Page 15 of 18

Re: "Perfect" LED current limiting challenge

Posted: Sat Jan 21, 2006 12:50 pm
by philba
Hey Jarhead, this is great info. I (and I think most on here) really appreciate the effort you put into it.

One sad and negative aspect of it, though. The intended recipient doesn't care, he just wants to have the last word.

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 22, 2006 7:54 pm
by Chris Smith
Jar head, having read all of the above authors and articles about pulsing LEDS heavy, many times over, over the years, they have little meaning to me as my experiments and the original authors of the pulse led have already proven wrong.

For starters most of the modern authors are much like Philba, "you cant do it because the manufacture says it cant be done”,... type of attitude. I can only laugh. And they werent doing past the amp for the most part.

Their concept does not include the actual work that the original authors have done from the 60s, nor my experiments, nor any of the successes.

When you tell me to my face that my experiments did not light up a led as bright as a mag light or brighter, then you will be making your real statment known much like the other nay sayers.

You can quit now or proceed, your choice. There is a lot more out there. I have done it and I didnt have access to great scopes at the time for all my experiments. I worked from a light point of view, not from the math. All my variables were adjustable on demand. I used many combinations, many transistors, and many timming circuits, and then I suceeded in getting a lot of light out from the LED, my main goal.

But sorry you cant change history.

As to Philba,.... I know you didn’t write this, but it’s the story of your life...."Pulsing these diodes at 10 or 100 amps is obviously total bullshit" ......and some wonder why they remain just ordinary in this field.

When you quit before you even start, the results and your learning process are always predictable. Zero bites.

Even jar head put you to rest while you claim victory.

<small>[ January 22, 2006, 08:43 PM: Message edited by: Chris Smith ]</small>

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 22, 2006 8:50 pm
by philba
Sheesh chris, you'll say just about anything including fabrications.

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 22, 2006 9:45 pm
by Chris Smith
Philba,...From your point of view, nothing must seem very large?

But then you have done nothing in this field and yet you continue to speak out as if you had a leg to stand on in the first place.

Go back to your back seat and watch others live life and accomplish what you dream cant be done. Jar head has taken it head on, while you still pine. But the back seat of life at least for some, is their whole life?

Jar head, in the June 95 issue of N&V [for starters] you might find some interesting data to help you along.

Anthony Charlton describes a lot of the early pulse laser techniques,.. and the how and why of success.

It might help you where I cant fill in the some of your blanks.

Re: "Perfect" LED current limiting challenge

Posted: Mon Jan 23, 2006 11:17 am
by rshayes
Ian:

You might find this web page (http://sound.westhost.com/appnotes/an003.htm) of interest. It describes a three transistor driver for 350 milliamp LEDs. Operation at battery voltages from about 4 volts to 15 volts is described with efficiency of about 60 percent.

This circuit uses the .7 volt Vbe of a silicon transistor to set the current level. The sampling resistor consumes a fair ammount of power. Replacing Q1 with an NPN germanium transistor and reducing the value of the sampling resistor accordingly should push the efficiency up into the 70 percent range.

Re: "Perfect" LED current limiting challenge

Posted: Mon Jan 23, 2006 11:18 am
by rshayes
Ian:

You might find this web page (http://sound.westhost.com/appnotes/an003.htm) of interest. It describes a three transistor driver for 350 milliamp LEDs. Operation at battery voltages from about 4 volts to 15 volts is described with efficiency of about 60 percent.

This circuit uses the .7 volt Vbe of a silicon transistor to set the current level. The sampling resistor consumes a fair ammount of power. Replacing Q1 with an NPN germanium transistor and reducing the value of the sampling resistor accordingly should push the efficiency up into the 70 percent range.

Re: "Perfect" LED current limiting challenge

Posted: Mon Jan 23, 2006 11:23 am
by rshayes
This was a retyping of the previous post caused by wierd behavior on the part of the bulletin board.

<small>[ January 23, 2006, 01:51 PM: Message edited by: stephen ]</small>

Re: "Perfect" LED current limiting challenge

Posted: Fri Jan 27, 2006 11:49 pm
by MrAl
Hello again,

stephen,

Looks like a fairly simple circuit. I tried
a Zetex PNP and got good results in the simulation.
A little experimentation showed that it's also
probably possible to add a couple resistors and
feed the sense transistor from the voltage of
the LED which would boost efficiency, although
wouldnt be as good for regulating current.

Take care,
Al

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 29, 2006 6:10 am
by jwax
I've grown to hating this thread since it has brought out the worst in some participants, but here's my contribution- a 87% efficient convertor from Maxim-
http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1167

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 29, 2006 6:21 am
by Mike6158
Originally posted by jwax:
I've grown to hating this thread since it has brought out the worst in some participants, but here's my contribution- a 87% efficient convertor from Maxim-
http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1167
Ditto... but we keep coming back... it must be a "car wreck thing"

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 29, 2006 8:28 am
by MrAl
Hi there,

Is kinda ironic isnt it? he he.
Most hated thread yet it's got something like 20
times more replies then any other thread. Wow :)

Take care,
Al

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 29, 2006 9:16 am
by jimandy
Yes, it's like real strong Roquefort cheese. If you get it on your fingers, they smell so bad you can't keep from sniffing 'em.

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 29, 2006 9:42 am
by rshayes
The 87 percent figure for the Maxim 767 is a little deceptive. If you connect it as a current regulator, the feedback voltage from your current sampling resistor equals the reference voltage in the integrated circuit, which is 1.25 volts. The power lost in the sampling resistor will reduce the efficiency to about 64 percent when driving a 3.5 volt LED. If you are driving a red LED (about 1.5 volts), the overall efficiency drops below 50 percent.

If this part was redesigned with a lower reference voltage, the overall efficiecy could probably be brought up to around 80 percent.

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 29, 2006 10:13 am
by Robert Reed
Jimandy
Or that tooth you chipped and no power on earth can keep your tongue from rubbing over it!

Re: "Perfect" LED current limiting challenge

Posted: Sun Jan 29, 2006 2:50 pm
by ian
I've done a lot of testing with various duty cycles, pulse widths, and inductors using some of the "direct drive" circuits. I have a PIC micro driving a FET so I can vary the pulses and make the circuit "ideal". What I find is the savings in current for driving a 3.2V LED is minimal.

I can get 25% savings i.e. under ideal conditions but there are wide current changes depending on voltage or pulse width. That's with 6.4V. Then the voltage drops the current drops FAST! If I use a micro, or additional circuitry to crate the pulse train the savings in current is cancelled out as compared to direct drive.

I guess I could compensate the voltage with the pulse width but it would be a hairy amount of PIC programming.

Still doing testing but it's not looking good.

Ian

<small>[ January 29, 2006, 03:16 PM: Message edited by: ian ]</small>