Current sense and efficient LED driver?

Did you miss this that I posted yesterday...

...Jim Thompson

Like I say, PAL, you have a serious HUB problem >:-}

But, turns out, it isn't needed...

...Jim Thompson

So, PAL with serious HUB syndrome, show us (with a LTspice schematic

...Jim Thompson

I saw that five minutes after I replied ;-)

All that nonsense, to replace one resistor!

Thanks for the laugh.

the

il

ed

om

room

why

ant

t at

mA.

n a

-Lasse

You need a schematic of a few LEDs and a resistor? That's better than your circuit.

Or maybe you can't envision a battery, and LED, and a depletion fet?

Past that, the OP has trickled out some more of his requirements, and it's an

Have fun! In spite of the stones >:-} I think you'll like it! ...Jim Thompson

I do digital. Who is it that reworked ALL of ON Semi's 74HC product when they switched to Chartered as their foundry. It wasn't you, it was me.

What don't you show us some LT schematics that actually work? ...Jim Thompson

I've posted dozens of LT Spice files. Including several switchers.

This particular LED problem is, at present, poorly defined.

It would be fun if he implemented 64, or maybe 192, of your circuits!

On a sunny day (Sun, 09 Feb 2014 11:53:37 -0800) it happened Daniel Pitts wrote in :

Way to complicated, For the OP, who needs RGB drive different voltage LEDs for R.G, and B, perhaps this cicuit makes more sense (and he very likely wants no switcher).

See:

On a sunny day (Sun, 09 Feb 2014 09:02:34 -0800) it happened John Larkin wrote in :

Yes, but BF245 comes as A, B, and C suffix, but indeed they have wide spread. I have used those a lot, but only ever came across B suffix AFAIR.

I have never used those, and still want to put some on the shopping list for experimenting some day. Do the small ones have a low enough current for a LED, like 10 to 20 mA with Vgs at 0V?

Yea, but he has a RGB matrix, switchers can make weird stroboscopic effects, will rectification and need filtering too.

al jfet.

ead.

linearly

for experimenting some day.

ith Vgs at 0V?

ts,

I'd say be the best bet for his size constraint is keeping the TLC5916 he is using already but supply the LEDs with three switchers, one for each color adjusted to the minimum required voltage

-Lasse

The LND150/160 are very consistant around 1.6 mA Idss.

The DN2530 (SOT89) part has high Idss, but a source resistor can set the operating current, pretty repeatability. I have some data on that, somewhere...

I think NXP or someone else makes depletion fets, too.

Right. The actual application is still a bit fuzzy.

Patent pending >:-} ...Jim Thompson

in

pical jfet.

spread.

up linearly

st for experimenting some day.

A with Vgs at 0V?

fects,

hopefully there is more to it than that, though it does seems like you can get a patent for anything these days no matter how obvious, as long as you pay the fee

-Lasse

There are switchmode solutions to LED drive efficiency (see Micrel MIC2287 for one example). There are also moderate-efficiency current mirror solutions (two transistors, base common, with different emitter resistors), that are OK if you start with a voltage source only slightly higher than the forward votage of the LED.

So, if you want to drive a UV, blue, or white LED, from +5V, a simple solution is fairly efficient. If you want to drive an IR LED from that +5V, the efficiency is going to be under 30%.

If you want to MODULATE the LED, efficiency and accuracy both matter. That's why there's LOTS of solutions for 'LED drive'.

The incessant asshole, PAL... you know, the one with the excessive HUB disorder, thinks that all you need to drive an LED is a resistor.

The question posted was whether a current limiter could be used to control the current level, and I posted how to do it... as requested.

Precision video requires exotic control mechanisms, as shown here in one of my several billboard controllers ;-)...

Of course PAL... you know, the one with the excessive HUB disorder, will further dump, but will not show you a high quality circuit... just denigration.

So PAL... bye >:-} ...Jim Thompson

Gawd! You're becoming more PAL-like every day. The patent application is for a method to set the minimum supply and still have current sources behave like current sources... and it tracks LED behavior and temperature and process variables on the chip.

And the fee was paid >:-} ...Jim Thompson