"Perfect" LED current limiting challenge

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
Robert Reed
Posts: 2276
Joined: Wed Nov 24, 2004 1:01 am
Location: ASHTABULA,OHIO
Contact:

Re: "Perfect" LED current limiting challenge

Post by Robert Reed » Wed Jan 18, 2006 4:25 pm

Chris
This is not to debate or deny your experimenting with various devices, but I have noticed in this post and several others, that you continully push devices to their limits and well beyond. It can be done occasioinally, as I have experimented with some of these myself, just out of curiosity. However, there is a lot to be said for the manufacturers maximum ratings.If you were to produce 100,000 units of the same type circuit,one would be very foolish to even begin to push the limits in the spec sheets (the chip maker has far more intrinsic knowledge of their products than any of us will ever have). By ignoring the maximum ratings, one could go bankrupt in less than a month after final production.

Jarhead
In looking at your graph, I see voltage and time axises. How did you determine the current pulses ? To reach 10 amps in 4 nanos would equate to a Di/Dt of 2500 amps per microsecond--a fantastic rate of rise. Any reactance present in the circuit- any at all-would severly limit this spec.

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: "Perfect" LED current limiting challenge

Post by Chris Smith » Wed Jan 18, 2006 5:35 pm

Robert

I agree for the most part.

However as experimenters we don’t follow the rules, we push them to find out why.

One bright author said the manufacture suggestions were there simply to break.

For the price of a Led, what’s not to experiment with?

I remember one great author stating His design of a Led Driver circuit and receiver with lens, was good for a "few inches".

I stopped experimenting when I got 40 feet out of the circuit.

I dare say I learned far more from these experiments than the author ever intended, and definately more than the average learner.

And my risk was a whole dollar with of chips.

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: "Perfect" LED current limiting challenge

Post by MrAl » Thu Jan 19, 2006 4:24 pm

Hello again,


Most of us here believe that the efficiency of
the LED does not go up...not that the instantaneous
brighness doesnt go up.
We believe that the brightness goes up, but only
for the pulse period. Averaged out using the
same average current (pulse vs constant I) we
believe the light output goes down simply because
efficiency goes down with increased current.
Also, since the eye integrates the light output
any light appearing to the human eye will be less,
and this is easily proved by using two similar
LED's and driving one with a pulse and one with
a constant current. The one with the constant
current (who's current is equal to the average
value of the pulse current) will be brighter
every time.

For the mathematical view, the Intensity vs
Drive Current curve is monotonic, and its slope
DECREASES with increasing drive current.
Read: *DECREASES* and NEVER *increases* -- NEVER.
In other words, the 'relative' light output does
not increase, but decreases, and by quite a bit.
Put another way, if you connected a whole bunch
of LED's that run at 20ma each so that the whole
bunch draws 10 amps*Duty Cycle, and compare the
light output to a single LED driven with a
repeating pulse with a peak of 10 amps, the
bunch of LEDs will put out *MUCH MORE* light
because the eff of this group is much higher
than the single pulsed LED.
(Note in the above the 'Duty Cycle' is that of
the pulsed LED and is defined as:
D.C.=ontime/totaltime). Also, please dont bring
extraneous definitions of Duty Cycle into the
discussion because it will only confuse the
issue. We all know and use this definition so
please, please, use this one...TIA !!

One other note brought up by Robert Reed is that
of reliability, which is also a key issue here
if we were to pulse LED's at high currents.
This is very important...we dont want to have to
replace the LED every three days...right?
Therefore, before i would use one in a design and
hand it to a customer i would have to have before
me a study that shows how life expectancy doesnt
change that much with a pulse current as compared
to an average constant current, or is at least
comparable to the amount of light output...even
if the light was two times brigher on average!
A study i read a while back (perhaps 2 years)
showed that the life expectancy goes down markedly
with increased current, although they were only
testing with constant currents within manufacturers
specs. If the LED gets hotter even for a short time
(pulsed) this could mean the LED ages much faster
even though the pulse is short. Of course i cant
verify this, but without proof to the contrary i
couldnt depend on this technique when a million
dollar (plus) contract is at stake...could you?
In other words, if you design a circuit like this
and the parts burn out week after week you end up
paying for the all the repairs related to this
issue for the next ten years...including labor.
Also, if it was a showcase project you *DONT* get
the bigger contract for your company.


Take care,
Al
LEDs vs Bulbs, LEDs are winning.

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: "Perfect" LED current limiting challenge

Post by Chris Smith » Thu Jan 19, 2006 5:00 pm

Al

Efficiency is a relative term.

For less amps over all, you can gain 400 or more times the brightness or luminosity from a led, thus this is “more efficient” especially to the eye.

And because the human eye is doing this interpretation, that’s all that counts in terms of the Eye.

This is called perceived efficiency.

And with fiber optics a lowly 20 milliamp green Led can not signal down hundreds of yards of fiber optics and still have the efficiency to trigger the receiver at the other end like a 10 amp pulse can, thus its also more efficient. Actual Efficiency.

Less over all amp draw is also, more efficient. More for less.

On the basis of less than 20 ma over all and 400 times the luminosity, yes it is more efficient.


I know my green Led, drawing less over all current produced the quantity and brightness of a four cell mag light, and If I spend less to get more [eye] perceived light, Then I am getting more for my money and its more efficient.

Considering the human eye perceives a mere 40 to 42 cycles as being a constant light source, the human Eye would be wasting any pulses past that number.

This gives you two choices. You can either drop the total quantity of pulses down to these numbers and use even less over all current, thus being even more efficient, and eye will be quite happy thinking it is a solid light.

Or you can use the greater off times to bump up the amperage, still keeping it well below the 20 milliamp over all current, and get a even brighter out put from the led with in limits.

As to how bright is a led with 10 amp short pulses? I wont stare it directly in the eye.

I will stare any 20 mill green DC led in the eye however.

As to burning out in three days, well that’s pure speculation not supported by any facts or even any experience from those here on this board. In fact so far, other than the actual experimenter, all have gotten these simple and very old facts quite wrong.

It cant be done still rings quite hallow.

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: "Perfect" LED current limiting challenge

Post by MrAl » Thu Jan 19, 2006 6:28 pm

Hello again,


First of all, 'efficiency' *IS NOT* a relative term...it
*CAN BE USED* as a relative term, as in *YOUR* context.
In *MY* context, i am *NOT* using it as a relative term.
I am, however, using it the way engineers and scientists
talk about the ratio of output energy to input energy so
they can better understand the operating characteristics
of a device.
Without stretching the English definition of 'efficiency'
i believe you can say that in a pulse communication system
(that needs pulses anyway) that a pulsed LED works more
efficient than non pulsed (after all you cant even use
a constant current can you), but a better word would be
more 'effective'. If you say 'more efficient than a
constant current LED', then yeah, since a constant current
LED cant convey ANY information i guess it sure is more
efficient: infinitely more efficient :) which tells you
something is wrong with that definition.
The pulsed LED still does not put out
more light relative to its current than a constant current
LED. You can say 'efficient' if you want to, but you'll
only confuse people when the discussion is more about
power input and output, and i dont believe anyone else
on this board was originally talking about power pulsing
from a communication standpoint, but rather from a
power and lighting standpoint--where total power in and
total light power out are the main concerns. You can also
possibly say that the total communication power is better,
but that's still not the same thing.

Secondly, when i did my own investigation i *ALREADY CONSIDERED*
the response to the human eye, so there is nothing else to
consider...is this clear now? In other words, my *input* was
current and my *output* was the *HUMAN BRAINS CONCEPT* of how
much light was perceived -- not how much light was output --
although total light output was an intermediate calculation.
In other words, how much light the person 'felt' was there.

Yeah you're right...i made a mistake when i said 'three days'
when i meant to say 'four days' :)
The main issue there is that we dont know what the reliability
is so it's hard to make assumptions about how good (or bad)
a final design will be for practical use.

Finally, no one here said you cant *pulse* an LED, they just
said the efficiency goes down. And please dont say again that
the efficiency goes up because it 'works' better in communication,
because that's not what we are talking about here. We are
talking about getting more total light output from an LED.
I've pulsed LED's (seven segment LEDs) in display circuits
(long time ago) using short pulses and higher currents because
it allowed me to 'multiplex' the display and thus use a much
simpler drive circuit and interconnection wiring, but i *STILL*
wont claim, "it's more efficient to multiplex LED's", in the
middle of a discussion about raising the efficiency of LED's
because
that's very confusing to other engineers that have a working
knowledge of circuitry and components...it immediately leads
to a response like, "Oh wow, did he/she figure out how to
boost the efficiency of an LED ?" (meaning more light
output for less current input). Multiplexing LED's is a very
specific application whereas boosting the efficiency of
LED's is a very broad subject.
Another way of saying this is: if you could boost the efficiency
of a generic LED it would work better in *EVERY* application.
This includes multiplexing, communications, and general lighting.

So the next time somebody asks you, "How much more efficient
is a pulsed LED over a constant current driven LED for use in
communications?", you can reply:

FIRST CHOICE:
"Infinitely", and then go claim your Nobel Prize, or...

SECOND CHOICE:
You can take my advice on talking about efficiency :)


Take care,
Al
LEDs vs Bulbs, LEDs are winning.

User avatar
MicroRem
Posts: 205
Joined: Tue Dec 04, 2001 1:01 am
Location: Vancouver, WA USA
Contact:

Re: "Perfect" LED current limiting challenge

Post by MicroRem » Thu Jan 19, 2006 8:56 pm

Mr Al is my hero
Mr Chris is not

cheers

Tom

Jarhead
Posts: 13
Joined: Sun Jan 01, 2006 1:01 am
Location: Molalla, OR
Contact:

Re: "Perfect" LED current limiting challenge

Post by Jarhead » Thu Jan 19, 2006 11:53 pm

Chris:

When I have a constantly on LED, consuming X watts, and this pulsed LED, consuming X watts (not counting losses in the drive circuit), the intensity of the pulse is brighter, but I cannot see the brief pulse and my eye does an averaging or integrating function on it.

My test equipment obviously shows a brighter pulse for the brief period it is there.

Setting the two on the bench, to shine at a wall, the pulsed LED, when consuming the same amount of average power, is considerably dimmer.

Please stop twisting my words and trying to pull out stuff that isn't there. The LEDs are definitely less efficient when pulsed at high currents.


ROBERT REED
"Jarhead
In looking at your graph, I see voltage and time axises. How did you determine the current pulses ? To reach 10 amps in 4 nanos would equate to a Di/Dt of 2500 amps per microsecond--a fantastic rate of rise. Any reactance present in the circuit- any at all-would severly limit this spec."

Robert-
- When one is experienced, one knows that having a copper plane in the immediate vincinity (say 2 mils away), from a conductor, severely lowers the conductor inductance. In reality, you do not even need to ground that plane, as the induced "eddy currents" in the nearby copper plane create a nearly equal and opposite field that opposes the one formed by the wire, which seriously acts as an inductor at these speeds. A one inch wire in space easily slowed my pulse edge down to 80nS.

- You can also use multiple paths, putting the traces/wires in parallel, and when you parallel two inductors, the inductance is decreased by half. I had more than two paths on purpose. One on each side of the LED, as well as another through the 0.31" PCB on the backside, which contained four copper planes, both outer layers 2 oz. copper, and two inner layers of 1 oz. copper.

- Combining both techniques really helps alot to reduce the inductance *quite* significantly.

- A series of 0603 sense resistors were put in parallel, with ground being brought across the resistor, right up to and almost touching the other end, to reduce the inductance of the resistor. Sense resistors(inductors) in parallel reduce the remaining inductance. One idea I have not had a chance to try, is to flip the SMT resistors upside down on purpose, to get the current path even closer to the copper plane.

- Current was basically measured across the sense resistors.

- I was quite amazed by the SOT-23 MOSFET being able to handle this amazing abuse. The gate of the MOSFET was driven with a "true" 9A gate driver, that uses both transistors and MOSFETs paralleled in the output stage. This combined approach gives you both speed and high current drive ability. Similar inductance reducing techniques, and capacitors placed right upon the chip power pins as they enter the device body helped to sweak every bit of edge speed from the MSOP-8 MOSFET driver chip.

- The MOSFET used as a gate threshold between 0.5V and 1V. The part is rated for Vgs operation of 2.5V, and has 150pf of input capacitance, and 45pf of gate to drain capacitance. I drove the gate with 15V (20V is the rated breakdown point).

- For this sort of rate of change, you create **absolutely** the smallest possible loop. This also helps reduce inductances. This took multiple iterations to achieve the smallest possible loop I could imagine.

*FYI, I can make the ceramic capacitor bank (yes, multiple ceramics, with most of the body over the ground plane in order to also reduce inductance here) actually SING *very nicely*, due to their piezoelectric properties, with the high dv/dt and di/dt rates present. X7R ceramics were used, since X5R and especially Y5V capacitors, in large values, actually reduce their capacitance with DC voltages on them.

- The ceramic caps were strategically located in immediate vicinity of the current loops, again to keep the current loops absolutely as short as possible. More than enough capacitance was utilized to essentially locally source the current needed during the pulse period.

Also, another technique to see things at these speeds, a fella has to remove the plastic o-scope probe shield on the tip, and provide a very short path from the ground that is brought right up to the tip, to the local ground in the circuit, or you see things that are not really there, being generated in the probe, due to the long ground lead that some folks are fond of. Some probes have a tip that accepts a little spring for this ground you find up at the tip of the probe. It is a great idea to use these in fast measurements.

ian
Posts: 251
Joined: Mon Feb 25, 2002 1:01 am
Location: toronto
Contact:

Re: "Perfect" LED current limiting challenge

Post by ian » Fri Jan 20, 2006 5:10 am

And now for something completely different......
ON topic!

I'll be experementing over the next few days with inductors and PWM, I'll be improving the design I used earlier and I'll post whatever results I see.

Also, I started this post with.....
"Here's a challenge for you hacks........."
Well, let me just say........... well, what CAN I say?

<small>[ January 20, 2006, 05:10 AM: Message edited by: ian ]</small>

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: "Perfect" LED current limiting challenge

Post by Chris Smith » Fri Jan 20, 2006 9:34 am

Jar head, mine was brighter, not dimmer.

Mine was too bright to stare at, and illuminated a area of over fifty feet as bright as a mag light.

I also used a simple lens and reflector to gather and focus all the light available to a usable area. This produced a light similar to a focus-able mag light, and just as bright but in green light instead of white or whitish blue.

Considering green at the same intensity is not as usable to the human eye, this made it more efficient or luminous than the Mag light.

From the stand point that a DC current puts our a long and continue stream of photons, more so than the pulsed version, yes you are right. But if you cant utilize them to gain more out of them, then your wrong. Scientifically or From the point of the eye.

So, which is more useful?

Useful defines both effective and efficient.

The pulse is more efficient and more effective to electronics, and to the human eye.

Its appears to be brighter, illuminates more area, and goes further down a light pipe than the DC’s larger mass of photons.

Why, because like a laser they are better organized into larger packets, ones that become more useful.

If you fired a BB gun at a brick wall a million times over, it might take you several years to wear down a hole through the bricks.

If you fire a single 50 caliber round at the wall, the wall WILL have a hole in it after one round.

The larger round contains about a million BB worth of kinetic energy, and thus it is more effective and more efficient than trying to “wear down” the wall.

It is more useful, efficient, and effective.

Total light out put as far as more photons that are weaker have little use to the eye or communications.

Smaller packets of more efficient or more effective photons do.

Both to the eye, and the fiber optics.

If your ever stuck on a raft and you try to signal a 747 at 50,000 feet with a one cell flash light, good luck.

Yet the same flash light in the strobe operation can easily be seen at 20 miles or more because the rearrangement of photons from the strobe light are such that they are in packets exactly like the pulsed led. Brighter, stronger, yet less overall mass of photons. And the battery last longer too.

Merely having more photons has little use in science or the real world, making DC less effective, less efficient, and less useful.

So are you defining “boosting efficiency or boosting effectiveness” as being a term to mean less useful in the real world of science or the eye?

Pulsing them with out a doubt makes them more useful, and thus more effective.

And because the original author was terrible at asking anything out of his post, no one here knows what criteria he actually had in mind. But, from the eye point of view, and from the signal point of view, both are more Effective and more Efficient.

As to Do we know how long the led will last?
40 years worth of history would disagree with your four day guess.

<small>[ January 20, 2006, 08:21 PM: Message edited by: Chris Smith ]</small>

Robert Reed
Posts: 2276
Joined: Wed Nov 24, 2004 1:01 am
Location: ASHTABULA,OHIO
Contact:

Re: "Perfect" LED current limiting challenge

Post by Robert Reed » Fri Jan 20, 2006 5:36 pm

Jar head
Now those are what I call replies! Not only did you give a detailed description of your results, but you also followed up with a detailed description of your test set up. In reading over these, I would have to say their isn't any room for argument here as I understood to the letter where you wre going with this. I have done some work with conductor over the ground plane in circuit building which in my case was viewed as transmission line theory i.e.--Any given size conductor at any given distance above a ground plane will have a given Zo. I always applied this to 50 ohm R.F. circuitry for fairly long hauls on board and it works quite well.

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: "Perfect" LED current limiting challenge

Post by Chris Smith » Fri Jan 20, 2006 5:58 pm

Jarhead has very good results, however, he is just starting where other left off more than two and a half decades back, and four if you count the original authors.

My suggestions would be to adjust many of your parameter to achieve the brightest output possible.

1] Decrease the pulse rate, increase the voltage. [more amps]

2] Increase the pulse width up to 200 ns max. Try several settings.

3] Repeat number one after this.

4] Play will as many repetition rates as you can adjust your input to.

5] try every combination of the above with different Leds, and alternate color leds.

But don’t try these on the White Led as its characteristics are not the same as the standard led, the capacitance is greater, and there are other problems associated with these LEDs.

Did you incorporate any protection diodes in your circuit?

Most other laser plans incorporate a very fast recovery diode at the feed [+] of the Laser or Led. Stops ringing which reduces effectivness and burn outs.

Mainly,..This stops ringing in the leads which burns out the emitter.

Short leads alway good.
Ground plane also good.
Keep Going, its all old but its a lost art.

Jarhead
Posts: 13
Joined: Sun Jan 01, 2006 1:01 am
Location: Molalla, OR
Contact:

Re: "Perfect" LED current limiting challenge

Post by Jarhead » Fri Jan 20, 2006 7:16 pm

Chris,

You are definitely digging yourself a nice hole from my point of view.

Could you please provide this LED pulsing paper you are referring to?

I know I've already provided some, showing the decrease in efficiency when an LED is pulsed at higher currents.

I've also done the experiement, and shinning them at the wall, the 4nS pulse occuring at 2.43KHz repetition rate, shows a considerably lower output than another LED driven at the same input power consumption level.

Remember, the human eye integrates light over a period of time.

Not being versed in the full blown analysis of the human eye, lets just pick a number here.

Lets say the human eye integrates light over a 30 Hz rate. (yes I know it varies over various lighting conditions and such, lets keep things real simple here.)

Now, I come along and hit the human eye with a four nanosecond pulse. Lets say the amplitude is 100. The eye is integrating all the light received over 1/30th of a second (30Hz), and thus it is going to accumulate light for 33 milliseconds. For well over 32 milliseconds, the eye is not recieving any light at all.

This pulse will not look like 100, as for the majority of the time it will not be recieving any light at all. So, how bright would it be? Well, lets start off with a simple look at things. Take 4 nanoseconds, and divide it by 33 milliseconds. We get 1.21e-7. Then take the pulse amplitude and multiply it by this number, you end up with 1.21e-5. So the output is quite dim.

Now, if you know just a little bit about the human eye, you will also know that the human eye does not have a linear response curve. What the eye sees as the difference from 1 mW to 2mW, is not at all the same as what it sees as the difference from 1W to 2W. The apparent intensity difference is essentially compressed at the higher power levels. We would need to stack these factors on top of the uber-bright pulse, and the apparent intensity actually decreases as the human eye is hit harder and harder with high intensity pulses.

From actual observations, and building up a circuit to create very well shaped pulses in the range of time you described, and looking at the light, comparing with a 100% duty cycle, constant on, much lower current, the pulsing technique definitely falls incredibly short on performance.

So, please, provide reference papers for pulsing LEDs and making them appear much brighter to the human eye, to get more light to back your claims. Please don't provide a bunch of hype that is propagated by a bunch of old wives tales from hobbyists sites.

User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: "Perfect" LED current limiting challenge

Post by Chris Smith » Fri Jan 20, 2006 8:39 pm

My papers were from the 80s, so the answer is no, I cant provide them, and We didnt have a WEB.

Having moved across the state from top to bottom more than 5 times since this date, I don’t even have one quarter of my library or files remaining.

The eye accumulates light as a steady source, after 42 cycles, so more cycles don’t mean more perception.

You have shown one example, at a rate that is far higher than necessary, and your limits only restrict you and the limits that can be achieved.

You are merely at the starting point, claiming defeat already.

So what was your experience with light before this, and I ask, because you don’t seem to grasp much of what light really is or what the led can do.

But you do have the equipment to find out.

The general information,... you do have,..... now proceed until you gain more luminosity.
[As this experiment only,.. requires]

This is but one gain from this lesson, while the fiber optics usage requires another.

Im not digging any holes, merely repeating history, while you have yet to challenge what has already been done fully.

Unlike you, I didn’t have access to your kind of accurate O-scopes at random in the 80s, and couldn’t check my timing down to the resolution you have shown.

Regardless of fancy equipment, I continued to attain my goals and succeeded because it has already been done, even back then.

Jarhead
Posts: 13
Joined: Sun Jan 01, 2006 1:01 am
Location: Molalla, OR
Contact:

Re: "Perfect" LED current limiting challenge

Post by Jarhead » Fri Jan 20, 2006 9:56 pm

Okay Chris, you don't seem to want to repent.

There is one fella that many electronics and optical experimenters often revere, and is a bountiful source of information for those starting out...

Don Klipstein:

"Many people have noticed that some LEDs, including most of the multi-LED moving display signs and most red LED digital readouts, are fed pulsed current instead of steady DC.

There is a common belief that this increases apparant brightness by taking advantage of nonlinearities in human vision. It is widely believed that apparant brightness of a pulsed LED is more determined by peak actual brightness rather than average actual brightness even if the pulse rate is fast enough for the LED to appear steadily lit. There is only a small grain of truth to this belief, and there is a different explanation why LEDs in red 7-segment readouts and many moving display signs work more efficiently when pulsed.

I have done lots of experimentation with 555 oscillators and various LEDs and solar cells in various room lighting conditions. Regardless of frequency and duty cycle, a pulsed LED that gives the same solar cell response as a continuously operated one of the same model will almost always match the visually apparant brightness of the continuously operated one. I have sometimes seen slight gains - less than 10 percent when I get any gain at all - in apparant brightness if the pulse frequency is barely fast enough to make the LED not visibly flicker. This frequency is usually around 60 Hz.

One thing many people don't realize is that most LEDs are not linear as light emitters. I have even seen one book stating falsely that they are linear as light emitters. Although photovoltaic detection is usually reliably linear, photovoltaic emission is a different story. Note the low quantum efficiency for emission and high quantum efficiency for detection (in a favored band) by most LEDs.

Human vision is nonlinear, but that nonlinearity is after a surprisingly accurate time-integration process. When a light is flashing rapidly enough to appear continuously on without flicker, what you see has a good correlation (although nonlinear) with average brightness and is surprisingly independent of peak brightness.

Most LEDs have maximum efficiency at currents near or somewhat over 20 mA. There are some exceptions - gallium phosphide red (low current red or "697 nM" red), old type silicon carbide blue, and most InGaN/GaN blue, blue-green, green, and white LEDs. These specific types are most efficient at lower currents."
http://members.misty.com/don/ledp.html

Towards the end, he does mention one exception, there are just a few LED types that are not running at their peak efficiency points, when run at their rated 15mA.

"Pulsing is in your favor if the LED is a type that has maximum efficiency at higher current and the average current must be low. LEDs that have efficiency maximized at currents 15 mA or more are GaP other than low current red, GaAsP, GaAlAsP ("super bright" and "ultrabright" red) and Hewlett Packard's similar "T.S. AlGaAs", and InGaAlP. Colors of these types are red through yellowish green, excluding gallium phosphide red."

Running them a little higher is quite different than blasting them with tremendous amounts of power.

Though this doesn't specifically cover the situation you claim, it does shed some light on the discussion here.

Jarhead
Posts: 13
Joined: Sun Jan 01, 2006 1:01 am
Location: Molalla, OR
Contact:

Re: "Perfect" LED current limiting challenge

Post by Jarhead » Fri Jan 20, 2006 10:08 pm

Sorry, I forgot to answer some of your other questions.

A piece of my job takes me into areas that are related to this article:
http://neasia.nikkeibp.com/neasiaarchivedetail/000503

I work as a Senior Avionics Design Engineer, doing EE work, but it takes me into many other disciplines, in order to take advantage of everything available.

Chris, as I answered your questions, could you please tell me exactly what you do for employment?


On a lighter note, I tried the same methods on some higher power devices (1 Watt LEDs), beefing things up further to handle the additional demands of the larger dies, and their higher associated capacitance and such.

Here we have the current pulse Mr. Reed was curious about:
<img src="http://www.molalla.net/~leeper/ledcur~1.jpg" alt=" - " />

And here is the voltage pulse:
<img src="http://www.molalla.net/~leeper/ledvol~1.jpg" alt=" - " />

Unfortunately, I could not get my hands on equipment for measuring very fast light pulses today, so I had to make due with what I had available. I do in fact know the light pulse edges are considerably faster than shown here:
<img src="http://www.molalla.net/~leeper/ledlig~1.jpg" alt=" - " />

Chris, I have already tried many different duty cycles, pulse repetition frequencies, pulse widths, and a combination of all these together. The circuit allows me to simply control the current as well as the pulse width, so it was very simple to do what you recommended already.

I however have not had the time to try a vast assortment of LEDs.

Chris- I'd like to envite you to my hobbyist ftp space, where I keep photos containing various pieces of information that some may find useful. Most of the photos there are for explaining various ideas, concepts, and little tests I've ran for folks. BTW, you will find just a few photos from other hobbies.
http://www.molalla.net/~leeper

Enjoy!

<small>[ January 20, 2006, 10:56 PM: Message edited by: Jarhead ]</small>

Post Reply

Who is online

Users browsing this forum: No registered users and 8 guests