Building an LED light/flash

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: Pau, rjlittlefield, ChrisR, Chris S.

mawyatt
Posts: 2479
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater

Post by mawyatt »

mjkzz wrote:Finally, I think I know why viktor got different result. I am calling you out for an answer. If you do not trust me, we can get someone here and send answers to that person.
I don't know what is causing Viktor's results, since I don't have his devices and setup. My original comment was about looking at the COB LED current rather than voltage to see if ripple was causing significant current modulation which likewise would modulate the optical output.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mjkzz
Posts: 1237
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

mawyatt wrote:
mjkzz wrote:Next (so we can keep track of things).

What is your simulation model with your "circuit" for the load? And this is the critical part for all other accusation :D

If you have a COB LED, go through the exercise I mentioned, plot a chart with voltage and current on it. After you finish that, you will see that graph is VERY DIFFERENT from a typical chart for diode and this is the base I mentioned before, you will see the 59mV thing is NOT applicable at all, you do not have a line for current going up as steep as a diode would. Sure you might observe some similar behavior, but in the operating region of it, it is NOT the same, far (fetched) from it. Modern LEDs are made to emit more light than being a perfect diode. So from experiment, hands-on experience, one would not model a modern LED as a diode. I can have a power supply with 200mV ripple and still not having flikering.

Like I said before, all the stuff you throw out are common knowledge and are well understood by even high school kids. Is the 59mV theory (not yours, I will explain what "yours" is, I think you mis-interpret it or I did not make it clear :D). or what you called Schottky prediction wrong? NO. It is how you apply that. Can you really apply it to a device that has a voltage-current chart so different from a diode? NO.

Of course, all this are based on one fact -- actual plot of voltage-vs-current chart of a modern LED
Peter,

Do you even read the references or what I've said!! Think you are getting mixed up again, it's not Schottky as you've stated above, but William Shockley, the co-inventor of the bipolar transistor at Bell Labs in 1947. The equation that describes diode behavior is the Shockley Diode Equation.

Agree, you can't apply Schottky but yes you can apply the Shockley Equation as has been stated and referenced before, an LED is a DIODE!

Please remember COB LEDs and LEDs are not the same, please reread my post about what a COB is!!

The 59mv theory you seemed obsessed with is derived from the Shockley Diode Equation, this equation describes an ideal diode characteristics with a ideality factor of 1 and without any series resistance. LEDs are also diodes and they have series resistance.

From:
https://ecee.colorado.edu/~bart/book/bo ... /ch4_4.htm

4.4.4. I-V characteristics of real p-n diodes

Next Subsection
First, there is the ideal diode region where the current increases by one order of magnitude as the voltage is increased by 60 mV. This region is referred to as having an ideality factor, n, of one.



From:
http://www.eng.auburn.edu/~niuguof/2210 ... diode.html

An ideal diode has the following Shockley I-V characteristics:


the voltage difference corresponding to 10x current difference is 60 mV theoretically for ideal diodes. Higher order physical effects degrade the I-V slope. As a result, it takes more than 60 mV voltage increase to increase the current by one decade, or 10x.

near room temperature we can deduce the 60 mV per decade rule, which says that Vf increases approximately 60 mV for every factor of 10 increase in Id


BTW we are all anxiously waiting for you to provide a reliable reference that indicates an LED is NOT a diode, even a reference that doesn't call it a diode or pn junction!!

Best,
Shockley Equation and Schottky . . . hahaha OK, my mistake, but it is irrelevant.

We are going in circle.

Anyways, please do that exercise like I mentioned, it is as real and as hands on as it could. It describes a true characteristic of an device even if there are equations for it. That said if you look at it within its operating range, you will find it is a very different device.

Do I have to quote Einstein to prove something? If a device characteristic is so different, sure, in physics, like I said before, a diode is a diode, but when design a circuit, it does not behave like a diode, thus should be modeled differently and treated differently.

If you do not agree with that, I have nothing to say.

mjkzz
Posts: 1237
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

mawyatt wrote:
mjkzz wrote:
mawyatt wrote:
mjkzz wrote:
THIS ACTUALLY IS THE BASE of my suggestion that you can not use it on all LEDs and I kept calling out for the LED model you used in your "circuit".
Peter,

If the "circuit" you reference is the "Fast LED Strobe for Macro" thread

https://www.photomacrography.net/forum/ ... hp?t=40999

This is current mode control and the LED characteristics don't matter since the current is forced thru the LED regardless of the LED parameters (within reason). This circuit is so immune to the load that you could put a LED, resistor, diode string, even a battery, or some combination of these and the load current wouldn't change unless the current source compliance voltage was exceeded. This is the benefit of negative feedback constant current mode operation. The model used in the simulation was just an ideal zener diode of the proper voltage and a resistor.

Best,
sure, I know this kind of circuit, I think we discussed this type of thing before, feedback.

I am not accusing that you are wrong, but when you design something to drive an LED, don't you think it is a good idea to put an LED model as load? No, no, nothing wrong with what you did, but like your signature says, you never know what you will find :-)
Since you claim to "know this kind of circuit", care to give us a detailed analysis of how it works?? Why many of the load characteristics don't matter.

Best,
Well, a simple word, it basically have sensing resistor, feeding back to comparator that turns of and off a transistor. But unlike you, I do have PHYSICAL device built and is being used by someone, a year or two ago, you probably know this person, too. And it can be controlled by a computer.

mjkzz
Posts: 1237
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

One last thing, there is something called numerical analysis where you do not need an equation, as long as you have some empirical data (ie, the charts, the chart), you can compute it. Again, with that chart, you do not need an equation whether it is Shockley Equation, or Schottky, or even Einstein, that is irrelevant.

You are so hang up on those equations, so fixed upon having a famous reference, in fact there are experiment you can perform, collect data and build a model with it. Do you get it?

And these data show a chart so different from a diode, it is wrong to use an equation for a diode in your circuit.

So which part of it you do not understand?

mawyatt
Posts: 2479
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater

Post by mawyatt »

mjkzz wrote: Shockley Equation and Schottky . . . hahaha OK, my mistake, but it is irrelevant.

Shockley and Schottky certainly are not irrelevant, maybe the person making these mistakes is irrelevant!!

We are going in circle.

Anyways, please do that exercise like I mentioned, it is as real and as hands on as it could. It describes a true characteristic of an device even if there are equations for it. That said if you look at it within its operating range, you will find it is a very different device.

Wrong again, the LED is a diode, behaves like a diode, will always be a diode, and follows the Shockley DIODE equation with a series resistance and ideality factor. Simple as that, references have been given. You can ramble all you want but you can't change that an LED is a DIODE. Can you prove otherwise, maybe quote a qualified reference (read certainly not YOU) that states an LED is NOT a DIODE?? We all are still waiting for this reference???

Do I have to quote Einstein to prove something? If a device characteristic is so different, sure, in physics, like I said before, a diode is a diode, but when design a circuit, it does not behave like a diode, thus should be modeled differently and treated differently.
See above, an LED certainly DOES behave like a DIODE since it is a pn junction, thus does not need to be modeled and treated differently.

If you do not agree with that, I have nothing to say. Now, I think I am wasting my time with you.



My comments in red.


How can someone agree with something that's is totally WRONG, LEDs are DIODEs period!!!

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2479
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater

Post by mawyatt »

mjkzz wrote:
mawyatt wrote: Since you claim to "know this kind of circuit", care to give us a detailed analysis of how it works?? Why many of the load characteristics don't matter.

Best,
Well, a simple word, it basically have sensing resistor, feeding back to comparator that turns of and off a transistor. But unlike you, I do have PHYSICAL device built and is being used by someone, a year or two ago, you probably know this person, too. And it can be controlled by a computer.

I think I have wasted too much time with you. Bye
Nope sorry Peter, not how it works! Again you've demonstrated your depth of knowledge of electronics and semiconductors. Maybe you ask some of your "high school" colleagues to help you with this simple circuit??

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2479
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater

Post by mawyatt »

mjkzz wrote:One last thing, there is something called numerical analysis where you do not need an equation, as long as you have some empirical data (ie, the charts, the chart), you can compute it. Again, with that chart, you do not need an equation whether it is Shockley Equation, or Schottky, or even Einstein, that is irrelevant.

You are so hang up on those equations, so fixed upon having a famous reference, in fact there are experiment you can perform, collect data and build a model with it. Do you get it?

And these data show a chart so different from a diode, it is wrong to use an equation for a diode in your circuit.

So which part of it you do not understand? If you still don't, please pick up a book about numerical analysis and read on it.
While in grad school I wrote a numerical methods program utilizing the Newton-Raphson Method to help solve some complex transcendental equations derived for predictive analysis, basically predicting in the future when an event would occur, in the specific case when phase coincidence would occur. This provided the foundation for the development of synchronous power transfer technology which was later patented (US Patent 4310771 and 4256972). Today this is still used and provides a safe means for power transfer in Nuclear Power Plants as one example.

SPICE (Simulation Program with Integrated Circuit Emphasis) is a numerical methods program that I've been using since in was introduced by Nagel at Berkeley in 1973! Got the pleasure of meeting Dr. Nagel at the IEEE ISSCC conference a few years ago, brilliant individual!!

Think I understand numerical methods quite well, thank you!!

Maybe you should read up on numerical methods, and while your at it be sure to get a text on Introduction to Electronics or better yet Electronics for Beginners.

https://www.explainthatstuff.com/electronics.html

Pay particular attention to the chapter on Diodes and LEDs:

"An LED is a special type of diode (a type of electronic component that allows electricity to flow through in only one direction). Diodes have been around for many decades, but LEDs are a more recent development. Let's take a closer look at how they work!"

Best,
Last edited by mawyatt on Wed Mar 11, 2020 8:22 am, edited 1 time in total.
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2479
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater

Post by mawyatt »

mjkzz wrote:
mawyatt wrote:
mjkzz wrote:
mawyatt wrote:
mjkzz wrote:
THIS ACTUALLY IS THE BASE of my suggestion that you can not use it on all LEDs and I kept calling out for the LED model you used in your "circuit".
Peter,

If the "circuit" you reference is the "Fast LED Strobe for Macro" thread

https://www.photomacrography.net/forum/ ... hp?t=40999

This is current mode control and the LED characteristics don't matter since the current is forced thru the LED regardless of the LED parameters (within reason). This circuit is so immune to the load that you could put a LED, resistor, diode string, even a battery, or some combination of these and the load current wouldn't change unless the current source compliance voltage was exceeded. This is the benefit of negative feedback constant current mode operation. The model used in the simulation was just an ideal zener diode of the proper voltage and a resistor.

Best,
sure, I know this kind of circuit, I think we discussed this type of thing before, feedback.

I am not accusing that you are wrong, but when you design something to drive an LED, don't you think it is a good idea to put an LED model as load? No, no, nothing wrong with what you did, but like your signature says, you never know what you will find :-)
Since you claim to "know this kind of circuit", care to give us a detailed analysis of how it works?? Why many of the load characteristics don't matter.

Best,
This post alone shows you how proud you are about your "circuit", but I have to tell you, it is just as simple as something can be done by a high school kid. There is nothing you should be proud of, as matter of fact, having 40 years career and showing off this type of circuit is kind of funny.
Yes it is a simple circuit, since you've already demonstrated that this "circuit" is beyond your understanding (your prior description of the operation is wrong, and you never addressed the load independence), maybe the "high school kid" you mentioned could help!!

BTW I have over 50 years experience, and this "circuit" isn't "showing off" (it's too simple to use for "showing off"!!) it was provided so that folks could develop their own COB LED Current Mode Controllers as DIY types. I even offer to help, but most folks here on PM don't require much help, if any. However, it seems this "circuit" is way to complex for your understanding, so you'll need to engage some help, maybe the "high school kid" you've mentioned??

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mawyatt
Posts: 2479
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater

LED is a DIODE

Post by mawyatt »

Peter,

Here's a few LED references for you, many more are available.

From Webster:
https://www.merriam-webster.com/dictionary/led

LED noun
\ ˌel-(ˌ)ē-ˈdē \
Definition of LED (Entry 2 of 2)
: a semiconductor diode that emits light when a voltage is applied to it and that is used especially in electronic devices (as for an indicator light)

From Dictionary.com:
https://www.dictionary.com/browse/led

LED[ el-ee-dee, led ]SHOW IPA
noun
light-emitting diode: a semiconductor diode that emits light when conducting current and is used in electronic displays, indoor and outdoor lighting, etc.

From Yourdictionary.com:
https://www.yourdictionary.com/led

led
pt. & pp. of lead
LED

a semiconductor diode that emits light when voltage is applied: used in electric lighting, alphanumeric displays, as on digital watches, etc.


From:
https://circuitglobe.com/light-emitting-diode-led.html

Light Emitting Diode (LED)
Definition: The LED is a PN-junction diode which emits light when an electric current passes through it in the forward direction. In the LED, the recombination of charge carrier takes place. The electron from the N-side and the hole from the P-side are combined and gives the energy in the form of heat and light. The LED is made of semiconductor material which is colourless, and the light is radiated through the junction of the diode.

Please note that an LED must have a diode junction to emit light, thus it must be a diode!!

So we all end up with a smile :D

If you don't have any "diode junction" how can you have any "LED light", how can you have any "LED light" if you don't have any "diode junction"!!

Since after all, we are all just another Brick in the Wall!! :lol:

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

ChrisR
Site Admin
Posts: 8564
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Gentlemen, the tone of this thread has gone off the rails.
I have deleted most of the ad-hominem comments.

Please let this one cool down.
Chris R

mawyatt
Posts: 2479
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater

Post by mawyatt »

Chris,

Sure, sorry for being so stubborn!!

That's why I introduced a little Pink Floyd to help lighten things up in my last post :D

Since I was asked to provide an LED model and measured LED data, which I'm delinquent on providing, here's a Cree XM-L2 LED model just created with SPICE using the default SPICE diode model & parameters. No attempt was made to include dynamic, capacitive or reverse bias conditions, just forward bias. Surprisingly it only required 2 parameters to give a pretty good fit, Series Resistance (RS) of 0.147 ohms and Ideality Factor (N) of 3.02. Cree provides a graph of measured I vs. V for the XM-L2 LED at 85C, so this temperature was used for the curve fit.

Please delete this is you think it's inappropriate.

Best,



Cree measured XM-l2 LED data @ 85C
Image

SPICE Simulated LED with diode default model @ 85C with RS=0.147 & N=3.02

Image
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

mjkzz
Posts: 1237
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

OK, things are getting back to normal, albeit slowly and my holidays are almost over, too :(

Got the reflector delivered, this is a 100W "driverless" LED with CRI of 95+. Anyways, to contribute something concrete, here it is. Key point is to look for "driverless" LEDs where you can simply put it on a switching power supply. NO FLICKERING even at 1000fps not sure about higher frame rate as I do not have a camera which can do that.

If you flash it, ie, pulse it, you can use a smaller heat sink and go fanless. It all depends on how you use it.

300W version sample coming soon . . . oh my eyes!!! :D

Image

chris_ma
Posts: 260
Joined: Fri Mar 22, 2019 2:23 pm

Post by chris_ma »

love the heat sink :)
chris

mjkzz
Posts: 1237
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

chris_ma wrote:love the heat sink :)
thanks. the fan is problematic for high magnification work due to vibration and air flow, but there are cool looking (if that is a factor for you) liquid CPU heat sink

Anyways, back to work soon, no time for this anymore. :cry:

mawyatt
Posts: 2479
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater

Post by mawyatt »

The heatsink is cool and looks like a car engine manifold, hot rod LED COB!!

Fans are usually a problem, here's a Godox 150watt 16,000 lumen Video Light where the fan is reported as too loud, so it's vibrating something.

https://www.bhphotovideo.com/c/product/ ... w_led.html
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Post Reply Previous topicNext topic