From: John Devereux
Subject: Re: current source for array of infrared leds
Date: Tue, 03 Dec 2002 07:10:44 +0000
References: <email@example.com> <firstname.lastname@example.org> <email@example.com>
X-Newsreader: Forte Agent 1.8/32.548
NNTP-Posting-Date: 03 Dec 2002 07:06:31 GMT
firstname.lastname@example.org (John Fields) wrote:
>On Mon, 2 Dec 2002 13:35:33 -0800, "Harry Dellamano"
>> Let's say light intensity at normal temp (25C) is like "Voltage Offset" and
>>light intensity at elevated temps (80C) is like "Voltage Gain" in our
>>voltage mode system. The offset is set at 25C by setting the supply's output
>>to the correct voltage (+/-0.25%) for the desired LED intensity (current).
>>The LED's negative voltage tempco is about right to get increased current at
>>80C for constant intensity. It's only a first order approximation and you
>>also have the voltage source's tempco. For closer intensity tracking a
>>thermistor, either PTC or NTC, can be added to the voltage feedback network
>>to set the gain. The units I've done, no thermistor was necessary.
>> Bottom line, there is a large intensity drop-off, to the eye, at 80C with
>>CC mode of operation and undetectable drop-off with CV mode.
>But, what you're doing when you drive an individual LED with a
>constant voltage supply is allowing the resistance of the LED to
>determine how much current the supply will push through the LED.
>Since the tempco of the LED is negative, if you have a perfect
>constant voltage supply supplying current to the LED, then as the
>LED warms up and its resistance decreases the supply will supply
>current as required to keep the voltage across the LED constant,
>which will heat up the LED and lower its resistance, which will...
>So, your plan sucks, blows, is wrong...
That's too strong; whether it runs away depends on what the
tempco actually is. Harry is implying that in fact the
tempco can be just what is needed to make up for the
decreased light output (at some particular operating level,