Taal Wechseln Sie zu deutschen Sprache Passa alla lingua italiana
Members: 71,050   Posts: 1,561,104   Online: 881
      
Page 4 of 5 FirstFirst 12345 LastLast
Results 31 to 40 of 44
  1. #31
    polyglot's Avatar
    Join Date
    Jun 2009
    Location
    South Australia
    Shooter
    Medium Format
    Posts
    3,308
    Images
    12
    Red LEDs have a Vf of about 1.5V, which means you'd need about 100 of them in series to make up the ~160V you get from rectified 115V. If there are fewer than that, then the efficiency with using a linear regulator is terrible because most of the voltage is dropped by the regulator, which must dissipate it as heat.

    Take that Amazon example (Simonh82); 24 red LEDs in series is 36V. 160V from the rectifier means that 77% of the energy would be going into the linear regulator and 23% into the LEDs. The heatsinking for that much power will cost 5-10x as much as putting in the proper switching regulator which is about 85% efficient. It's a 5W bulb, which probably means 5W into the LEDs, which means probably 5.8W input power; if it was linear then that would mean 22W input power and heatsinking for 17W, which is HUGE, not to mention less efficient than an incandescent. It'd also be incredibly unsafe because the LED terminals are effectively live.

    Transformers are heavy and very expensive. More expensive than switching regulators, the latter being based around tiny, high-frequency transformers rather than comparatively big brutish ones that must work at 50Hz and therefore require huge core cross-section to avoid saturation. The (50Hz) transformer to run this bulb would be about 3x larger than the bulb and pricier than the retail price of the bulb. The switching regulator would fit trivially on the back of that panel that holds the LEDs, with room to spare.

    The Amazon bulb is also "110-220V", which is about as solid an indication you can get that it's got a switching regulator on-board. Cheaper bulbs will have cheaper regulators that don't support the broad input range.

    Switching regulators, while being a complete bitch to design, are fantastically efficient in power, space and cost. They beat the pants off the other alternatives by an order of magnitude and manufacturers know that.
    Last edited by polyglot; 06-24-2013 at 01:30 AM. Click to view previous post history.

  2. #32

    Join Date
    Jan 2007
    Location
    Sydney, Australia
    Shooter
    Multi Format
    Posts
    303
    Images
    1
    Quote Originally Posted by polyglot View Post
    "Do not meddle in the affairs of wizards for they are subtle and quick to anger". Switching power supplies are well and truly into that territory.
    This is very well put. As someone with a degree in electrical engineering who has made non-mains rated switching supplies I would really discourage anyone from from fiddling with a mains rated switching supply...

  3. #33
    PeterB's Avatar
    Join Date
    Apr 2005
    Location
    Sydney, Australia.
    Shooter
    Medium Format
    Posts
    597
    Hi polyglot, you missed the fact that I included a transformer in the circuit (Tx). With that in place before the linear reg dropping down such that the rectified DC voltage is 17V then the power lost across a 15V (O/P) linear reg plus that lost in the Tx should be much less.

  4. #34

    Join Date
    Jan 2007
    Location
    Sydney, Australia
    Shooter
    Multi Format
    Posts
    303
    Images
    1
    Quote Originally Posted by PeterB View Post
    Hi polyglot, you missed the fact that I included a transformer in the circuit (Tx). With that in place before the linear reg dropping down such that the rectified DC voltage is 17V then the power lost across a 15V (O/P) linear reg plus that lost in the Tx should be much less.
    I'm not sure he did:

    Quote Originally Posted by polyglot View Post
    Transformers are heavy and very expensive. More expensive than switching regulators, the latter being based around tiny, high-frequency transformers rather than comparatively big brutish ones that must work at 50Hz and therefore require huge core cross-section to avoid saturation. The (50Hz) transformer to run this bulb would be about 3x larger than the bulb and pricier than the retail price of the bulb. The switching regulator would fit trivially on the back of that panel that holds the LEDs, with room to spare.
    The take-home message being that no transformer capable of doing the job would fit inside the bulb case, which is why all of these LED bulbs use SMPS.

    FWIW it's pretty easy to string a few LEDs together with plug-in power adapter (either transformer or SMPS based) and a resistor as a DIY safelight. It's exactly what I did when I was printing in my bathroom at home, but it's not quite as easy as the direct bulb replacement.

  5. #35
    polyglot's Avatar
    Join Date
    Jun 2009
    Location
    South Australia
    Shooter
    Medium Format
    Posts
    3,308
    Images
    12
    Quote Originally Posted by PeterB View Post
    Hi polyglot, you missed the fact that I included a transformer in the circuit (Tx). With that in place before the linear reg dropping down such that the rectified DC voltage is 17V then the power lost across a 15V (O/P) linear reg plus that lost in the Tx should be much less.
    Initially I did, yes: some people do make very long series strings of low-voltage bulbs, e.g. xmas lights. Then I added a paragraph regarding baseband (50Hz) transformers. Getting 5W DC out of one requires it to be much larger than a GU10 bulb and it's still going to be about half as efficient as a switching supply at several times the total BOM cost.

    The short answer is: if there are LEDs in a 115V or 220V GU10 (or similar) bulb, there is absolutely a switching regulator in there. It's the only physically and economically viable approach. And they have a designed-for input voltage range, which is set not just by component tolerances by also by the actual switching design (selection of frequency, feedback topology, duty cycle range, magnetics-reset approach, etc, etc). Moving the input voltage outside of the design range will cause the switcher to fail even if no components go outside of their voltage tolerances.

    Raised input voltage will usually cause (depending on the control approach) either reduced duty cycle or reduced frequency. Both can lead to it operating in discontinuous mode, which means (unless it was specifically designed for that mode) that the transformer (edit) flux gets a bias and can saturate the core. So no energy gets through and everything gets hotter until something lets go.

    Sorry to be a downer, but there are plenty of 240V-compatible red-LED bulbs on the market. Much easier and cheaper to buy one of those than fiddle with a $10 sealed/disposable product in a way that's not really possible unless you want to spend $30 on replacement parts (and a couple days of EE labour) to go in it.
    Last edited by polyglot; 06-24-2013 at 08:24 AM. Click to view previous post history.

  6. #36
    clayne's Avatar
    Join Date
    Sep 2008
    Location
    San Francisco, CA | Kuching, MY | Jakarta, ID
    Shooter
    Multi Format
    Posts
    2,838
    Images
    57
    And the safety of your house while we're at it.
    Stop worrying about grain, resolution, sharpness, and everything else that doesn't have a damn thing to do with substance.

    http://www.flickr.com/kediwah

  7. #37
    PeterB's Avatar
    Join Date
    Apr 2005
    Location
    Sydney, Australia.
    Shooter
    Medium Format
    Posts
    597
    Thanks polyglot (and David). Yes I overlooked some very important facts about using a stepdown Tx in such an application, that while it might work and be not too inefficient, it would be way too large and costly to justify using. My main reason for wanting to modify that LED light (if easy/safe to do so) was that it has been thoroughly tested by others on this thread to be paper-safe. This puts it streets ahead of other cheap 240V LED luminaires out there. I might contact the retailer to see if they have a 240V model from the same manufacturer.

  8. #38
    PeterB's Avatar
    Join Date
    Apr 2005
    Location
    Sydney, Australia.
    Shooter
    Medium Format
    Posts
    597
    I think I found a 240V version here. I have also emailed superbrightleds and will let you know what they say.

  9. #39

    Join Date
    Aug 2011
    Shooter
    Medium Format
    Posts
    166
    Peter, the one I linked to earlier is 240v and as I mentioned I've had paper under it for up to an hour. There is zero fogging.

    The bulb shipped from the China or Hong Kong. I bet the seller is also on amazon AU or ebay.

  10. #40
    PeterB's Avatar
    Join Date
    Apr 2005
    Location
    Sydney, Australia.
    Shooter
    Medium Format
    Posts
    597
    Quote Originally Posted by PeterB View Post
    I think I found a 240V version here. I have also emailed superbrightleds and will let you know what they say.
    I noticed it is about half the power of the original 110V one (1W c.f. 1.9W) which now makes me think they are indeed different globes and manufacturers.

Page 4 of 5 FirstFirst 12345 LastLast


 

APUG PARTNERS EQUALLY FUNDING OUR COMMUNITY:



Contact Us  |  Support Us!  |  Advertise  |  Site Terms  |  Archive  —   Search  |  Mobile Device Access  |  RSS  |  Facebook  |  Linkedin