On 2023-01-20 17:34, bitrex wrote:> On 1/20/2023 11:00 AM, John Larkin wrote: >> >> It would be fun to connect a photodetector to an amp and headphones >> and listen to the light in our world, indoors and out. I might do >> that. >> >> I did that that with a coil and mag fields when I was a kid. It would >> be different now. >> > > It'll maybe need to be an amp with a logarithmic gain if you're going to use it both indoors and out, the dynamic range between lighting in an average office and out doors at noontime on a sunny day is like 90 dB.Just use a photo diode, without parallel resistor, and AC-couple it to an amplifier. The logarithmic response is provided by the forward biased diode itself. E.g. 10% modulation, at whatever light intensity, will provide the same AC voltage. Arie
LED Bulb Efficiency vs. Operating Life
Started by ●December 26, 2022
Reply by ●January 20, 20232023-01-20
Reply by ●January 20, 20232023-01-20
On Fri, 20 Jan 2023 11:34:46 -0500, bitrex <user@example.net> wrote:>On 1/20/2023 11:00 AM, John Larkin wrote: >> On Fri, 20 Jan 2023 09:03:29 +0000, Martin Brown >> <'''newspam'''@nonad.co.uk> wrote: >> >>> On 20/01/2023 00:58, Ricky wrote: >>>> On Thursday, January 19, 2023 at 5:31:57 PM UTC-5, bitrex wrote: >>>>> On 1/19/2023 5:20 PM, Ricky wrote: >>>>>> On Thursday, January 19, 2023 at 4:45:52 PM UTC-5, bitrex wrote: >>>>>>> On 12/27/2022 11:39 AM, John Larkin wrote: >>>>>>>> On Tue, 27 Dec 2022 15:47:22 +0000, Martin Brown >>>>>>>> <'''newspam'''@nonad.co.uk> wrote: >>>>>>>> >>>>>>>>> On 26/12/2022 17:16, Ricky wrote: >>>>>>>>>> I know it costs more to make a more efficient LED bulb >>>>>>>>>> for room lighting. Is there also a tradeoff between >>>>>>>>>> efficiency and operating life? I can't think of a >>>>>>>>>> mechanism, but I'm not so familiar with LED light bulb >>>>>>>>>> design. >>>>>>>>> >>>>>>>>> If you push the blue light flux and/or residual heating >>>>>>>>> high enough it can damage/darken the phosphor used to >>>>>>>>> generate the yellow light. >>>>>>>>> >>>>>>>>> The surface brightness of recent LEDs is now about the same >>>>>>>>> per unit area as the sun - which makes it a bad idea to >>>>>>>>> look directly at them. Blue, violet or UV ones especially. >>>>>>>> >>>>>>>> I have a 12-volt, roughly 1 cm square, array that looks like >>>>>>>> a welding torch. It's the kind used in street lamps. >>>>>>>> >>>>>>>> Why don't they diffuse LED street lamps? They are annoying. >>>>>>>> >>>>>>> Especially when they start blinking >>>>>> >>>>>> I can't stand the fast blinking LED tail lights on cars. >>>>>> Everytime I move my line of view, they create a dozen spots of >>>>>> light. When there are more than one it gets insane looking. Seems >>>>>> like most people don't even see this. The first time it happened >>>>>> to me I was trying to merge where there was no merge lane and the >>>>>> ramp was coming from an angle, rather than merging while driving >>>>>> parallel, so the rear view mirror didn't show anything useful. A >>>>>> quick look over my shoulder showed one car passing me, just as I >>>>>> needed to either go or stop. As I turned my head back, the tail >>>>>> lights (those tall Cadillac tail lights) suddenly blossomed into >>>>>> a dozen pairs of lights and I thought it was a bunch of cars! I >>>>>> had to hit my brakes to avoid an accident, only to find there was >>>>>> only the one car. Insane that they would create this sort of >>>>>> hazard, even if everyone doesn't see it. It's the sort of thing >>>>>> that would be changed on an airliner, after the first accident it >>>>>> causes. >>>>>> >>>>>> I've never found out how rapid the blinking is. >>>>>> >>>>> I usually see those flashing brake light mods on street racer-type >>>>> cars around Providence RI, kids put them on their old Eclipse etc. >>>>> to make them look cool. I don't know that they're much improvement >>>>> over a regular high-level lamp at avoiding being rear-ended, >>>>> though. >>>> >>>> We aren't talking about the same thing. I'm talking about tail/brake >>>> lights where the brightness is adjusted by PWM. It saves a few cents >>>> by leaving off the inductor too smooth the waveform into a level. >>> >>> The problem is that they PWM pulse them at a frequency that some people >>> can see in their peripheral vision (which is much more flicker >>> sensitive). I guess the engineers who designed it didn't think about >>> their choice of frequency too hard. Anything above about 300HZ frequency >>> would look pretty much continuous but there seem to be several car >>> makers (and street furniture makers that use ~100Hz at a guess). >>> >>> People who see this as a problem are the same ones who don't get on with >>> walls full of TVs or large screen monitors at 60Hz refresh rates. >> >> It would be fun to connect a photodetector to an amp and headphones >> and listen to the light in our world, indoors and out. I might do >> that. >> >> I did that that with a coil and mag fields when I was a kid. It would >> be different now. >> > >It'll maybe need to be an amp with a logarithmic gain if you're going to >use it both indoors and out, the dynamic range between lighting in an >average office and out doors at noontime on a sunny day is like 90 dB.We's only want to hear the AC component of modulated lights. A bit of optics, even a cardboard tube, would allow aiming and help reject ambient DC light. Gain is cheap.
Reply by ●January 20, 20232023-01-20
On Thu, 19 Jan 2023 16:11:35 +0100, albert@cherry.(none) (albert) wrote:>In article <nnd$4d9a769b$61464ae6@e846562bf142425e>, >Sjouke Burry <burrynulnulfour@ppllaanneett.nnll> wrote: >>On 26.12.22 18:16, Ricky wrote: >>> I know it costs more to make a more efficient LED bulb for room >>lighting. Is there also a tradeoff between efficiency and operating >>life? I can't think of a mechanism, but I'm not so familiar with LED >>light bulb design. >>> >>Lots of light, high chip temp, shorter chip life. > >This can be offset by operating the leds vastly below capacity. >E.g. I have a flash light costing 3.5 euro with 37 leds. >I estimate that the leds have an effectively infinite life span. >Unlike light bulbs you can diminish the current and have approximately >proportional light. > >Groetjes AlbertThe problem with 'white' LEDs is the degradation of the phosphor with high currents. Running the LED at Imax and the light output will drop significantly after a few hundred or a thousand hours. Running at Imax/2 or even Imax/3 and the light output may be strong after claimed 30000 hours. The efficiency (in lm/W) is also better for the lower current. Reputable LED manufacturers specify the light characteristics well below maximum allowed Imax current, often at Imax/2 or Imax/3 so with IMax=1 A, the characteristics are specified at 350 mA. Of course, this requires two or three times the number of LEDs to get the same light output and hence the lamp is more expensive, but this extends the usable life time with more than 3 times, thus being more economical in the long run.
Reply by ●January 20, 20232023-01-20
upsidedown@downunder.com wrote:> On Thu, 19 Jan 2023 16:11:35 +0100, albert@cherry.(none) (albert) > wrote: > >> In article <nnd$4d9a769b$61464ae6@e846562bf142425e>, >> Sjouke Burry <burrynulnulfour@ppllaanneett.nnll> wrote: >>> On 26.12.22 18:16, Ricky wrote: >>>> I know it costs more to make a more efficient LED bulb for room >>> lighting. Is there also a tradeoff between efficiency and operating >>> life? I can't think of a mechanism, but I'm not so familiar with LED >>> light bulb design. >>>> >>> Lots of light, high chip temp, shorter chip life. >> >> This can be offset by operating the leds vastly below capacity. >> E.g. I have a flash light costing 3.5 euro with 37 leds. >> I estimate that the leds have an effectively infinite life span. >> Unlike light bulbs you can diminish the current and have approximately >> proportional light. >> >> Groetjes Albert > > > The problem with 'white' LEDs is the degradation of the phosphor with > high currents. Running the LED at Imax and the light output will drop > significantly after a few hundred or a thousand hours. Running at > Imax/2 or even Imax/3 and the light output may be strong after claimed > 30000 hours. The efficiency (in lm/W) is also better for the lower > current. > > Reputable LED manufacturers specify the light characteristics well > below maximum allowed Imax current, often at Imax/2 or Imax/3 so with > IMax=1 A, the characteristics are specified at 350 mA. > > Of course, this requires two or three times the number of LEDs to get > the same light output and hence the lamp is more expensive, but this > extends the usable life time with more than 3 times, thus being more > economical in the long run.If the phosphor were responsible, you'd expect the light output to get bluer and bluer as the lamp aged, which I don't think it does. The phosphor/fluor is inorganic, so it doesn't degrade the way organic dyes do. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Reply by ●January 20, 20232023-01-20
On 20/01/2023 17:09, Ricky wrote:> On Friday, January 20, 2023 at 11:57:20 AM UTC-5, Clive Arthur wrote: >> On 20/01/2023 16:00, John Larkin wrote: >> >> <snip> >>> >>> It would be fun to connect a photodetector to an amp and headphones >>> and listen to the light in our world, indoors and out. I might do >>> that. >> The visible spectrum covers about an octave, so translate the light >> frequencies into audio frequencies. >> >> (Not sure how you'd do that.) > > Don't think it could be a direct conversion, but using three photodectors like the eye does, you can get an idea of the color and select a frequency to play based on that. > > I knew someone with a cochlear implant. She knew a bit about how it worked. I think she said it had only 16 electrodes, yet she could hear music well enough to enjoy it. That impressed me. Imagine how different it could be with 256 electrodes. Music is such a rich medium!RGB detectors could work for a single tone, but I was hoping for polyphony. Simply divide the light frequencies by 2^39 or 2^40. Jan could probably do it with a PIC. I just remembered - my great niece has a toy Xylophone with a single octave and the notes are coloured like a rainbow. -- Cheers Clive -- Cheers Clive
Reply by ●January 20, 20232023-01-20
On Friday, January 20, 2023 at 5:50:39 PM UTC-5, Clive Arthur wrote:> On 20/01/2023 17:09, Ricky wrote: > > On Friday, January 20, 2023 at 11:57:20 AM UTC-5, Clive Arthur wrote: > >> On 20/01/2023 16:00, John Larkin wrote: > >> > >> <snip> > >>> > >>> It would be fun to connect a photodetector to an amp and headphones > >>> and listen to the light in our world, indoors and out. I might do > >>> that. > >> The visible spectrum covers about an octave, so translate the light > >> frequencies into audio frequencies. > >> > >> (Not sure how you'd do that.) > > > > Don't think it could be a direct conversion, but using three photodectors like the eye does, you can get an idea of the color and select a frequency to play based on that. > > > > I knew someone with a cochlear implant. She knew a bit about how it worked. I think she said it had only 16 electrodes, yet she could hear music well enough to enjoy it. That impressed me. Imagine how different it could be with 256 electrodes. Music is such a rich medium! > RGB detectors could work for a single tone, but I was hoping for > polyphony. Simply divide the light frequencies by 2^39 or 2^40. Jan > could probably do it with a PIC.Yeah, they have a special instruction on the PIC for that, but it takes 2^39 cycles. Probably added just for Jan, knowing how many of the devices he uses.> I just remembered - my great niece has a toy Xylophone with a single > octave and the notes are coloured like a rainbow.LOL, maybe it will sound from the light beams. But isn't each note on a xylophone binary, 1/0? I would think you'd want an analog device. If you really did that, (not impossible with three sensors, the human eye does it) it would probably sound a lot like noise, unless you had a way to present the sounds in at least 2D. An eye with no spacial distinction wouldn't be very useful, to anyone smarter than planaria. What sort of processing can turn three receptors, RGB, into a spectrum of color? It might take more than a PIC, just sayin... -- Rick C. --- Get 1,000 miles of free Supercharging --- Tesla referral code - https://ts.la/richard11209
Reply by ●January 20, 20232023-01-20
On Friday, January 20, 2023 at 3:50:25 PM UTC-8, Ricky wrote:> On Friday, January 20, 2023 at 5:50:39 PM UTC-5, Clive Arthur wrote:> > I just remembered - my great niece has a toy Xylophone with a single > > octave and the notes are coloured like a rainbow. > LOL, maybe it will sound from the light beams. But isn't each note on a xylophone binary, 1/0? I would think you'd want an analog device. > > If you really did that, (not impossible with three sensors, the human eye does it) it would probably sound a lot like noise, unless you had a way to present the sounds in at least 2D. An eye with no spacial distinction wouldn't be very useful, to anyone smarter than planaria. > > What sort of processing can turn three receptors, RGB, into a spectrum of color? It might take more than a PIC, just sayin...An octave or two output is eight to 24 distinct frequencies, not just three. There are linear-array sensors that make a spectrum breakdown for visible light easy with prism or grating, and worst-case the translation to sound from light would just be a 24x24 matrix operating on a 24-element light sensor 'vector'. A DSP processor could handle it fine; slight bonus for using multiple high-Q filters on the 24-element output-with-modulation.
Reply by ●January 20, 20232023-01-20
On Friday, January 20, 2023 at 7:25:26 PM UTC-5, whit3rd wrote:> On Friday, January 20, 2023 at 3:50:25 PM UTC-8, Ricky wrote: > > On Friday, January 20, 2023 at 5:50:39 PM UTC-5, Clive Arthur wrote: > > > > I just remembered - my great niece has a toy Xylophone with a single > > > octave and the notes are coloured like a rainbow. > > LOL, maybe it will sound from the light beams. But isn't each note on a xylophone binary, 1/0? I would think you'd want an analog device. > > > > If you really did that, (not impossible with three sensors, the human eye does it) it would probably sound a lot like noise, unless you had a way to present the sounds in at least 2D. An eye with no spacial distinction wouldn't be very useful, to anyone smarter than planaria. > > > > What sort of processing can turn three receptors, RGB, into a spectrum of color? It might take more than a PIC, just sayin... > An octave or two output is eight to 24 distinct frequencies, not just three. There are linear-array > sensors that make a spectrum breakdown for visible light easy with prism or grating, > and worst-case the translation to sound from light would just be a 24x24 matrix operating > on a 24-element light sensor 'vector'. A DSP processor could handle it fine; slight bonus for > using multiple high-Q filters on the 24-element output-with-modulation.I'm talking about sensing the color of light. The human eye uses three wavelength specific sensors to respond to various spectra of light frequencies. I don't know why you talk about 24 distinct frequencies. The human ear can distinguish much finer resolution in frequency than that! If you used an array of wavelength specific light sensors, there would be no need for a matrix. Each one would correspond to a wavelength of sound. That would only require a single output channel and it would not require a DSP to handle the processing. An Arduino could do it. -- Rick C. --+ Get 1,000 miles of free Supercharging --+ Tesla referral code - https://ts.la/richard11209
Reply by ●January 21, 20232023-01-21
On Friday, January 20, 2023 at 6:17:48 PM UTC-8, Ricky wrote:> On Friday, January 20, 2023 at 7:25:26 PM UTC-5, whit3rd wrote: > > On Friday, January 20, 2023 at 3:50:25 PM UTC-8, Ricky wrote: > > > On Friday, January 20, 2023 at 5:50:39 PM UTC-5, Clive Arthur wrote: > > > > > > I just remembered - my great niece has a toy Xylophone with a single > > > > octave and the notes are coloured like a rainbow.> > An octave or two output is eight to 24 distinct frequencies, not just three. There are linear-array > > sensors that make a spectrum breakdown for visible light easy with prism or grating, > > and worst-case the translation to sound from light would just be a 24x24 matrix operating > > on a 24-element light sensor 'vector'. A DSP processor could handle it fine; slight bonus for > > using multiple high-Q filters on the 24-element output-with-modulation.> I'm talking about sensing the color of light. The human eye uses three wavelength specific sensors to respond to various spectra of light frequencies.But, mapping that to a sonic output, there's a polyphonic opportunty. Consider the light is actually a complete spectrum (to be mapped to a multiplicity of notes, to make a 'chord').> I don't know why you talk about 24 distinct frequencies. The human ear can distinguish much finer resolution in frequency than that!The idea was to use a non-high-fidelity sonic range, like the two-or-three octave range of telephony, and the diatonic scale, 8 full notes spanning an octave... Visible light (400 to 700 nm) is less than an octave, some modest frequency-gain in the range seemed appropriate.> If you used an array of wavelength specific light sensors, there would be no need for a matrix. Each one would correspond to a wavelength of sound. That would only require a single output channel and it would not require a DSP to handle the processing. An Arduino could do it.Or, you could pump energy into a tuning fork array (similar to a music box) and no computation is necessary at all, just a power-of-light dependence for multiple oscillators' amplitudes.
Reply by ●January 21, 20232023-01-21
On Saturday, January 21, 2023 at 1:28:07 AM UTC-5, whit3rd wrote:> On Friday, January 20, 2023 at 6:17:48 PM UTC-8, Ricky wrote: > > On Friday, January 20, 2023 at 7:25:26 PM UTC-5, whit3rd wrote: > > > On Friday, January 20, 2023 at 3:50:25 PM UTC-8, Ricky wrote: > > > > On Friday, January 20, 2023 at 5:50:39 PM UTC-5, Clive Arthur wrote: > > > > > > > > I just remembered - my great niece has a toy Xylophone with a single > > > > > octave and the notes are coloured like a rainbow. > > > An octave or two output is eight to 24 distinct frequencies, not just three. There are linear-array > > > sensors that make a spectrum breakdown for visible light easy with prism or grating, > > > and worst-case the translation to sound from light would just be a 24x24 matrix operating > > > on a 24-element light sensor 'vector'. A DSP processor could handle it fine; slight bonus for > > > using multiple high-Q filters on the 24-element output-with-modulation. > > > I'm talking about sensing the color of light. The human eye uses three wavelength specific sensors to respond to various spectra of light frequencies. > But, mapping that to a sonic output, there's a polyphonic opportunty. Consider the light is actually a complete > spectrum (to be mapped to a multiplicity of notes, to make a 'chord').Yes, and color is not a 1D phenomenon.> > I don't know why you talk about 24 distinct frequencies. The human ear can distinguish much finer resolution in frequency than that! > The idea was to use a non-high-fidelity sonic range, like the two-or-three octave range of telephony, > and the diatonic scale, 8 full notes spanning an octave...That's my point. There's a lot more than 8 notes in a music octave. The diatonic scale has 12 notes and musicians often bend those notes. Other scales use even more notes.> Visible light (400 to 700 nm) is less than an octave, some modest frequency-gain in the > range seemed appropriate. > > If you used an array of wavelength specific light sensors, there would be no need for a matrix. Each one would correspond to a wavelength of sound. That would only require a single output channel and it would not require a DSP to handle the processing. An Arduino could do it. > Or, you could pump energy into a tuning fork array (similar to a music box) and > no computation is necessary at all, just a power-of-light dependence for multiple oscillators' > amplitudes.Power of light? You've lost me. -- Rick C. -+- Get 1,000 miles of free Supercharging -+- Tesla referral code - https://ts.la/richard11209
