Post by thread:Re: [EE] White LEDs on mains

Jinx wrote: > A few weeks ago someone mentioned seeing white LEDs powered > by mains. I'd seen it too but couldn't remember where > > Found this in one of my pdf stashes when looking for something else > > http://home.clear.net.nz/pages/joecolquitt/acwhiteleds.pdf > > 11kB > > Why have the 'bridges' between the parallel strings of LED's? I'm talking about how the nodes between the LED's that conduct on one cycle of AC are connected to the nodes that conduct on the other cycle... What purpose do they serve. I initially thought it could serve as a mechanism to keep the system running in the event of a 'fail open' LED, but no, that would not help... Rolf

> > Why have the 'bridges' between the parallel strings of LED's? > > Don't know. The LEDs won't be conducting half the time. It > would help with assembly a bit I guess, as you'd solder two > LEDs in anti- parallel and then join them to the next pair. Maybe ? Has any ever made a bridge out of LEDs? It rectifies and lights up! I've seen LEDs run off mains simply by using a resistor. The last one I saw was in one of those cheap main filter things. Not the best for a lamp, but it seemed to work ok as an indicator, if the layer of dust was any guide to its age. Usually it's a physically large resistor, either because dropping 240v produces a few watts or to give a large gap for isolation in case something goes wrong (sure!), but sometimes 2 or 3 small wattage ones in series. The series method is for the really cheap stuff. Tony

Jinx wrote: > A few weeks ago someone mentioned seeing white LEDs powered > by mains. I'd seen it too but couldn't remember where > > Found this in one of my pdf stashes when looking for something else > > http://home.clear.net.nz/pages/joecolquitt/acwhiteleds.pdf > > 11kB > > RE the PDF design: Why use back-to-back LEDs? One can simply use a full wave rectifier and then use only one direction of LEDs. Second, since each LED has a significant voltage drop, the voltage drop of the string could be quite large. This reduces the drop across the resistor. A 1K resistor can't be optimal for a single LED *and* a string of 12? I've built some "fireflies in a bottle" decorations. These operate on 120V using a full wave rectifier, fuse and ~500ohm resistor. The resulting rectified DC current feeds a series of 55 yellow LED's. These have proved to be very durable. ...Alan

When running LEDs from the mains, I would be wary of the LED's peak-reverse-voltage rating (typically only 5 or 10V). I *think* that if you exceed that you simply cause avalanche breakdown (which, as long as the resistor is there to limit current is probably OK) but technically you are greatly exceeding the ratings of the LED. You should use a silicon diode (like 1N4004) in series with the LED to prevent it from seeing a couple hundred volts across it backwards. Also, I'd want to use more than one current limiting resistor in series. That way, in the unlikely event that one fails shorted, at least the others are there to limit the current. Sean On Sun, Aug 24, 2008 at 11:44 AM, Tony Smith <spam_OUTajsmithTakeThisOuTbeagle.com.au> wrote: {Quote hidden}>> > Why have the 'bridges' between the parallel strings of LED's? >> >> Don't know. The LEDs won't be conducting half the time. It >> would help with assembly a bit I guess, as you'd solder two >> LEDs in anti- parallel and then join them to the next pair. Maybe ? > > > Has any ever made a bridge out of LEDs? It rectifies and lights up! > > I've seen LEDs run off mains simply by using a resistor. The last one I saw > was in one of those cheap main filter things. Not the best for a lamp, but > it seemed to work ok as an indicator, if the layer of dust was any guide to > its age. > > Usually it's a physically large resistor, either because dropping 240v > produces a few watts or to give a large gap for isolation in case something > goes wrong (sure!), but sometimes 2 or 3 small wattage ones in series. The > series method is for the really cheap stuff. > > Tony > > -

On Sun, 24 Aug 2008 14:53:25 -0400, you wrote: >When running LEDs from the mains, I would be wary of the LED's >peak-reverse-voltage rating (typically only 5 or 10V). I *think* that >if you exceed that you simply cause avalanche breakdown (which, as >long as the resistor is there to limit current is probably OK) but >technically you are greatly exceeding the ratings of the LED. You >should use a silicon diode (like 1N4004) in series with the LED to >prevent it from seeing a couple hundred volts across it backwards. > >Also, I'd want to use more than one current limiting resistor in >series. That way, in the unlikely event that one fails shorted, at >least the others are there to limit the current. > Resistors don't fail short - even the standards for Intrinsically safe equipment for use in flammable atmospheres state that resistors can be assumed to be infallible as regards failure to lower values. however using multiples in series helps spread heat load. {Quote hidden}> > >On Sun, Aug 24, 2008 at 11:44 AM, Tony Smith <.....ajsmithKILLspam@spam@beagle.com.au> wrote: >>> > Why have the 'bridges' between the parallel strings of LED's? >>> >>> Don't know. The LEDs won't be conducting half the time. It >>> would help with assembly a bit I guess, as you'd solder two >>> LEDs in anti- parallel and then join them to the next pair. Maybe ? >> >> >> Has any ever made a bridge out of LEDs? It rectifies and lights up! >> >> I've seen LEDs run off mains simply by using a resistor. The last one I saw >> was in one of those cheap main filter things. Not the best for a lamp, but >> it seemed to work ok as an indicator, if the layer of dust was any guide to >> its age. >> >> Usually it's a physically large resistor, either because dropping 240v >> produces a few watts or to give a large gap for isolation in case something >> goes wrong (sure!), but sometimes 2 or 3 small wattage ones in series. The >> series method is for the really cheap stuff. >> >> Tony >> >> --

Sean Breheny wrote: > You > should use a silicon diode (like 1N4004) in series with the LED to > prevent it from seeing a couple hundred volts across it backwards. That would work only if the leakage of the LED is (much) higher than that of the diode. To avoid that (potential?) problem I would put the diode across the LED, not in series. -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

> When running LEDs from the mains, I would be wary of the LED's > peak-reverse-voltage rating (typically only 5 or 10V) One of my earliest discoveries about LEDs was a little circuit for running them on 230V. 0.47uF and 1k in series, with a 1N4148 anti-parallel across the LED. I used a yellow LED, and it's still there after mumble-mumble-mumble years, as an illuminator for the counter on a reel-to-reel tape deck

Hi Mike, Well, I can think of some design mistakes which could lead to a resistor failing with a lower resistance than nominal. For example, some small resistors have a maximum working voltage lower than what you would expect based on power dissipation (200 to 250V for 1/4W thru hole carbon film resistors). If you exceed this, I would imagine that you could get arcing across part or all of the resistor. Also, if the resistor overheats and burns material around it, any carbon-containing material around it might be able to conduct once it burns. Yes, those are a bit of a stretch, so I suppose that you are correct. Sean On Sun, Aug 24, 2008 at 3:30 PM, Mike Harrison <mikeKILLspamwhitewing.co.uk> wrote: {Quote hidden}> On Sun, 24 Aug 2008 14:53:25 -0400, you wrote: >>Also, I'd want to use more than one current limiting resistor in >>series. That way, in the unlikely event that one fails shorted, at >>least the others are there to limit the current. >> > > Resistors don't fail short - even the standards for Intrinsically safe equipment for use in > flammable atmospheres state that resistors can be assumed to be infallible as regards failure to > lower values. > > however using multiples in series helps spread heat load. > >

Hi Wouter, I thought about that but decided that putting the diode anti-parallel to the LED would double the power dissipation of the current limiting resistor. Also, the series diode would limit the inverse current through the LED to (at maximum) the reverse leakage current of the diode. I think it would be very difficult to damage the LED with such a tiny current. I suppose what you say would be safer and if resistor dissipation were an issue, TWO silicon diodes could be used, one anti-parallel to the LED and one in series with the LED/diode combination. This would guarantee that the reverse voltage on the LED would be never more than 0.6V but there would still be negligible current during the "off" part of the AC cycle. Sean On Sun, Aug 24, 2008 at 5:11 PM, Wouter van Ooijen <.....wouterKILLspam.....voti.nl> wrote: {Quote hidden}> Sean Breheny wrote: >> You >> should use a silicon diode (like 1N4004) in series with the LED to >> prevent it from seeing a couple hundred volts across it backwards. > > That would work only if the leakage of the LED is (much) higher than > that of the diode. To avoid that (potential?) problem I would put the > diode across the LED, not in series. > > -- > > Wouter van Ooijen > > -- ------------------------------------------- > Van Ooijen Technische Informatica: http://www.voti.nl > consultancy, development, PICmicro products > docent Hogeschool van Utrecht: http://www.voti.nl/hvu > > -

On Mon, Aug 25, 2008 at 3:30 AM, Mike Harrison <EraseMEmikespam_OUTTakeThisOuTwhitewing.co.uk> wrote: > Resistors don't fail short - even the standards for Intrinsically safe > equipment for use in flammable atmospheres state that resistors > can be assumed to be infallible as regards failure to > lower values. Not all resistors are equal. For IS (intrinsic safety) application, what you say can only apply to certain type of resistors (eg: metal film and wire-wound). Xiaofan

> On Mon, Aug 25, 2008 at 3:30 AM, Mike Harrison <mikespam_OUTwhitewing.co.uk> > wrote: > >> Resistors don't fail short - even the standards for Intrinsically safe >> equipment for use in flammable atmospheres state that resistors >> can be assumed to be infallible as regards failure to >> lower values. > > Not all resistors are equal. For IS (intrinsic safety) application, what > you > say can only apply to certain type of resistors (eg: metal film and > wire-wound). > > Xiaofan I believe carbon composition resistors fail to a short. Film and wire wound fail to an open. When I started in electronics, carbon composition resistors were the only ones around. They had a cylindrical body. I remember seeing my first carbon film resistor with the now familiar "dog bone" shape and thought they were quite strange. I haven't been following this thread closely, but have seen stuff about protecting the LEDs from reverse voltage. As I recall from the original schematic on the thread, short strings of LEDs were connected in parallel with polarity reversed. That handles the reverse polarity protection and also allows AC current to flow so the current limit capacitor does not just charge up to the peak voltage and have everything stop. Harold -- FCC Rules Updated Daily at http://www.hallikainen.com - Advertising opportunities available!