High Power LED Stroboscope Problem

Hello guys
Sorry for wrong topic selection and bad English

I'm working on a project with 3*10 Watt High Power LED's wich act as a frequency & duty variable stroboscope via a Micro Controller . The problem is for accurately stopping a high speed rotary device the duty should be below 2% ( as an experience ) . but with such a low Duty, brightness of LEDs are so low . I know that with higher current pumping to LEDs the problem can be solved . but the question is HOW? and is there a way that current flowing through LEDs be adjusted with selected duty in Code ? would you please help me on over current protection circuity too ?

Thanks a lot

With low duty cycles you increase the current greatly in order to maintain brightness - however, LED's are never going to come close to the brightness of a proper xenon strobe tube.

Nigel Goodwin said:

With low duty cycles you increase the current greatly in order to maintain brightness - however, LED's are never going to come close to the brightness of a proper xenon strobe tube.

Click to expand...

Thank you for your reply
The problem is How?
because the LED forward voltage is 10-12 volt and to get greater currents , voltage should be multiplied by a factor of about 50 (2% Duty) means 500-600 volts!, and i cant afford such a voltage nor LED candle even 20% of it!
( I get about 5 mA with LED directly connected to a 20 A - 12V Computer power supply without any current limitting equipment , while strobing with 60Hz,1% Duty ) - and i cant even think about high voltage transformers like the one use in crt TVs - and i couldnt fount any step up regulator to do the job

Is there a specific type of circuit to do the job with a "high current capable but low voltage power supply" ?
thanks in advance ...

If your LED's are 12V types, then they probably have current limiting inside, and may well have constant current circuitry - in which case you can't increase their current.

You need to check the makers datasheets about using them.

For a normal LED you just reduce the value of it's series resistor, or increase the current of any constant current source.

Nigel Goodwin said:

If your LED's are 12V types, then they probably have current limiting inside, and may well have constant current circuitry - in which case you can't increase their current.

You need to check the makers datasheets about using them.

For a normal LED you just reduce the value of it's series resistor, or increase the current of any constant current source.

Click to expand...

Thanks again and again and ...

I dont think that it has such a thing inside .
Its a chineese low cost in production LED with the model HLS-HW-R10W100 and search for a standard data sheet had no result . The one and only data available from dealer :
**broken link removed**

Link(https://www.alibaba.com/product-gs/340586289/10watt_high_power_led.html)

Product Attributes
Place of Origin:
Guangdong China (Mainland)
Brand Name:
Highlights
Model Number:
HLS-HW-R10W110
Color:
White
Life:
50,000h
Detailed Description

10w power led




Power LED Applications
Indoor Lighting: Spotlight,Ceiling light,Bulb
Architectural and Landscape Light: Down light,Wall lamp,Garden light
Roadway Lighting: Street light,Garden light,Tunnel light
Portable and Personal lighting: Torch,headlight
Display Lighting: Display case

High-power LED as light source for illumination has the following advantages:

1, less power consumption: light effects to 75lm / w of the LED power consumption than equivalent incandescent brightness reduced by 80%;

2, long life: product life of up to 50,000 hours, 24 hours of continuous light can be used for 7 years;

3, nanosecond response time, brightness and color make it easy to dynamically control: enables dynamic color changes and digital control;

4, a large design space: the organic architecture can be realized with the integration to only see the effect of light not seen light;

5, environmental protection: no toxic metals mercury, no infrared and ultraviolet radiation;

6, color: different wavelengths produce different colored light, bright saturated, no filter, can control red, green and blue primary colors to form a variety of colors, can realize full-color gradients, and other color effects.



Electrical Parameters
Product Model Color Luminous Intensity(φV) Electrical Parameters
view angle Voltage(Vf) Current(IF)
10W(round)3 parallel,3 series HLS-HR-R10W red(R) 30-40lm/W 180° 6-8v 1050mA
HLS-HY-R10W yellow(Y) 30-40lm/W 6-8v 1050mA
HLS-HB-R10W blue(B) 10-25lm/W 10-12v 1050mA
HLS-HG-R10W green(G) 70-80lm/W 10-12v 1050mA
HLS-HW-R10W90 white(W) 80-90lm/W 10-12v 1050mA
HLS-HW-R10W100 90-100lm/W 1050mA
HLS-HW-R10W110 100-110lm/W 1050mA
12W(round)3 parallel,4 series HLS-HR-R12W red(R) 30-40lm/W 9-12v 1050mA
HLS-HY-R12W yellow(Y) 30-40lm/W 9-12v 1050mA
HLS-HB-R12W blue(B) 10-25lm/W 12-16v 1050mA
HLS-HG-R12W green(G) 70-80lm/W 1050mA
HLS-HW-R12W90 white(W) 80-90lm/W 1050mA
HLS-HW-R12W100 90-100lm/W 1050mA
HLS-HW-R12W110 100-110lm/W 1050mA

Yeah, Nigel Pointed out, No LED have 12V forward voltage drop. Its near, 3.5 Volts, I think. So, first get a proper LED. Then, by using higher supply voltage and lower value of current limiting resistance, you can get acceptable Brightness even at low duty cycle. You can always use more of them in Parallel, to increase the brightness.
No matter how hard you try, you can't beat the sun-light, and even the poorest Illumination works at night. So, actually, its a compromise.

PICMICRO said:

Yeah, Nigel Pointed out, No LED have 12V forward voltage drop. Its near, 3.5 Volts, I think. So, first get a proper LED.

Click to expand...

Many White LED stacks are rated at 10 to 12V.? They are proper LED's

PICMICRO said:

Yeah, Nigel Pointed out, No LED have 12V forward voltage drop. Its near, 3.5 Volts, I think. So, first get a proper LED. Then, by using higher supply voltage and lower value of current limiting resistance, you can get acceptable Brightness even at low duty cycle. You can always use more of them in Parallel, to increase the brightness.
No matter how hard you try, you can't beat the sun-light, and even the poorest Illumination works at night. So, actually, its a compromise.

Click to expand...

Its possible to have such forward voltage because of nine 1 watt cells are in its die with a Series-parallel arrangement

Thank You all for your usefull guides
but the main problem is still not answered
My strobeoscope should have the ability to change the duty in its user interface(digitally)
if i set current for a specific one , the LED may harm in higher selectable duties and i really need a way to dynamically change this current .
i can set a resistor network and dynamically add or omit some resistors from a series of networks ( or even use a digital pot ) , but then it will be stepwise and its noticable change of brightness is my concern . i want the brighness to be always the same even by changing duty cycle ! ( just the opposite of dimming methods with PWM )

Is there a way ?
thanks ...

You don't need a variable MARK ratio for a strobe, only a variable SPACE - so the ON time can be keep constant, and vary the OFF time to alter the speed, just as with a real strobe.

It really depends if you can get enough brightness from an LED? - real xenon strobes are exceptionally bright.

If you have a variable power supply try feeding the LED from it, whilst monitoring the current, see if the current increases as you gradually turn the voltage up. Assuming there's no current limiting inside the LED, then the current should increase hugely for a small voltage increase.

You'll also have to be aware that the limiting factor with the LED will be the thermal characteristics -- how much heat can be produced in the die during the pulse without damaging the life of the die. This is the upper limit on what you'll be able to do. LEDs just don't make good strobes compared to xenon flash tubes, especially in normal room lighting conditions.

Nigel Goodwin said:

You don't need a variable MARK ratio for a strobe, only a variable SPACE - so the ON time can be keep constant, and vary the OFF time to alter the speed, just as with a real strobe.

It really depends if you can get enough brightness from an LED? - real xenon strobes are exceptionally bright.

If you have a variable power supply try feeding the LED from it, whilst monitoring the current, see if the current increases as you gradually turn the voltage up. Assuming there's no current limiting inside the LED, then the current should increase hugely for a small voltage increase.

Click to expand...

Thats the point
I should disable Duty change option or set limit to about 10% max
Appreciate you nice guys

I decided to use single 1 watt LEDs in series instead of 10 watt High power LED . but i get only 6-7 mA from each when using 1% Duty cycle with no use of any current limiting resistor . My Power supply is producing 20A of current in every of its 12 volt regulated output rails . and i really prefer to use this because of production costs ( The rotary system uses this power supply installed and i want to make a stroboscope for it)

even when i use simple 5mm white LED , i get 4 ma current ! Any suggestion to make it better ?
i used a N-channel power mosfet to drive these LEDs with logic output of uC ( Panasonic K2129 about 3A continuous and 6A in pulse current absolute rating)

And because it may be used in wet areas low voltage LED type is preferred

squishy36 said:

You'll also have to be aware that the limiting factor with the LED will be the thermal characteristics -- how much heat can be produced in the die during the pulse without damaging the life of the die. This is the upper limit on what you'll be able to do. LEDs just don't make good strobes compared to xenon flash tubes, especially in normal room lighting conditions.

Click to expand...

I really want to know absolute ratings . but these Chinese manufacturers are so weak in documenting their products , and i don't have any other non-Chinese option to select .
Thanks anyway

The data sheet:the LEDs are (3 in parallel) and (3 or 4 in series). You can see in one of the pictures a 3x3 block of LEDs. So because there are 3 LEDs in series the voltage could be about 12 volts.

Again: LEDs should never be driven with voltage (with out a resistor). LEDs are current devices and should be driven with constant current.

If we had a real data sheet I could tell you how much they could be over driven for small duty cycles.

I guess the 1W led you are trying to use is 3V type ? So, even with that the current isn't enough at 1% duty cycle?
Then, there is only one way to go, I think, Increase the supply Voltage. I guess computer Power supply has 24 Volt supply as well.

ronsimpson said:

The data sheet:the LEDs are (3 in parallel) and (3 or 4 in series). You can see in one of the pictures a 3x3 block of LEDs. So because there are 3 LEDs in series the voltage could be about 12 volts.

Again: LEDs should never be driven with voltage (with out a resistor). LEDs are current devices and should be driven with constant current.

If we had a real data sheet I could tell you how much they could be over driven for small duty cycles.

Click to expand...

yes , it has a 3*3 matrix inside
unfortunately i want to use PWM for strobing ,but its general use is dimming
use a method that everybody uses to dim the light , in a project which maximum brightness is an important factor makes me nervous

PICMICRO said:

I guess the 1W led you are trying to use is 3V type ? So, even with that the current isn't enough at 1% duty cycle?
Then, there is only one way to go, I think, Increase the supply Voltage. I guess computer Power supply has 24 Volt supply as well.

Click to expand...

Thanks again
Power supply only has multiple 12V,5V and 3.3V voltage
i used an unregulated (stabilized transformer type) 2A 24Volt supply and now i have broken 1Watt LEDs , and i dont know why ?!
The LEDs where mounted on their own heat sinks and burned in less than a second after turned on

is it possible that the 2200uF 50V capacitor in my supply be the problem ?

m_seifi_e said:

I decided to use single 1 watt LEDs in series

Click to expand...

Don't put them in series, put them in parallel. Give each one a limiting resistor to balance the minor differences in voltage developed across each led - this causes the slightly lower one to hog most of the current, and you have a very demanding application.

Before you get too far into this, find a way to rig up a quick test to see if you can get them to blink as fast as you need them to. I had dismal results making a white led stroboscope - leds are slower than xenon tubes or lasers because of the large die capacitance (using a push-pull driver will help), and white are even slower because of the phosphor.

duffy said:

Don't put them in series, put them in parallel. Give each one a limiting resistor to balance the minor differences in voltage developed across each led - this causes the slightly lower one to hog most of the current, and you have a very demanding application.

Before you get too far into this, find a way to rig up a quick test to see if you can get them to blink as fast as you need them to. I had dismal results making a white led stroboscope - leds are slower than xenon tubes or lasers because of the large die capacitance (using a push-pull driver will help), and white are even slower because of the phosphor.

Click to expand...

Thank you for reply
I already tested with just one LED directly connected to power supply and had no success
May I ask that in your designed Stroboscope how much current in every pulse satisfied your desires ? and what type of LED you used ?
thanks again

Have you thought about a integrated High Power LED Driver?
There are some types availible that supports analog dimming, like the CAT4201 from CATALYST.

That means the current througt the LED's can be steered by an analog voltage.
When you give your pulses to a adjustable voltage divder, e.g. digital Pot, you can bias the current in a wide range. Or you change the Value of the Current setting resistor.

Every LED has an maximum current limit. When that limit was exceeded for shortest time the LED is damaged. So the current driver should never overshot that limit.