On Thu, 18 Jan 1996 10:05:52 +0000, Simon Watson <simon.watson@dur.ac.uk>
wrote:
>Hi! If anyone could help me with ideas/suggestions for solving the
>following problem I'd be really grateful.
>Basically, I need to use the output of a DAC to control the intensity of
>two infra-red LEDs (only one lit at a time). Each LED will only be
>pulsed for around 10uS, allowing forward currents of up to 1A (from the
>datasheet), with the whole thing under control of a microprocessor. The
>DAC is a 10-bit voltage output device, allowing for a linear resolution
>of 1mA or so. An important requirement is that each intensity level is
>repeatable, so is some sort of temperature compensation required?
>What's the best way of achieving this? Any help will be much
>appreciated.
Here's a wild shot in the dark. I'm a complete newbie, but this might
*just* work...
Basic circuit:
DAC output -> matched (monolithic) transistor current mirror ->
current multiplier -> current-controlled LED.
---o----o-------------o-----------------o--- Vcc
| | | |
> | | |
R1 < | | |
> / \ |
| |v v| _|_
o--| Q1 Q2 |--. \ /
| |\ (pnp) (pnp) /| | _V_ \\
| \ / | |
| | | | | LED
o----|-------------|----' |
| | | .-------------'
o------|----o | |
| | `--( )
From DAC < < ( ) 1:20
> > R3 | Current multiplier
R2 < < |
| | |
| | |
-o----o-----------------o------------------ Gnd
The DAC voltage will control the Ic of Q1, which will then cause Q2 to
reflect the same current. That current is then multiplied by the current
multiplier (TLxxx series or something or other ... don't really know) through
the LED.
Q1 and Q2 should be a monolithic pair (to compensate for temperature
variation), and R3 should be 1% tolerance. As long as you can control the
variation of the DAC's voltage over temperature, the current through the LED
should be pretty much proportional to the input voltage.
Typical values (for 0-5V DAC output, 10V Vcc):
R1 : 1.5k Ohm
R2 : 20k Ohm
R3 : 100 Ohm (1%)
This will give a 0-1A current swing through the LED for 0-5V swing from the
DAC. Max. dissipation of Q1 is 500mW, Q1+Q2 dissipates 1W.
Comments anyone? I'm still trying to figure this out myself.
/+_,--___ _|
++( P \_\ \ ---------------------------------
|++ U ) ) \_\ Dave Rahardja at ORU, Tulsa, OK
-++ ( P /__/\:: ---------------------------------
/++++__--~ |/
Date: Mon, 22 Jan 1996 07:22:27 GMT
Original Subject: Re: Controlling LED intensity