XPost: uk.d-i-y, alt.home.repair   
   From: Gothberg"@internet.co.is   
      
   On Fri, 21 Dec 2018 01:06:46 -0000, Rod Speed  wrote:   
      
   >   
   >   
   > "William Gothberg" <"William Gothberg"@internet.co.is> wrote in message   
   > news:op.zucezsmio5piw3@desktop-ga2mpl8.lan...   
   >> On Thu, 20 Dec 2018 22:14:31 -0000, Rod Speed    
   >> wrote:   
   >>   
   >>>   
   >>>   
   >>> "William Gothberg" <"William Gothberg"@internet.co.is> wrote in message   
   >>> news:op.zucb6od6o5piw3@desktop-ga2mpl8.lan...   
   >>>> On Thu, 20 Dec 2018 21:35:49 -0000, Rod Speed    
   >>>> wrote:   
   >>>>   
   >>>>>   
   >>>>>   
   >>>>> "William Gothberg" <"William Gothberg"@internet.co.is> wrote in message   
   >>>>> news:op.zub8lqzpo5piw3@desktop-ga2mpl8.lan...   
   >>>>>> On Thu, 20 Dec 2018 19:11:36 -0000, Rod Speed   
   >>>>>>    
   >>>>>> wrote:   
   >>>>>>   
   >>>>>>> "William Gothberg" <"William Gothberg"@internet.co.is> wrote in   
   >>>>>>> message   
   >>>>>>> news:op.zubnqbkho5piw3@desktop-ga2mpl8.lan...   
   >>>>>>>> On Thu, 20 Dec 2018 09:36:02 -0000, Jon Fairbairn   
   >>>>>>>>  wrote:   
   >>>>>>>>   
   >>>>>>>>> "William Gothberg" <"William Gothberg"@internet.co.is> writes:   
   >>>>>>>>>> Agreed. All I can detect (with my digital camera) is that   
   >>>>>>>>>> one brand of LED light I have flickers about 5 times less   
   >>>>>>>>>> (not sure if it's smother or faster) than the others.   
   >>>>>>>>>   
   >>>>>>>>> Try a longer exposure and move the light rapidly relative to the   
   >>>>>>>>> camera.   
   >>>>>>>>   
   >>>>>>>> I wonder, if I fed the lamps with mains voltage DC, simply a bridge   
   >>>>>>>> rectifier and a huge capacitor, they'd reduce their flicker.   
   >>>>>>>   
   >>>>>>> Wont work at all if they use capacitor droppers and   
   >>>>>>   
   >>>>>> I made a few of those to power LEDs to indicate the function of my   
   >>>>>> central   
   >>>>>> heating.  I'm looking inside the flickery lamp just now (£15, 20W).   
   >>>>>> Without undoing the glue holding the PSU onto the inside of it, all I   
   >>>>>> can   
   >>>>>> see is probably: the mains going through a large bipolar cap, a tiny   
   >>>>>> resistor (to discharge it safely?), a bridge rectifier, another very   
   >>>>>> large   
   >>>>>> resister (to limit the LED current more accurately?), then a 400V   
   >>>>>> 4.7uF   
   >>>>>> capacitor (which is bulged).   
   >>>>>   
   >>>>>> A capacitor dropper with a rectifier and smoothing capacitor after it?   
   >>>>>   
   >>>>> Yep, that's what it is.   
   >>>>>   
   >>>>>> The one I made has no smoothing cap, just mains to cap to resistor to   
   >>>>>> bridge to LED.   
   >>>>>   
   >>>>> Yeah, not need for one if you don't mind the 100Hz flicker.   
   >>>>   
   >>>> It was just indicator LEDs to tell me what water circuit was running.  3   
   >>>> zones from the one boiler switched with valves.   
   >>>>   
   >>>>>> Perhaps this bulged cap is why I'm getting flicker, I'll try replacing   
   >>>>>> it   
   >>>>>> tomorrow.   
   >>>>>   
   >>>>>>> they very likely do because those are the only cheap   
   >>>>>>> droppers for dropping such a large voltage.   
   >>>>>   
   >>>>>> Aren't miniature SMPS units pretty cheap?   
   >>>>>   
   >>>>> Not as cheap as the cap and the bridge rectifier.   
   >>>>>   
   >>>>>> I just bought a 12V 6A SMPS for £4.50.   
   >>>>>   
   >>>>> Yeah, I did too.   
   >>>>>   
   >>>>>> Designed for powering LEDs   
   >>>>>   
   >>>>> Mine will run anything 12V. I currently use it to power a water pump.   
   >>>>   
   >>>> It was sold for LEDs, presumably it will run anything provided I don't   
   >>>> exceed the 6A.   
   >>>>   
   >>>> However I've noticed they scrimp on the caps (or cooling).  Loads of   
   >>>> them   
   >>>> get bulged caps after a while, in particular a 3A PSU I ran 2A of LEDs   
   >>>> 24/7 from, failed in 1 year.  It kept cutting out - I discovered the   
   >>>> bulk   
   >>>> capacitor had dried out.  Same happened (over a longer period) with two   
   >>>> monitor PSUs.   
   >>>>   
   >>>>> The LEDs I use are all Hues and have their   
   >>>>> own power supply with the led strips.   
   >>>>   
   >>>> The one I mentioned above was for an insectocuter, I removed the flours   
   >>>> and ballast and fitted strips of UV LEDs instead.   
   >>>>   
   >>>>>> - but I've looked inside it and it's definitely a switched mode, not a   
   >>>>>> capacitor dropper.  Now this flickery LED lamp I'm looking inside,   
   >>>>>> it's   
   >>>>>> about 20W, so 12V at 2A is all that's required, it could have had an   
   >>>>>> SMPS   
   >>>>>> in it similar to the one I just described.   
   >>>>>   
   >>>>> Yeah, but the cap and bridge are cheaper.   
   >>>>   
   >>>> Well I've got 9W £4 strips with a switched mode PSU in them, so they   
   >>>> can't   
   >>>> cost that much.   
   >>>>   
   >>>>>> I'm now looking inside one of the better LED lamps (the non-flickery   
   >>>>>> model).  It has a basic SMPS inside it.  They're 9W and £4 each for   
   >>>>>> the   
   >>>>>> whole lamp.  I'm sure it's more than just a standard SMPS though,   
   >>>>>> because   
   >>>>>> when some LEDs fail short circuit (it has about 40 in series), the   
   >>>>>> voltage   
   >>>>>> coming from the PSU drops, to maintain the correct current for the   
   >>>>>> remaining good LEDs.   
   >>>>>   
   >>>>> Yeah, its best to drive leds in constant current mode.   
   >>>>   
   >>>> I'm surprised that the LEDs always fail short circuit, new type of LED   
   >>>> designed to do this?   
   >>>   
   >>> Think its just the way leds fail naturally   
   >>> with the higher powered lighting leds.   
   >   
   >> What's quite weird is with the decent strips I've got, the LEDs are wired   
   >> in pairs.  Each pair is in parallel, then there are 20 such pairs in   
   >> series.  When one single LED fails, I'd expect either it shorts and the   
   >> neighbouring one in the pair gets 0 volts, or it fails open circuit and   
   >> the neighbouring one gets double current and soon dies.  But neither   
   >> happens.  The neighbouring LED stays lit at the same brightness.  Any idea   
   >> how this is possible?  Could the dead LED still have the same current   
   >> going through it?   
   >   
   > There are only 3 possibilitys.   
   >   
   > The pairs arent actually wired in parallel, it only   
   > looks like they are.  This is the most likely.   
      
   They're definitely in parallel.  I've tested a broken strip in depth with a   
   meter, and also looked at the circuit tracks.  It's most definitely 2 LEDs in   
   parallel, then 20 of those pairs in series.  70V DC is applied to the whole   
   strip by the PSU.  I've    
   just looked at the voltage across the LEDs in the half busted strip pictured   
   in the link below.  Working pairs are 3.3V across each LED.  Broken pairs are   
   2.6V across each LED.  Pairs with one of the two lit are 3.6V per LED.  No   
   idea what that means!:     
   https://www.dropbox.com/s/eml663rsozgtlzj/Broken%20LED%20strip.jpg?dl=0   
      
   > When a led fails, it just stops emitting light but is still electrically   
   > identical to before it failed. This is the least likely.   
      
   I'd say it was the most likely, although from the above measurement in the   
   photo, it looks like they stuill conduct, but a different amount, and the   
   neighbour in the pair doesn't care that much.  I think the neighbour is more   
   likely to be the next to    
   fail, so it might be getting a bit of a rough time.   
      
   > Or the leds don't actually vary in light they put out   
      
   [continued in next message]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)