1. Home
  2. Electronics Design
  3. LED driver

LED driver

This is a single-cell, single inductor, single transistor LED driver. The values are all my first wild guesses, none tuned or tweaked, so it could be improved.

Efficiency is 70% with 1.6 mW into the LED.

Version 4 SHEET 1 880 680 WIRE -144 64 -208 64 WIRE -16 64 -64 64 WIRE 80 64 -16 64 WIRE 240 64 160 64 WIRE 336 64 240 64 WIRE 400 64 336 64 WIRE 480 64 400 64 WIRE 336 112 336 64 WIRE 480 112 480 64 WIRE -16 128 -16 64 WIRE -208 176 -208 64 WIRE 240 208 240 64 WIRE 480 240 480 176 WIRE -16 256 -16 208 WIRE 64 256 -16 256 WIRE 176 256 64 256 WIRE -208 304 -208 256 WIRE 576 304 480 304 WIRE 672 304 576 304 WIRE 240 320 240 304 WIRE 288 320 240 320 WIRE 336 320 336 176 WIRE 336 320 288 320 WIRE -16 336 -16 256 WIRE 336 336 336 320 WIRE 480 336 480 304 WIRE 672 336 672 304 WIRE 240 352 240 320 WIRE -16 464 -16 400 WIRE 240 464 240 432 WIRE 336 464 336 400 WIRE 480 464 480 400 WIRE 672 464 672 416 FLAG -16 464 0 FLAG 336 464 0 FLAG -208 304 0 FLAG 240 464 0 FLAG 480 240 0 FLAG 480 464 0 FLAG 672 464 0 FLAG 576 304 Led_Test FLAG 400 64 A FLAG 64 256 B FLAG 288 320 E SYMBOL LED 464 112 R0 WINDOW 0 39 77 Left 2 WINDOW 3 32 106 Left 2 SYMATTR InstName D1 SYMATTR Value NSCW100 SYMBOL ind 64 80 R270 WINDOW 0 -32 56 VTop 2 WINDOW 3 -38 55 VBottom 2 SYMATTR InstName L1

SYMBOL npn 176 208 R0 WINDOW 0 -54 25 Left 2 WINDOW 3 -90 73 Left 2 SYMATTR InstName Q1 SYMATTR Value BC547C SYMBOL cap 320 112 R0 WINDOW 0 52 26 Left 2 WINDOW 3 48 60 Left 2 SYMATTR InstName C1 SYMATTR Value 33n SYMBOL cap 320 336 R0 WINDOW 0 49 23 Left 2 WINDOW 3 42 59 Left 2 SYMATTR InstName C2 SYMATTR Value 300n SYMBOL cap -32 336 R0 WINDOW 0 52 9 Left 2 WINDOW 3 53 44 Left 2 SYMATTR InstName C3

SYMBOL res -32 112 R0 WINDOW 0 53 44 Left 2 WINDOW 3 44 79 Left 2 SYMATTR InstName R1 SYMATTR Value 10K SYMBOL voltage -208 160 R0 WINDOW 0 54 44 Left 2 WINDOW 3 52 79 Left 2 SYMATTR InstName V1 SYMATTR Value 1.5 SYMBOL res 224 336 R0 WINDOW 0 -51 43 Left 2 WINDOW 3 -48 74 Left 2 SYMATTR InstName R2 SYMATTR Value 50 SYMBOL LED 464 336 R0 WINDOW 0 36 77 Left 2 WINDOW 3 27 103 Left 2 SYMATTR InstName D2 SYMATTR Value NSCW100 SYMBOL current 672 416 R180 WINDOW 0 -69 55 Left 2 WINDOW 3 -75 19 Left 2 SYMATTR InstName I1 SYMATTR Value 1m SYMBOL res -160 80 R270 WINDOW 0 -39 60 VTop 2 WINDOW 3 -47 55 VBottom 2 SYMATTR InstName R3

TEXT 664 184 Left 2 !.tran 100m TEXT 608 88 Left 2 ;Single-Cell LED driver TEXT 616 128 Left 2 ;J Larkin Jan 29 2020

--

John Larkin         Highland Technology, Inc 

The cork popped merrily, and Lord Peter rose to his feet.  
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"
Reply to
jlarkin
Loading thread data ...

snipped-for-privacy@highlandsniptechnology.com wrote:

Wierd oscillator. You have ignored the fact LTspice sets inductor ESR to

1e-3 ohms when you neglect to define it. This raises the unloaded Q to 83,000.

What is strange is the oscillator runs along at zero amplitude, then suddenly jumps to maximum amplitude in a few cycles. That is normally impossible with a Q that high.

Another wierdness is if you do not define the time to start saving data and maximum time step, the oscillator waits for 78 msec before it starts oscillating.

However, if you set the stop time to 5msec and the maximum time step to

10us, the oscillator starts at 960 us.

I don't think I have ever seen another oscillator like this.

Watch the wrap.

Version 4 SHEET 1 1056 712 WIRE -16 64 -208 64 WIRE 80 64 -16 64 WIRE 240 64 160 64 WIRE 272 64 240 64 WIRE 336 64 272 64 WIRE 448 64 336 64 WIRE 336 112 336 64 WIRE 448 112 448 64 WIRE -16 128 -16 64 WIRE -208 176 -208 64 WIRE 240 208 240 64 WIRE 448 240 448 176 WIRE -16 256 -16 208 WIRE 48 256 -16 256 WIRE 176 256 48 256 WIRE -208 304 -208 256 WIRE 240 320 240 304 WIRE 288 320 240 320 WIRE 336 320 336 176 WIRE 336 320 288 320 WIRE -16 336 -16 256 WIRE 336 336 336 320 WIRE 240 352 240 320 WIRE -16 464 -16 400 WIRE 240 464 240 432 WIRE 336 464 336 400 FLAG -16 464 0 FLAG 336 464 0 FLAG -208 304 0 FLAG 240 464 0 FLAG 448 240 0 FLAG 272 64 Q1C FLAG 288 320 Q1E FLAG 48 256 Q1B SYMBOL LED 432 112 R0 WINDOW 0 39 77 Left 2 WINDOW 3 32 106 Left 2 SYMATTR InstName D1 SYMATTR Value NSCW100 SYMBOL ind 64 80 R270 WINDOW 0 -32 56 VTop 2 WINDOW 3 -38 55 VBottom 2 SYMATTR InstName L1 SYMATTR Value 200u SYMBOL npn 176 208 R0 WINDOW 0 -54 25 Left 2 WINDOW 3 -90 73 Left 2 SYMATTR InstName Q1 SYMATTR Value BC547C SYMBOL cap 320 112 R0 WINDOW 0 52 26 Left 2 WINDOW 3 48 60 Left 2 SYMATTR InstName C1 SYMATTR Value 33n SYMBOL cap 320 336 R0 WINDOW 0 49 23 Left 2 WINDOW 3 42 59 Left 2 SYMATTR InstName C2 SYMATTR Value 300n SYMBOL cap -32 336 R0 WINDOW 0 52 9 Left 2 WINDOW 3 53 44 Left 2 SYMATTR InstName C3 SYMATTR Value 1u SYMBOL res -32 112 R0 WINDOW 0 53 44 Left 2 WINDOW 3 44 79 Left 2 SYMATTR InstName R1 SYMATTR Value 10K SYMBOL voltage -208 160 R0 WINDOW 0 54 44 Left 2 WINDOW 3 52 79 Left 2 SYMATTR InstName V1 SYMATTR Value 1.5 SYMBOL res 224 336 R0 WINDOW 0 -51 43 Left 2 WINDOW 3 -48 74 Left 2 SYMATTR InstName R2 SYMATTR Value 50 TEXT -168 -64 Left 2 ;'Single-Cell LED driver TEXT 56 -8 Left 2 ;J Larkin Jan 29 2020 TEXT -136 520 Left 2 ;CT = (33*300) / (33 + 300) = 29.e-9\n \nFc = 1 / (2 * pi * sqrt(200e-6 * 29e-9)) = 66,085\n \nXL = 2 * pi * 200e-6 * 66,085 = 83 \n \nQ = 83 / 1e-3 = 83,000\n \nXC2 = 1 / (2 * pi * 66,085 * 300e-9) =

8.02 TEXT -168 0 Left 2 ;.tran 0 5m 0 10u TEXT -168 32 Left 2 ;.tran 0 100m TEXT -168 -32 Left 2 !.tran 0 5m 0 10u
Reply to
Steve Wilson

It's a Colpitts

Reply to
bitrex

"unknown schematic syntax" : pi * sqrt(200e-6 * 29e-9)) = 66,085\n \nXL = 2 * pi * 200e-6 * 66,085 = 83 \n \nQ = 83 / 1e-3 = 83,000\n \nXC2 = 1 / (2 * pi * 66,085 * 300e-9) =

8.02
Reply to
Robert Baer

snipped-for-privacy@highlandsniptechnology.com wrote in news: snipped-for-privacy@4ax.com:

Make me a LED amp driver with precision current adjustment in 50 steps via two buttons. But make it for the high wattage units. So vary from between your 1.6 mW and say 2 W into the load (LED). Or even bigger for balanced LED arrays. Hey!

I know... Make a lab grade bench device that has current readout to say 3.5 digits for testing LEDs with a small log that can be plotted to show failure events in order to characterize a given batch of LEDs. For say matching and culling and proofing of a vendor's claims of performance.

Reply to
DecadentLinuxUserNumeroUno

He *did* say watch the wrap (which you cut). I got the same message then unwrapped the lines and it runs fine.

Reply to
John S

Make me an LED driver that can pulse a 3-volt theatre LED to 350A at 20V (7kW). Make me an LED driver that can pulse a 30V 3.3A 100W COB LED array to 33A (2kW).

--
 Thanks, 
    - Win
Reply to
Winfield Hill
[Snip...]

Well, mind the wrap!

Regarding the startup of the thing, it's basically a common base LC oscillator, so you could add 680nF across R1. If you also replace V1 by a pulsed source, it will start instantly, without relying on uncertain numerical noise.

The associated reduction of ripple current in the power supply is a bonus.

Jeroen Belleman

Reply to
Jeroen Belleman

Really? How do you define ripple current? I'm seeing a sine waveform of

+20mA peak and -16mA peak. The average is 1.36mA and RMS is 13mA. It seems to be a lot of ripple current to me.
Reply to
John S

Indeed, you are correct. I missed the vertical plot scale change caused by my modifications. Sorry about that.

Jeroen Belleman

Reply to
Jeroen Belleman

on"

Definitely scraping the bottom of the topics barrel with this one.

Reply to
bloggs.fredbloggs.fred

[...]

Cheap shot that contributes nothing. Your opinions don't matter.

Jeroen Belleman

Reply to
Jeroen Belleman

Then it's a good thing it's loaded.

Not strange, obvious. I knew it would do that.

Ditto.

But it's just a Colpitts.

--

John Larkin         Highland Technology, Inc 

The cork popped merrily, and Lord Peter rose to his feet.  
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"
Reply to
jlarkin

Boring... Make it flash with just one transistor.

A relaxation/squegging Hartley, the second inductor can be an adjacent moulded inductor - no need for fancy transformers or dual inductors. This circuit predates the ca 1976 LM3909.

Simulation runs better with a noise starter V2.

Version 4 SHEET 1 880 680 WIRE -64 -288 -208 -288 WIRE 240 -288 -64 -288 WIRE -64 -256 -64 -288 WIRE -64 -160 -64 -176 WIRE 240 -96 240 -288 WIRE -64 -48 -64 -80 WIRE 176 -48 -64 -48 WIRE -64 -32 -64 -48 WIRE -64 48 -64 32 WIRE 16 48 -64 48 WIRE 16 80 16 48 WIRE -208 176 -208 -288 WIRE -64 192 -64 48 WIRE 16 192 16 160 WIRE 240 192 240 0 WIRE 240 192 16 192 WIRE 16 240 16 192 WIRE 240 240 240 192 WIRE -208 352 -208 256 WIRE -64 352 -64 256 WIRE -64 352 -208 352 WIRE 16 352 16 320 WIRE 16 352 -64 352 WIRE 240 352 240 304 WIRE 240 352 16 352 WIRE 240 368 240 352 FLAG 240 368 0 SYMBOL LED 256 304 R180 WINDOW 0 24 64 Left 2 WINDOW 3 24 0 Left 2 SYMATTR InstName D1 SYMATTR Value NSCW100 SYMBOL ind2 0 64 R0 SYMATTR InstName L1

SYMATTR Type ind SYMATTR SpiceLine Rser=1 SYMBOL npn 176 -96 R0 WINDOW 0 75 17 Left 2 WINDOW 3 47 49 Left 2 SYMATTR InstName Q1 SYMATTR Value BC547C SYMBOL cap -80 192 R0 WINDOW 0 30 6 Left 2 WINDOW 3 28 59 Left 2 SYMATTR InstName C1 SYMATTR Value 10n SYMBOL cap -80 -32 R0 WINDOW 0 30 7 Left 2 WINDOW 3 30 63 Left 2 SYMATTR InstName C2 SYMATTR Value 100n SYMBOL res -80 -272 R0 WINDOW 0 47 45 Left 2 WINDOW 3 44 79 Left 2 SYMATTR InstName R1 SYMATTR Value 1Meg SYMBOL voltage -208 160 R0 WINDOW 0 54 44 Left 2 WINDOW 3 50 79 Left 2 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value 1.5 SYMBOL ind2 0 224 R0 SYMATTR InstName L2

SYMATTR Type ind SYMATTR SpiceLine Rser=1 SYMBOL voltage -64 -176 R0 WINDOW 3 -261 31 Left 2 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 WINDOW 0 -81 -4 Left 2 SYMATTR Value SINE(0 100n 100k) SYMATTR InstName V2 TEXT 344 280 Left 2 !.tran 500m TEXT 328 -224 Left 2 ;Single-Cell LED flasher TEXT 336 -184 Left 2 ;EPW Jan 2020 TEXT -40 392 Left 2 !K1 L1 L2 0.5

piglet

Reply to
piglet

formatting link

formatting link

formatting link

We don't sell many of this kind of driver, because so many people do it, it's a price war to the bottom.

The D140 was developed to drive a bar-laser pumped YAG laser for a battery-powered military rangefinder, but we got outbid for production.

--

John Larkin         Highland Technology, Inc 

The cork popped merrily, and Lord Peter rose to his feet.  
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"
Reply to
jlarkin

I appoint you a member in full standing of the Crabby Old Gits Who Never Design Stuff And Probably Never Did Very Well Society.

I was peeling shrimp, minding my own business, when this popped up in my head, values and all. The Shrimp and Grits and Abita beer were excellent.

We could discuss Shrimp and Grits but that might be OT.

--

John Larkin         Highland Technology, Inc 

The cork popped merrily, and Lord Peter rose to his feet.  
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"
Reply to
jlarkin

I've done it many times. No apologies needed.

Reply to
John S

The LED current spike seems to average about 25 mA and lasts about 5 microseconds, which won't be very visible. It's more of a blocking oscillator than a squegger. Efficiency looks like about 30%.

--

John Larkin         Highland Technology, Inc 

The cork popped merrily, and Lord Peter rose to his feet.  
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"
Reply to
jlarkin

Yes, Blocking oscillator - thanks that was the description I was looking for.

I may have the component values all wrong, the ones I remember seeing made quite fat light blips of at least tens milliseconds long. There was no explicit C across the tank just stray winding capacity. Crazy sensitive to cell voltage - optimized for 1.35V button cell probably Hg in those days and went crazy above 1.5V. I may have got the LED polarity wrong too - it kinda works either way.

I admired the way to get functionality at extrenme low cost.

piglet

Reply to
Piglet

Is there nothing new under the sun? Well, OK, maximum current so far is 600A vs 250A, and rise-fall time is ~1us, faster than 10us. Plus 100W capability. But this must be an easy-to-copy design. OK, it's in AoE x-Chapters, and new design details are on DropBox.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.