Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Why is anybody using High Frequency Switched Mode LED drivers to power LED streetlights?

Status
Not open for further replies.

Flyback

Well-Known Member
Hello,
My question is why is anybody using High Frequency Switched Mode LED drivers to power LED streetlights?

As the attached LTspice simulation and schematic shows, simple ‘Switched Linear Regulator’ based LED streetlights are very good and have admirable efficiencies.

They also need no electrolytic capacitors. Also, they are far easier to protect from mains transients (the biggest streetlight ‘killer’ out there) than High Frequency Switched Mode LED drivers. ‘Switched Linear Regulator’ based LED streetlights are easier to protect from mains transients than High Frequency Switched Mode LED drivers because the FETs can be very high voltage rated, without this badly affecting the efficiency. If a High Frequency Switched Mode LED driver uses a high voltage FET with a low enough Rds(on), then its switching losses become excessive. Therefore, a High Frequency Switched Mode LED driver based LED streetlight will use a lower voltage rated FET and be more susceptible to mains transient related failure.

Also, the ‘Switched Linear Regulator based LED streetlight’ needs no custom wound transformer and uses few different components.

The only disadvantage is the larger number of LEDs that are needed to make the ‘Switched Linear Regulator’ based LED streetlight. However, the cost of this is totally offset by the reduction in susceptibility to mains transients…..which means a dramatically lower failure rate and massively less replacement costs.

The power factor of the “Switched Linear Regulator based LED streetlight”, is not as good as the High Frequency Switched Mode LED drivers can be, however, it is good enough to pass regulations. After all, streetlights have only low Power Factor and Mains Harmonic regulatory requirements because they are only on at night, when the electricity grid is under-utilised, and so power factor does not matter anywhere near so much.

So why on Earth is anybody using High Frequency Switched Mode LED drivers for LED streetlighting?

The attached is a Streetlight with 63.5W of LED power. It is 76% efficient. It uses only switched Linear current regulators (which at certain times are “swamped” and have little dissipation during those times).

The linear regulators are switched ON/OFF at around the twice mains frequency.

The efficiency actually varies with the mains input level, and in fact at 265VAC input, it is actually 84% efficient, and gives 77.5W of LED power.

If you wanted to add a microcontroller and change the reference voltage of the linear regulators as the mains input voltage changed, then you could make it >84% efficient over the whole 206 to 265 VAC range. It wouldn’t be workable over the full 100-265VAC mains range however.

Attached is the LTspice simulation, pdf schematic and mains input voltage and current waveform scope shot.

So indeed, why on Earth is anybody using High Frequency Switched Mode LED drivers for LED streetlighting?
 

Attachments

  • LED streetlight_schematic.pdf
    205.9 KB · Views: 135
  • LED streetlight_Input waveforms.pdf
    37.8 KB · Views: 112
  • LED streetlight.asc
    52.4 KB · Views: 105
Hmm, this seems like it is a rather niche area, but there are always going to be trade-offs with any approach. I wonder if the lack of smoothing in the waveform would result in 100-120Hz flickering? In most cases it might not be noticeable, but with fast-moving objects like cars, it might create a noticeable strobing effect.

As you say, having to use a larger number of LEDs is also a reasonable factor here. I'm sure one could crunch the numbers on the subject and determine whether or not the cost of the extra LEDs would be offset by the lower failure rate due to mains transients, but sadly the temptation of lower up-front costs often wins over long-term strategies anyway.
 
Most LED bulbs have good power factor and your design does not.

Inside a light bulb there is much heat generated from the LEDs. There is not room for more heat from a linear regulator.
 
I have seen many segments in street lights and traffic lights burned out. There is probably not a lot of manufacturers of LED street lights so they make them burn out soon and sell many replacements? Maybe they are made in the country where nothing lasts long?
 
Maybe they are made in the country where nothing lasts long?
If all the LEDs were in parallel; when one dies it would not be too objectionable.
When the LEDs are in strings of 10; when one dies it also removes 9 neighbors. This is easy to see.

Example; we have 30LEDs in three strings of 10. A failure is defined as 1/3 of the LEDs not working.
If the LEDs were in parallel, it would take 10 dead LEDs to make a failure. With the series//parallel thing it only takes one dead LED to get a failure. (the failure rate is 10X)

(LED bulbs in street lights) I read a paper that stated the cost of changing light bulbs (in man hours) has dropped way down, because of LED light bulbs.
 
Somebody makes an IC that shorts a burned out LED in a series string so that the remaining LEDs work. Or LEDs in a street or traffic light are paralleled because some parts are off while most of the remainder still works. Nobody cares that a few parts are not lighted so they will not replace the entire light until it is really bad.
 
thanks i know the ic you mean, onsemi make them, they are called "LED open" ic's , or something like that
 
Most LED bulbs have good power factor and your design does not.

If you run the simulation in the top post and calculate the power factor, you see it is 0.829
…That is actually within limits for streetlighting, which doesn’t have much power factor requirement because they are always only on at night.
So PFC is no problem here.

These kind of lights also get used in dock areas etc where they need night time lighting outdoors.....these are within company premises, and all business's can have any power factor they like, because they actually pay for the reactive power that they draw....so even if the power factor was 0.2, it would be allowable...because they (business's) pay money for their reactive power draw.

It is quite simply the case, that due to the improved mains transient withstand of this switched linear regulator LED streetlight, it is far superior to a high frequency SMPS based LED streetlight.
Its all about how long it lasts, because maintenance and replacement cost is high.

Also, the efficiency is between 76% and 96% depending on mains voltage level...that's well good enough for an offline LED light, and is well comparable with a high frequency switched mode offline LED driver based street light
 
Last edited:
This Switched Linear regulator LED streetlight (attached) is over 96% efficient at 240VAC input.

It has a power factor of 0.89

I really am wondering why anybody is doing LED streetlights driven by High Frequency Switched Mode LED drivers. It is surely madness?

The “Switched Linear regulator” type is far more resilient against mains transients, which are the biggest killer of LED streetlights by far. And it costs a fortune to replace failed streetlights.

Attached is pdf schematic, LTspice simulation and Mains input voltage and current waveforms of an even better Switched Linear regulator LED streetlight. It uses far less LEDs than the one in the top post.

As discussed, streetlights are Government property. Your Government is ultimately responsible for Electrical Regulations. Your Government is going to prefer streetlights which have a slightly lower power factor (0.89 in this case) but save money due to the fact that they need replacing less.

Streetlights that are used in large company premises such as dockyards can have any low power factor that is wished, because big companys pay for their reactive power, so therefore they can pollute the mains as much as they like.

I can see a worldwide revolution towards “Switched Linear regulator” LED streetlights. Do you agree?
 

Attachments

  • Offline Linear LED drive.asc
    19.6 KB · Views: 111
  • Offline Linear LED drive _Schematic.pdf
    99.4 KB · Views: 138
  • Offline Linear LED drive _Input Waveforms.pdf
    33.2 KB · Views: 107
https://led-driver.power.com/sites/default/files/PDFFiles/der548.pdf
upload_2017-3-16_20-46-10.png

upload_2017-3-16_20-48-46.png
 
Thanks, as you allure with your “1” sticker on the electrolytic capacitor, that LYTswitch design is no good for a street light because it uses an electrolytic.

The “Switched Linear Regulator” method that I show does have some drawbacks…..The dissipation in the low side FET can be very high at high mains voltage, unless extra circuitry discussed below is added. Also, the LED current can vary with mains voltage and also the LED current varies with the forward voltage variation of the LEDs.

The solution to these problems would be to have a microcontroller to switch in/out various strings of LEDs, at various points in the overall LED string in order to match the LED string length and current to the mains level, and to the overall forward voltage of the LEDs. This would be like an adaptive control. The problem is, that each switch would not be ground referenced, and so would need its own high side supply and digital isolator to interface to it. This gets a little messy.

Does anyone know of any FET drivers which pass through a little power to the high side without needing a pulse transformer or high side supply? As you know, hardly any power is needed to power the high side fet drives....if some kind of RF chip were available which could drive the fet and pass through enough power to do so?
 
Last edited:
Does anyone know if there are off-the-shelf modules for the ferrite pulse transformer gate drive circuits which exist in the attached version of the “Switched Linear Regulator” LED streetlight?
Otherwise its going to be a lot of circuitry…..because more series FETs will be needed than shown in order to handle the variation in LED Vf and Mains voltage.
 

Attachments

  • Schematic_Offline Linear LED drive_1.pdf
    92 KB · Views: 122
  • Offline Linear LED drive_1.asc
    30.1 KB · Views: 122

Attachments

  • da23xx_gate[1].pdf
    223.4 KB · Views: 117
  • fa2659[1].pdf
    81 KB · Views: 131
  • da2099[1].pdf
    127.4 KB · Views: 124
  • pwrbase[1].pdf
    449.6 KB · Views: 122
  • P732[1].pdf
    531.4 KB · Views: 147
  • P774[1].pdf
    283.5 KB · Views: 132
You still on about switched mode supplies for lights? You have years and years worth of this stuff on here! I seriously thought you had asked just about every possible combination of questions on this, maybe time to go see the DR! You have a unhealthy fixation on this.
 
You still on about switched mode supplies for lights? You have years and years worth of this stuff on here!
Thats just it, the LED streetlight that i have proposed here is totally and utterly new!..it is a world first!!...totally different than anything either i or anyone else has ever asked about before!!!....it uses the revolutionary "switched linear regulator" method to drive LED streetlights.

Thanks RonSimpson for the gate drive sample show...i will look through those, but actually, i am wondering about modules containing the whole gate drive circuitry that i drew out in the above posts, rather than just the transformer. As you know, we will need a lot of gate drives, and our components count will go too high if the entire gate drive circuit is not in a module. There are lots of led bank to switch in and out as the mains half cycle rises and falls.
 
Last edited:
Thats just it, the LED streetlight that i have proposed here is totally and utterly new!..it is a world first!!...totally different than anything either i or anyone else has ever asked about before!!!....it uses the revolutionary "switched linear regulator" method to drive LED streetlights.

Thanks RonSimpson for the gate drive sample show...i will look through those, but actually, i am wondering about modules containing the whole gate drive circuitry that i drew out in the above posts, rather than just the transformer. As you know, we will need a lot of gate drives, and our components count will go too high if the entire gate drive circuit is not in a module. There are lots of led bank to switch in and out as the mains half cycle rises and falls.
No its exactly one of your first posts, the one you mention a lighting company and NEW/revolutionary office light system...............

Ok my bad, that was inside an office, this is outside. So yeah totally new and never thought of, every light engineer is an idiot for not doing it this way. Why not call a company and ask why they dont do it differently? The answer might save you time and effort.

Get them to sign a NDA, you can get legal binding ones done over the internet.
 
If it is so revolutionary and awesome, you need to be talking to either street light manufactures or public utility districts. Preaching here isn't going to get the ear of anyone who matters.

Good ideas are simple. Putting them into use is the hard part.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top