Rock stable LED lightintensity and photodetector circuits

A

Anton

Guest
Hi!

I need light from a LED and photodetector to independetly be "rock stable".
ie output light intensity compensated for temperature variations at the LED,
and the same for the photodetector. Also power supply variations needs to be
considered. The compensations/regulations has to be done separately for the
LED and the detector. It is not possible to let the photodetector (or any
other photodetector) control the LED intensity.


Anyone got nice circuit(s) suggestion ??


/Anton
 
Since the output of a LED depends on things over which you have no control -
eg age, then you cant hope to compensate unless you can measure the
intensity.

--
Regards,

Adrian Jansen
J & K MicroSystems
Microcomputer solutions for industrial control
"Anton" <flaska@telia.com> wrote in message
news:3RLNc.99507$dP1.340855@newsc.telia.net...
Hi!

I need light from a LED and photodetector to independetly be "rock
stable".
ie output light intensity compensated for temperature variations at the
LED,
and the same for the photodetector. Also power supply variations needs to
be
considered. The compensations/regulations has to be done separately for
the
LED and the detector. It is not possible to let the photodetector (or any
other photodetector) control the LED intensity.


Anyone got nice circuit(s) suggestion ??


/Anton
Click to expand...
 
by assuming that the ageing process of the LED is slow compared to the time
it's suppost to be used between calibrations/checks. This implies defining a
method and criteria in order to decide when the LED has passed its "use
before" date (time). Ie regulary, as you suggests, control the leds
intensity with a photodetector when not used in the measurement.


/Anton

"Adrian Jansen" <qqv@noqqwhere.com> skrev i meddelandet
news:41082166$1@duster.adelaide.on.net...
Since the output of a LED depends on things over which you have no
control -
eg age, then you cant hope to compensate unless you can measure the
intensity.

--
Regards,

Adrian Jansen
J & K MicroSystems
Microcomputer solutions for industrial control
"Anton" <flaska@telia.com> wrote in message
news:3RLNc.99507$dP1.340855@newsc.telia.net...
Hi!

I need light from a LED and photodetector to independetly be "rock
stable".
ie output light intensity compensated for temperature variations at the
LED,
and the same for the photodetector. Also power supply variations needs
to
be
considered. The compensations/regulations has to be done separately for
the
LED and the detector. It is not possible to let the photodetector (or
any
other photodetector) control the LED intensity.


Anyone got nice circuit(s) suggestion ??


/Anton
Click to expand...
 
"Anton" <flaska@telia.com> wrote in
news:3RLNc.99507$dP1.340855@newsc.telia.net:

Hi!

I need light from a LED and photodetector to independetly be "rock
stable". ie output light intensity compensated for temperature
variations at the LED, and the same for the photodetector. Also power
supply variations needs to be considered. The
compensations/regulations has to be done separately for the LED and
the detector. It is not possible to let the photodetector (or any
other photodetector) control the LED intensity.


Anyone got nice circuit(s) suggestion ??


/Anton
Click to expand...
I think you need to give some more details of your requirement. What is
"rock solid"? ie. tolerable intensity drift with time. NO electronic
circuit does not drift at all. Also, it would be helpful if you describe
your application/problem because someone may have solved the same problem
in a different way.

An easy way to get a stable signal is by using a current source, or more
correctly a current sink. This is a configuration of an opamp, a
transistor, a current sense resistor, feedback loop, and a voltage
reference diode. This is stable for most apps but only guarantees a certain
current flow. The more stability required in time the higher spec the parts
and the more care required in design. These current sink ccts are easy
enough to find. eg National Semi applications handbook.

Also, calibration of the intensity is non-trivial and instrumentation to
do this is expensive.
 
Look for "Ocean Optics" on the web and look for their "calibration lamps" .
They have calibrated outputs and uncalibrated (but stable) output devices. .
but you would have to pay, of course.

FWIW

Dave


"Anton" <flaska@telia.com> wrote in message
news:3RLNc.99507$dP1.340855@newsc.telia.net...
Hi!

I need light from a LED and photodetector to independetly be "rock
stable".
ie output light intensity compensated for temperature variations at the
LED,
and the same for the photodetector. Also power supply variations needs to
be
considered. The compensations/regulations has to be done separately for
the
LED and the detector. It is not possible to let the photodetector (or any
other photodetector) control the LED intensity.


Anyone got nice circuit(s) suggestion ??


/Anton
Click to expand...
 
My specs are:

The output of the detector consists of to parts; DC-level and an AC-signal.
I'm interested in the AC-signal which has a bandwith from 0.0001Hz up to say
100Hz. (maybe strange to call it AC-signal but it is timevariable ;-) )
Acceptable resulting noise (from both LED and detector) is DC-level/10000.
For example ; DC-level out from the detector (and following
current-to-voltage transformer) is 4 V then maximum acceptable noise is 0.4
mV. The main problem has been to reduce temperature dependence on LED and
detector.

Therefore I need an LED-driver which is temp stab and also a photodetector
with following current-to-voltage convertion temp stab.

Has anyone used a avalance photodetector??

/

Anton
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top