LED reference current source

Must have been, since I just saw some discussion about it on LTspice group less than a month ago.

Jon

message=20

He spec'd an RF transistor and JL sticks base resistors on BJT bases like others put bypass caps on ICs. Everywhere.

Jon

What is it supposed to buy you? Does the LED forward TC match a Silicon Vbe TC? ...Jim Thompson

-- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

Do you need or want the base resistor?

Only if there's a resistor in the emitter.

message=20

Yup. ;)

Jon

Toob guys have been doing it for ages. Where 'ya been?

Tim

-- Deep Friar: a very philosophical monk. Website:

"John Larkin" wrote in message news: snipped-for-privacy@4ax.com...

Like, as in a VR tube and a p-channel triode?

As far as doing it with semiconductors, there seems to be a lot of different chemistries in LEDs, namely lots of different forward voltage curves for a diveb color, and not much data about tempcos.

--

John Larkin Highland Technology Inc

Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom timing and laser controllers Photonics and fiberoptic TTL data links VME analog, thermocouple, LVDT, synchro, tachometer Multichannel arbitrary waveform generators

Is there any significant matching between the LED and the Si BE junction? I've used this (or the opposite sex) with6 really good results:

Vcc +--------+ | | | V ->

| - .-. | | | | | | ||-+ '-' ||

That will have a tempco of around -0.4% per degree C. In theory an LED tc cancels the Vbe tc some. I just don't know how much. Google is ambiguous on the subject.

I guess I'll have to build it and try it. We have some Osram LEDs that are a gorgeous orange color, about 1.8 volts drop at 1 mA, so they are a good candidate.

If it's just an LED, who cares?

I saw it decades ago for audio amp bias, claim was green LED (old-style now, not high modern intensity types) balanced Vbe tempco.

Grant.

"John Larkin" wrote in message news: snipped-for-privacy@4ax.com...

No, as shown... They do it in immitation of the SS amp guys, who have been doing it for even longer.

I once had a chemistry lab about bandgap. We took a battery, LED and styrofoam cup of LN2, and proceeded to observe the color change upon putting them together. Theory from lecture said, cold --> crystal shrinks --> wavelength shrinks.

So naturally, I had grabbed a green (GaP) LED out of the bin (versus red GaAsP, or blue GaN, which apparently are more similar to each other than to GaP). The professor noted its behavior and suggested I demonstrate with a red or blue LED instead. ;-)

Tim

That LN2 demo was always much more fun with the early generation of low efficiency red LEDs. The quantum efficiency rockets up when the lattice is cooled and thermal vibrations are decreased much more rigid lattice. Unfortunately they die after a few hundred cycles or less.

I don't recall seeing the wavelength shift perceptably though the brightness changed by a very significant amount. Admitedly on a red led this could be both quantum efficiency and it producing shorter wavelength red light which is perceived as bright.

I still have a few original 1970's red, yellow and green LEDs that are now in series with modern high efficiency ones to show them off.

ISTR a circuit using green LEDs and BJTs is as old as the hills - certainly dates from the era when LEDs first became available.

Next time I have LN2 to hand and a spectrograph I will try it.

Analog Device SSM-2220 data sheet uses that scheme. Nelson Pass amps have used it in production. LED biasing was a wave or sorts in high end audio.

The RF transistor I presume is for low capacitance. The low beta could be an issue.

The high efficiency LEDs are double heterojuntion. When they first came out, the existing generation LED drives had insufficient compliance and produced a dimmer light than the standard LEDs. That is the high voltage of the double heterojunction LED made the drive chips put out less current.

It wasn't like the LED companies gave the chip companies a heads up on the specs of the new generation LEDs so that the new driver chips would be available.

seen it done since forever for audio amps, I believe some claim they are also less noisy than regular diodes, just remember to shield the LED from light

looks like somewhere between red and UV will get close to zero tempco

-Lasse

Maybe decap the transistor and shine the led on it :-)

Also in SQUID readout amplifiers, like Pasquarelli & Rossi in the ISEC'97 conference.

What about NESG4030, hFE=3D400 typ, Cre=3D0.12pF typ? I know, costs way too much...

Regards, Mikko