Safe Speed Forums

Not so much as Improving Road Safety? .........

Strobed LED light designs:

> Does anyone else share this annoyance? New LED based traffic lights and rear light clusters on cars both now using LEDs.
>
> I can see the advantage in that, if a filament light blows then all is out,
> but why do they insist in having the lights strobed? Reason, to save
> power and extend LED life.
>
> OK fair enough, but my problem is that I can see the flickering, especially when I turn my head across the lights ('tis distracting at least - and may even be dangerous to certain people potentially affected by strobed lights?) .

Is there a regulation
which stipulates the minimum refresh frequency eg 100Hz(flashes per
second) and is it just me? eg. I seem to be the only person in my office
who notices PC monitors strobing below a setting of about 85Hz!

I thought that the general rule for vehicle lighting was 'No flashing lights'?

It may be OK as at present when there are few rear lights strobed, but
imagine approaching a junction and turning ones head to check for traffic
when there are say 20 cars in sight, all strobing and more importantly
(and inevitably) all out of synch!

I haven't yet been able to locate the design regs for car design on
this, any idea where one might find them?


Also, what are the regs. on 'what %age of LEDs can fail' before a light is considered 'out'?

Thanks for looking.

I think the theory is that the LED is REALLY brighter when strobed, regardless of the ower saving aspect.

_________________
Time to take responsibility for our actions.. and don't be afraid of speaking out!

Flickering LED lamps annoy the hell out of me too.

I notice that some modern London buses strobe their rear lights for side lights and illuminate them continuously for brake lights.

Most car makes with LED rear light clusters seem to run them continuously (no strobing) but there are exceptions.

Something in my visual system is also unhappy about the very fast rising edge of LED lamps coming on. I perceive a 'bounce' after full illumination. I don't think it's real, but the fast rising edge seems to set something 'ringing'.

['Ringing' is an electronics term frequently applied to unwanted oscillation.]

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

Fast switch on of LED lights. Yes, another point well made. I had forgotten to include that in my post.

It would be interesting to hear about any research on this from the point of view of driver perception. ie Is it just a case of getting used to it or is there some perception in the brain which causes one to react differently.

TRRL may/not? have done something on this perhaps.

Not too long ago I designed an LED alarm/alert beacon for the deaf (work). Within this I researched the efficiency/noticability of pulsed LEDs. At frequencies above that where you notice flicker there is no gain in intensity or noticability (for a given average current) regardless of duty cycle and pulse rate (static eyeball). At frequencies below ~30hz, it becomes very annoying/attention grabbing and seemingly brighter the lower the frequency. However, people who suffer from epilepsy are at risk from seizure at frequencies where flicker is noticeable (not good for our application), hence we decided upon a constant intensity output.

This is not the same as movement of the eye leaving persistence steaks on the retina which is noticeable even at very high pulse rates (up to 10khz), the vehicle rear LED clusters pulse at ~100hz (my estimate); you can’t see them flicker if you fix your gaze upon them. I think these are pulsed because of the way the low cost PWM controller simply reduces the light output (the duty cycle rises to 100% - no pulsing - when the brake pedal is pushed).

What you might sometimes experience is a beating effect. If you have one set of LEDs flashing at 100Hz, and another at 110Hz, you may notice a 10Hz beat going on which could be distracting. While the LEDs in isolation may not appear to be flashing, when they are close together and reflecting off the same surfaces you may start to see a beat effect.

While the higher the frequency the less distracting it is on the eye, there are other downsides such as higher switching losses in the electronics and increased RF emissions.

I couldn’t understand how that would be possible given (from my experience) that the eye will seen an average intensity above a certain frequency, so I tried it

One sig-gen at 100hz, the other at 105hz, each independently driving an LED (equal wavelength and intensity) with a square wave – no beat noticed (I was careful to reflect both off a diffuse surface, other light sources were extinguished).
I confirmed the setup by rearranging both sig-gens in series (modulation) – I could see the 5hz beat.

I've not tried it, but I think if the LEDs are fast enough to respond (ie visible in their own right and not illuminating a long decay phosphor), you WILL see the Doppler beat. Were you using normal red or amber LEDs, or were these the IR-based white ones with phosphor, Smeggy? Also what were your mark/space ratios approx?

LED intensity risetime is 1us, sig-gen output risetime is much quicker
Wavelength 650nm (red), no phosphor conversion
Duty cycle exactly 50%
Frequencies confirmed with frequency meter

The result makes sense to me. Given that the eye is taking the average intensity (if the pulse rate is above a certain frequency), it wont percieve a difference between 1 unit of intensity constantly on and 2 units of intensity on for half the time, hence no beat can be noticed. I could try again with a massively reduced duty cycle but I can’t see how that could make any difference.

I would like to know your reasoning for a perceptible (Doppler?) beat.

I don't think you're right on this. Remember persistance of vision? The perceived brightness is far above zero during the off part of the cycle.

As far as I remember (with a suitable switching frequency) the differece in brightness between 100% duty cycle and 10% duty cycle was quite small (hence the large battery life gains available from switching).

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

A thought -Smeggy( not wishing to get into the egg cracking,so apologies if out of turn) - sig gen in series - were both outputs equal - ( thinking about fc + fm & fc - fm as modulation sums and differences) - one about 205Hz,the other 5Hz, and of course all the other ones at differing levels,depending on the sum/difference chosen.( 305, 310, 10 etc), because you have after all fed AC into rectifiers, with possible non linearities( used to be a problem in RF/HF circuits, when valve aging took effect)

Just another possible angle on the result.

I know about that, that’s also what I thought before my testing.
The persistence effect is very slow hence it is very significant below a certain frequency, but not above. More importantly it’s linear, meaning you get equal benefit with constant light.

The latest LED traffic lights have a single LED per aspect. There are some on the way into Poole if you come down the Holes Bay Road, junction with Sterte Road.

I haven't noticed any strobing, except in one bicycle headlight. This can strobe the lights in order to save battery power. There's little point doing it to extend the life of the LED as they are supposed to last something like 50 years aren't they?

_________________
Andrew.

My info is not as much as that for traffic type - 25 years. But wheras lamps give an almost constant output, LED types do drop off over the years.

Well this thread has explained something to me! Living, as I do, in the last century, I had no idea that these LED clusters were strobed. The main thing that I noticed about them was that they are very difficult for me not to look at. If I am near one, I constantly feel drawn to stare at it, making other road observation very difficult.
Human visual accuity is based on motion and/or change of light intensity monitored via a group of rods and cones in the centre of the retina, near the fovea, called the Bundle of Loos (I think. It could be the Bundle of Fun for all I know - don't ask me, I'm a bloody physicist; I can think of nothing less entertaining than poking about in some poor dead bastard's eyeball. Eurgh!).
Whatever. The brain is conditioned to respond to optical stimulus by aiming this 'sweet' area at the stimulus to find its point of highest magnitude.
If these LED's are batting out some kind of optical Rhumba, that could explain why I can't 'take my eyes off them'.

Paul, as a matter of interest, living as you do with the barbarous Scotia, have you ever seen the Aurora Borealis? If so, did you feel that you could 'hear' a rushing noise that wasn't there? I ask you this as, on the few times that I have seen an Aurora, I have experienced this auditory hypersensitivity (which is not uncommon) and your very apt description of the 'ringing' (or, as we used to call it 'Q') sensation may be explained as having similar visual accuity to me (and about 20% of the population).

_________________
Aquila



Licat volare si super tergum aquila volat...

Claire and I have seen the Aurora a few times. We've never heard a sound associated with it.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

I’ve just had total recall :banghead:

A couple of years back I had a chap round from Future Electronics demonstrating their Luxeon Lumileds for potential use on my beacon project. I asked why the literature recommended using a PWM controller instead of a low level current source for low-level illumination, he stated that these LEDs suffer wavelength shift at low currents. I tried this with my home collection to find he was indeed right! The blues turned a nasty shade of turquoise and the whites turned yellow. I can only estimate that the reds also shift to a longer wavelength at low currents, hence reducing their photometric efficiency (they are already towards the extreme of the long end of the visible spectrum).

It's not so much actually hearing a sound as getting an auditory sensation as if you are expecting a sound - a bit like looking at a UV light in a disco (do they still have them? They were great for detecting the girls that were wearing 'front-loader' bras!), the wavelength is shorter than the diameter of the crystals in the lens and thus cannot be refracted to a focus, so you see an 'aura' around the tube and your accomodatory muscles start to ache as they try and focus the UV onto your 'sweet-spot'.
A fellow astronomer I know, when observing aurorae, says he can smell frying onions and taste blood in his mouth, as well as the auditory response, but his visual accuity is exceptional, being able to see objects in 10X binoculars that I need at least a 30X telescope to see...

_________________
Aquila



Licat volare si super tergum aquila volat...

Sorry I missed this - and I don't think I understand it. What is linear?

I think that visual persistance 'charges' quickly and 'discharges' slowly. It's similar to a circuit that charges a capacitor via a resistance of R and discharges the same capacitor with a resistance of (say) 20R (like a charge pump or a power supply reservoir capacitor).

At very low frequencies, obviously the slow discharge will have time to decay, but above 'medium' frequencies the decay will be negligible because there's no time for any significant 'discharge' before the next charging pulse.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

Very interesting. Fried onions eh?

Sounds like Synesthesia.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety