Author Topic: My questions  (Read 9156 times)
Ash
Member
*****
Offline

View Posts
View Gallery


Re: My questions « Reply #15 on: October 16, 2015, 05:30:39 AM » Author: Ash
With FCO or deep clear cover (not refractor) MH lantern, at the 70 deg side angles most of the light is thrown by significant part of the reflector area, not directly from the arc tube (the arc tube's contribution is closer to the lantern, where it is allready out of scene for the driver looking forward). The reflector does provide the required spread of light

While this does not apply to SOX, i must say modern HID optics come to the point where they spill very little light, no worse than LED lanterns (lens panels haver their own performance  limitations too)

I seen some places where ligting have been converted form FCO/deep clear cover HPS or refractor optics HPS to lens panel LED. The change is allways to the less visibility side due to the "laser in the eye" effect - Without it, it would be better visible what is going on in the dark areas.... The HPS does provide even lighting of the stretch of the road, it is not "light under the lantern" at all
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: My questions « Reply #16 on: October 16, 2015, 01:55:40 PM » Author: Medved
Indeed, most of the light at 70deg comes either from the reflector (in case of FCO), or the refractor lens, in both cases it is spread over wider area. The same is valid with a good LED lens and well sized LED source behind - the lens usually spread the light over significant part of it's cross section (as seen from the 70deg angle). The thing is, the lantern should be made so, from that angle all the ~1cm "circles" from the individual LED's are barely touching each other, so when viewed from the angle, it forms about the same surface area as the HID. Well, that assumes each LED has a lens making practically the same beam pattern as the lantern, so if you take any angle, you always have to see all the individual LED's contributing the same. Then there won't be any more glare than with the HID.

The problem is indeed, by far not all LED lanterns are really made according to such rules. Way too frequently each individual LED has it's optics designed so, it covers just part of the complete beam pattern That mean when you are in one direction, it get served by just very few individual LED's, so the light intensity from these is way higher. That is problem not only for the eventual glare, but as well for the overall reliability: When some of the LED's fail, you just get dark "holes" in the pattern, while with the proper arrangement just the overall intensity would slightly drop, but the pattern shape remains the same.
The reasons may be higher flexibility in creating even custom beam patterns, eventually a possibility to shape the beam electronically (if the ballast allows individual LED brightness control), but it is costing more glare, so can not be used for far throw lanterns (with angles above 55deg or so).

With LED's there is more freedom to make it better, but that means more room to make it worse...
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: My questions « Reply #17 on: October 16, 2015, 03:52:44 PM » Author: Ash
In the LED lights i see the lenses are the same over all LEDs : From any distance or direction, all lenses appear to emit the same light. But in each single lens, the light emission is still only seen from a point behind the lens (allthough magnified), not the entire lens area

If the entire lens would be emitting, we stil have a problem :

 - In the X direction (row of lenses parallel to the road direction) this will only be effective at one set angle - Somewhat farther away the lenses are obscuring each other so its inefficient (ok, we can make this to happen above the 70 deg where no light is intended to be emitted), somewhat closer and they will again be seen as separate lenses... But the lantern is still in the visible scene for the driver much closer than 70 deg

 - In the Y direction (row of lenses normal to the road direction) to make control of the light in the Y plane (light hitting the road vs light hitting out of the sides of the road), each lens have to be really individual lens, they cannot be long "bars" covering the row of LEDs. But then, the gaps between the rows of lenses will allways be visible the same, from any point along the road...
Logged
xelareverse
Member
****
Offline

View Posts
View Gallery

Re: My questions « Reply #18 on: October 17, 2015, 11:33:16 PM » Author: xelareverse
What does the OV in most street light model numbers stand for?
Logged
mdcastle
Member
***
Offline

View Posts
View Gallery

mdcastle 26956281@N02
WWW
Re: My questions « Reply #19 on: October 18, 2015, 10:03:30 PM » Author: mdcastle
Besides people hating the yellow light, sending someone out at union wages in a bucket truck is a bigger cost impact than energy use, especially for utility company owned lights.
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: My questions « Reply #20 on: October 19, 2015, 12:40:23 AM » Author: Medved
... sending someone out at union wages in a bucket truck is a bigger cost impact than energy use, especially for utility company owned lights.

That part at this time makes the LED's one of the more expensive light sources: Random failures, needing either each one failure at least one such trip, or leave the lights not working till multiple fail in the same area and send the truck afterwards, the LED's are still just a reliability gamble, so you have to count on the related losses....
There are millions of promises about LED lifetime among their makers, but many were (at the time of installation) about a design style present just few months, so no chance it has undergo any test of time at all...
Although the situation is improving, the odds of choosing a problematic design makes the LED's one of the more expensive technologies...

With HID's you just tune the periodic group relamp scheme and with that you have one upfront planned bucket trip serving the maintenance of the complete street at once, so majority of the cost is shared among many lights.
That makes the HPS as the cheapest light of today - still good system efficacy, cheap lamps and over the years the reliability became well predictable (using quality lamps and scheduling lamp replacement in 2/3 of their rated life means practically no failures, so no need for any bucket trips outside of the relamping schedule).
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: My questions « Reply #21 on: October 19, 2015, 04:28:51 PM » Author: Ash
The LED markting personnel promise 100K hours in their presentation, and the bean counters dive in. There are no tech people involved in making decisions
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: My questions « Reply #22 on: October 20, 2015, 01:31:00 AM » Author: Medved
The thing is, there really are many designs really meeting the 50k+ lifetime, even among the cheapest products. The problem is, from the modern ones nobody could tell now, which ones these are, because just nobody knows it at all. You can tell which ones have at least the chance to meet that based on the present experience, but the thing is, the new designs are so new, there are too many possibilities for some new phenomenon to pop up, so you still have very high risk of frequent problems.

Today you may tell which of the 10 years old designs meets which life rating, but the problem is, these are half of the efficacy and many times more expensive. And to keep the collected data valid, you can not make any modifications to them, so these designs will stay that expensive and that inefficient - so unsellable at all.

The HPS have the advantage the technology haven't changed the last 25 or 20 years, so from that time you know which designs are the good ones and these are still competitive today, so using them means a guaranteed success.
But you do not have anything like this with any of the so fast evolving technologies like the LED lighting. So either you suffice with worse performing HPS, or you play Poker using the LED's - you may win (have extremely efficient really maintenance free installation for decades), but more likely you will loose (have failures every other week).

Ironically, the maker, who happen to have one of his first product (so jut the luck) trouble free, will have the updated products most likely very problematic. Because having the first one trouble free means he does not learn anything from that and the odds are never favorite all the time...
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: My questions « Reply #23 on: October 20, 2015, 03:40:45 AM » Author: Ash
Worse performing HPS ? As in - no blue peak to mess with your eye, no glare, and higher Lm/W ?
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: My questions « Reply #24 on: October 20, 2015, 06:51:02 AM » Author: Medved
Worse performing HPS ? As in - no blue peak to mess with your eye, no glare, and higher Lm/W ?

HPS loose mainly on the efficiency of the optic - overilluminate the road just underneath and close around the lantern, so for the same minimum illumination level you need higher power input.

Plus except really high wattages, even the brutto efficacy (not counting the optic) the efficacy is about the same (70W HPS has barely 86lm/W, 150W has about 120lm/W, while modern LED's are at about 120lm/W as well; extra ballast losses apply for both, there is no difference)

The glare comes from improper optical design (mainly using optical elements originally designed and sized for 60lm chips with modern chips of the same size emitting 200lm, obviously that means 3x higher brightness, exceeding the critical level), you can quite easily have the same problems with HPS as well, if you use 150W lamp with a refractor sized for 50W lamp...
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: My questions « Reply #25 on: October 20, 2015, 08:13:29 AM » Author: Ash
Most HPS lanterns designed in the last 20..30 years have excellent optical performance, unless there is a problem with the lantern like a yellowed plastic cover. I dont see why it is allways assumed that they dont perform as well
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: My questions « Reply #26 on: October 20, 2015, 10:17:54 AM » Author: Medved
I didn't say they are generally bad, I just compare what is achievable with HPS (or any other pointy light source radiating in really all directions) compare to LED's, where all the light output goes only into one half space.
The thing is, the optic can focus into an accurate beam the light only from one half space, the other half becomes effectively not focused at all. With discharges it means you may focus only half of the light into the far reaching lobes (whose need 80..90% of the total light flux in case of absolutely even illumination level on the road), the part of the light you are not able to focus is then used to illuminate the area underneath or close around the lantern. And because for the exact even illumination the "close around" needs just 10..20% from the total light output, but you spend there 50% of that, the 30..40% of the extra light just illuminate that above the level what is needed, so are what is sometimes called "over illumination losses". And of course generating this extra light you do not need means part of the power consumption does not generate any needed work (therefore it is equal to any other losses).
Because the LED already emit all of it's light only into one half-space, you may easily focus it by a refractor into any pattern shape you want, so by that reach exactly even illumination level, so do not spend any extra power than really needed for the specified minimum illumination level.
That means with the same task, you need 20..30% less raw lumens when you use LED's, compare to when you use a HPS (or any other pointy HID; with large surface lamps like coated lamps or mainly LPS the difference would be way greater, in the case of LPS the over illumination losses are in the order of at least 80% or so).
So when you need for certain stretch of road a fixture with a 6000lm HPS lamp (so typical 70W), the same minimum illumination level on the same area could be reached using just ~4000lm (so about 45W of 90lm/W average over life) LED's (or using 20000lm LPS; there even with the 220lm/W efficacy it means more than 90W for the lamp), that is the performance difference I'm talking about.
But as I said before, in most cases the 15W and about three planned group relamping (through the 12years expected life of the lanterns) related saving is not so significant to really pay off the risk when compared to the LED reliability gamble...
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: My questions « Reply #27 on: October 20, 2015, 01:10:37 PM » Author: Ash
I tried modelling it, and i get better results...



For area light, all light ends up in the intended area. The over illumination losses are ~0.21



For road light, the road appears thin from distance (at the 70 deg angle), so the lantern have to light in a thin beam in that direction. In deep clear cover lantern (Tungsram/Schreder Z3, ...) the light coming from the reflector can be focussed into this angle. The light coming directly from the arctube can not be controlled, so part of it will be spilled outside of the sides of the road (in front of and behind the lantern)

With the light spill losses, i calculate the light spill losses first, then the over illumination losses on the light that still hit the road

column height / road widthlight spill lossesover illumination lossestotal losses
1/40.130.10.22
3/80.140.090.23
1/20.160.080.24
5/80.170.070.25
3/40.190.060.25
7/80.210.050.26
1/10.230.040.27
9/80.240.030.27
5/40.260.020.28
11/80.270.020.29
3/20.290.010.3

So yep i get the 0.3, but that includes the light spill losses too - not sure if that was included in the 0.3..0.4 you provided, and the 0.3 is in extreme condition with the pole being 1.5 times taller than the width of the road....
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: My questions « Reply #28 on: October 20, 2015, 05:35:55 PM » Author: Medved
Indeed, the spill and overillumination I have summed up all together (the model I have used didn't allowed to distinguish).
The losses calculation came from the pole height/spacing being 1/4, with the pule height to the road width of about 2/3.
I just did a more detailed numeric calculation (the spread-sheet table calculators are really an universal simulator), capable to distinguish among these two components and indeed, the result shows all the losses come mainly from the light spill (16m pole spacing, 5m wide road plus 2m wide side walk, 4m poles and 10lx minimum level requires 1650lm of HPS and/or 1120lm of LED's).
But still this light spill is the direct consequence of the inability to focus half of the light output from an omnidirectional light source...

By the way the rough calculation with an LPS-like optic (so only what gets radiated above the horizon gets redirected below the horizon, no focusing at all), the example above would need nearly 17000lm lamp. There the vast majority of the losses would really go on the overillumination account...
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: My questions « Reply #29 on: October 21, 2015, 04:09:51 AM » Author: Ash
That would require 1240 Lm of LED, with 0.25 HPS losses as i modelled, and assuming 1.0 efficiency LED optics. But i think assuming 1.0 efficiency for LED optics right away is wrong, and here are the places where i think it will go wrong :

 - The lens size / chip size relation in a LED lantern is not great - the LED chip is not a point source. If the lenses are designed to aim precisely the light from the center of the chip, they spill the light from the perimeter.. Compare that to HPS, with 1ft wide optical box for 6mm wide arctube

 - The lens cannot be "infinite" on the sides where light does not have to go. There is a point where the lens of chip 1 must end, so the next lens may start. That mean high angled lens "dome" surfaces, but that also mean throwing the light to the distance.. Compare that to HPS, there the ends of the optic box see the arctube from its end, so dont get any significant amount of light

 - The chip is flat. The intensity it supplies into the lens is at peak vertically down and at peak cos(angle) for any other angle. How well can the bottom part of the lens redirect that in the wanted directions only ? The lens have to be complete i.e. there have to be some shape to it on the "unwanted" directions too. This part sees the chip as well, and the light coming to it will continue where it goes...

Logged
Print 
© 2005-2024 Lighting-Gallery.net | SMF 2.0.19 | SMF © 2021, Simple Machines | Terms and Policies