Split topic: Pulse/Multiplexed LED current

[JohnW2]

In another thread ( click here ), a spin-off discussion arose about the used of pulsed (maybe also multiplexed) current used to drive LED elements, the most recent contributions to which discussion were ....

5x brighter? You're talking nonsense. I've worked on display systems and the only reason we multiplex is to simplify the drive circuitry, not for any perceived brightness changes. JohnW2 is correct - as soon as you have multiple elements multiplexed there is no perceived brightness change anyway. The effect you describe is only for a single PWM source where the eye is fooled into thinking something is brighter than it actually is. It also only applies to a relatively low (50-78Hz) frequencies which is several times lower than the PWM rate for a typical LED display.

Indeed.

As I implied before, the one thing that presumably is true is that if one has N (not chronologically overlapping) multiplexed channels, each with a mark-space ratio of 1:N, then one will have constant (not pulsatile) light output, but can use current pulses of a magnitude that the LED elements would not survive if it were a continuous current?

However, I'm far from convinced that such an approach would achieve anything particularly useful. Once one has reached a stage at which current is flowing continuously through LEDs (albeit different LEDs at different points in time), I would have thought that one might just as well use a (probably smaller) number of LEDs which are powered continuously, but not beyond their continuous rating. Not only would that probably increase product life, but (in view of the slightly 'convex' shape of the light/current curve) I imagine that it could be more efficient (in terms of light/current, or light/power), couldn't it?

Since some are unhappy after 'drifted threads', I'd be grateful if any further comments on this topic could be posted in THIS thread. Thanks.

Kind Regards, John

 

[ericmark]

When doing my degree, which was not that long ago, did it after I had finished working, we were given a project to test the pulsing of LED elements to produce more light, we used red LED's and 555 timers and a optical isolator as driver, so rather a simple affair, we had a lux meter to measure the output, it was rather a failure, we go no more output by flashing the LED as to solid drive, however I suspect the Lux meter was the problem, it likely averaged out the output so they did seem brighter to the eye by flashing them, but did not record any increase, in fact recorder results were the reverse.

Moving from tungsten to CFL to LED I noted more changes than expected, one being due to direction of the light, most LED bulbs project no light towards the base, so a fitting where the base is at bottom gives different results to where base is on the top. But I had base at bottom and Philips SES globe 8W CFL which had a very short life, Lidi were doing cheap LED candle bulbs so I decided to give them a try, only 1.8W intention was to use in reading lamp, but yet more CFL failed so as a temporary measure they were fitted into the ceiling lamps.

I was rather surprised, they seemed brighter than the CFL they replaced, although that could be due to colour temperature, so 10 bulbs in the room so when Lidi did LED bulbs again I got 8 x 3W bulbs they only had 8 so two were 1.8W. The result seemed far brighter than the CFL which always did look dim, but when we tried to read, we found not bright enough.

Mother had a smaller living room also using 10 bulbs, so I fitted 10 x 5W from HomeBargains in my house and fitted my old bulbs in her house, so ended up with 50W of lighting using LED, where house as built used 200W of tungsten. As to if my eyes have got worse hard to quantify, but likely they have, but seems in practice LED uses around 1/5th of the power of tungsten allowing for my eyes to have degraded slightly. When using the light to read with, however it seems much brighter simply walking through the room.

Using my camera with 100 ASA film and a f2 lens and a speed of 1/20 second I could take indoor photos with it, however my digital SLR with same lens and set to ISO 800 still needs a shutter speed of 1/20 second. Maybe all those years ago I had 400 ASA or 600 ASA film? Only have the prints so can't be 100% sure, but it seems those two 100W tungsten had a lot more light than the ten LED's. I regular took indoor photos without a flash. And today I rely on the anti shake technology with floating CCD to reduce speed to 1/15 second and 800 ISO (which is same as old ASA) to get same results.

The camera will be an average of the light over 1/15 second, so flashing the LED will not make it brighter for the camera, where maybe the eye can see more due to flashing, but really the only way to judge light to the human eye is to read some thing. The human brain fills in the gaps, so as out eyes degrade the brain uses the knowledge of what is likely there to fill the gaps, macular degeneration sufferers can get road signs mixed up with people, and see people without any heads, so with reading, we may not be seeing every letter of the word, but our brain guesses it.

Also Irlen syndrome can result in colour being very important for some people, so if not true white light, then the results may vary person to person so near impossible to quantify how much light the human can see. And flashing light also causes medical problems.

So in conclusion even if for some people flashing the light using the same power can help, that is not universally true, so really can't claim it is an improvement.

Flash in photography is really odd, my flash gun can produce light for 1/25,000 of a second far shorter of a time that any shutter can work for, and it can freeze movement, picture of shower with a blacked out room and only flash gun is really interesting, however the focal plane shutter although it can be set to 1/2000 second, to expose the whole of the sensor max speed is around 1/180 second, so to use the flash the shutter speed it limited. There are today high speed flashes which in real terms are slower, it sends out a series of pulses over 1/100 second so the camera shutter can be set at a higher speed, but also replacing the mechanical shutter with electronic switches, so this is moving on.

However the pictures taken with LED and discharge lights have produced some odd effects, we can see the effects with a camera, but our brain filters out the effects with our eyes.

 

[JohnW2]

When doing my degree, which was not that long ago, did it after I had finished working, we were given a project to test the pulsing of LED elements to produce more light, we used red LED's and 555 timers and a optical isolator as driver, so rather a simple affair, we had a lux meter to measure the output, it was rather a failure, we go no more output by flashing the LED as to solid drive, however I suspect the Lux meter was the problem, it likely averaged out the output so they did seem brighter to the eye by flashing them, but did not record any increase, in fact recorder results were the reverse.

As you imply, it sounds as if that was a rather silly experiment (IF the expectation was that the meter would show a higher reading with the pulsed light) - the pulsed light appears brighter, despite not having a higher average light intensity, because the human eye does not 'average out' (whereas the meter very probably would).

... unless, of course, the experiment was perhaps designed to illustrate just that - i.e. that the human eye, but noit an 'averaging meter' perceived the pulsed light as brighter?

Kind Regards, John

 

[SimonH2]

Indeed, what you'd need to do would be to setup a light comparison bench so that you could adjust a (static) light to exactly match the perceived brightness of the pulsed light. Then you could measure the actual brightness of the static light to get an accurate proxy for the perceived brightness of the pulsed light.

 

[JohnW2]

Indeed, what you'd need to do would be to setup a light comparison bench so that you could adjust a (static) light to exactly match the perceived brightness of the pulsed light. Then you could measure the actual brightness of the static light to get an accurate proxy for the perceived brightness of the pulsed light.

Indeed so.

Kind Regards, John

 

[bernardgreen]

Use a grease spot photometer to compare a steady current LED with a pulsed current LED with the same average current.

Grease Spot

 

[JohnW2]

Use a grease spot photometer to compare a steady current LED with a pulsed current LED with the same average current.

I almost jumped in and suggested that (happy memories of school!), but then I wondered exactly how it would work - since, again, we would be relying on human visual perception.

Kind Regards, John

 

[SimonH2]

The whole point is to rely on human perception - that's what you're trying to measure. Instead of the candles, one light source is a steady DC driven LED, the other is a pulsed LED. get the system balanced so that they have the same perceived brightness and you then have a measure of the relative brightness AS PERCEIVED BY THE EYE.
And that's what the discussion is about.

As an aside, the version of the experiment I've had described to me was similar, except that instead of the screen with grease spot, there are two matt reflectors, each set at 45˚ to the line between the light sources. You look from the side and see the two matt sheets, each lit from one source. When both are the same brightness (set by moving the screens as an assembly left or right) then you've found a balance point and can use the inverse square law to work out the relative brightness.

 

[JohnW2]

The whole point is to rely on human perception - that's what you're trying to measure. Instead of the candles, one light source is a steady DC driven LED, the other is a pulsed LED. get the system balanced so that they have the same perceived brightness and you then have a measure of the relative brightness AS PERCEIVED BY THE EYE. ... And that's what the discussion is about.

I realise all that, so maybe I was not clear enough when I wrote that a grease-spot photometer "would again be relying on human visual perception" - by which I meant "would again be relying on the nuances of human visual perception".

In other words, I, for one, do not know how the human visual perception process would respond to looking at one side of a grease spot, one side of which was being illuminated by constant light, and the other side of which was being illuminated by pulses of light - and therefore what would be the meaning/interpretation of the situation when the spot was, under those circumstances, perceived to disappear - do you?

You seem to be assuming that the spot would be perceived to disappear in the situation in which the brightnesses of the two LEDs ('steady' and 'pulsed') would, if viewed singly and directly, be the same - (as would be the case if both LEDs were 'steady') - but I don't think that is a foregone conclusion. As I said, I have no idea.

Kind Regards, John

 

[SimonH2]

Ah, I see now.
I also wonder about the grease spot, the alternative of having two matt panels side by side would (to my thinking) be a better test. I can't help wondering what effect the different "texture" of the grease spot will have. But either way, if the perceived brightnesses are the same, then I'd expect the spot to disappear/the two panels to merge - if they don't then the two sources aren't "the same" as far as the observer is concerned.

 

[JohnW2]

Ah, I see now. I also wonder about the grease spot, the alternative of having two matt panels side by side would (to my thinking) be a better test. I can't help wondering what effect the different "texture" of the grease spot will have. But either way, if the perceived brightnesses are the same, then I'd expect the spot to disappear/the two panels to merge - if they don't then the two sources aren't "the same" as far as the observer is concerned.

Maybe. However, as I think you now understand, my point was this ... if one doen't think about it too deeply (which I doubt that any of us usually does), it may seem 'obvious' why the grease spot appears to disappear when 'equally illuminated' on both sides, but the reality might be much more complicated in terms of perceptive physiology - so, as I said, I wouldn't personally want to speculate about what the situation might be if the nature of the light (e.g. constant vs pulsed) was different on the two sides of the spot.

Kind Regards, John