using a LED as a light sensor

hum..

I am doing exactly this with an FPGA
but I only see about 0.2% timing difference between dark and light

with some other methods the difference is about 10%
Antti

On 7 May 2006 23:37:15 -0700, the renowned "Antti"
<Antti.Lukats@xilant.com> wrote:

>hum..
>
>I am doing exactly this with an FPGA
>but I only see about 0.2% timing difference between dark and light
>
>with some other methods the difference is about 10%
>Antti

Maybe there's an internal pull-up or pull-down you need to turn off. 


Best regards, 
Spehro Pefhany
-- 
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

well I specially turned it ON not off

Antti

"Spehro Pefhany" <speffSNIP@interlogDOTyou.knowwhat> wrote in message
news:619u52tt1vjveq72agp691mm9d86ustcak@4ax.com...
> On 7 May 2006 23:37:15 -0700, the renowned "Antti"
> <Antti.Lukats@xilant.com> wrote:
>
> >hum..
> >
> >I am doing exactly this with an FPGA
> >but I only see about 0.2% timing difference between dark and light
> >
> >with some other methods the difference is about 10%
> >Antti
>
> Maybe there's an internal pull-up or pull-down you need to turn off.

I'm a bit surprised that LED's are sensitive enough to use a simple input
port discharge approach. Microcontroller ports typically spec input leakage
in th 1 uA to 10 uA range without pullups. OK, that's over the temperature
range. But we are not trying to make a temperature sensor but a light
sensor. 1 uA of photocurrent is a lot of current - about 5 x the saturation
level of the systems I use with 5x5 mm photodiodes. so, yes, I would
surprised if the timing change from dark to light was very great - and
swamped by the change with temperature.

Peter

well I was able to make a iopad oscillator that had frequency change
10% based on light intensity, and even that was possible bad solution,
eg the change could be even larger.

but getting stable absolute reading is probably more difficult or
impossible due to temperature, effects, etc

Antti

On 8 May 2006 04:00:56 -0700, the renowned "Antti"
<Antti.Lukats@xilant.com> wrote:

>well I specially turned it ON not off
>
>Antti

There's your problem. The photodiode current is negligible compared to
the pullup. 


Best regards, 
Spehro Pefhany
-- 
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

On Mon, 08 May 2006 11:02:36 GMT, the renowned "Peter Dickerson"
<first{dot}surname@tesco.net> wrote:

>"Spehro Pefhany" <speffSNIP@interlogDOTyou.knowwhat> wrote in message
>news:619u52tt1vjveq72agp691mm9d86ustcak@4ax.com...
>> On 7 May 2006 23:37:15 -0700, the renowned "Antti"
>> <Antti.Lukats@xilant.com> wrote:
>>
>> >hum..
>> >
>> >I am doing exactly this with an FPGA
>> >but I only see about 0.2% timing difference between dark and light
>> >
>> >with some other methods the difference is about 10%
>> >Antti
>>
>> Maybe there's an internal pull-up or pull-down you need to turn off.
>
>I'm a bit surprised that LED's are sensitive enough to use a simple input
>port discharge approach. Microcontroller ports typically spec input leakage
>in th 1 uA to 10 uA range without pullups. OK, that's over the temperature
>range. 

Typically something like 1nA at room temperature, and say doubling
every 10&#4294967295;C.  

>But we are not trying to make a temperature sensor but a light
>sensor. 1 uA of photocurrent is a lot of current - about 5 x the saturation
>level of the systems I use with 5x5 mm photodiodes. so, yes, I would
>surprised if the timing change from dark to light was very great - and
>swamped by the change with temperature.
>
>Peter

Yes, that was my WAG that it might be close even under typical
figures. With a discrete 5mm SR led I measure about 4nA of photodiode
current under normal room illumination with 5V reverse bias. A
large-die green 0.56" LED display (decimal point) gives only about
half of that current. 


Best regards, 
Spehro Pefhany
-- 
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

In comp.arch.embedded Don Klipstein <don@manx.misty.com> wrote:

>   Although I have found the small area to be a big consideration, I have 
> also found some to be better at converting sunlight received by that area 
> into electrical power than same area of monocrystalline solar cell.  

Well, that figures.  A LED may have a very small active silicon area.
But it's contained in a housing optically designed for the express
purpose of getting as much light from that very small area out into
the world.  Since light paths generally work equally well in both
directions, that implies that the same housing is also quite good at
*collecting* light from the world onto that tiny area where the LED is
sensitive to light.

-- 
Hans-Bernhard Broeker (broeker@physik.rwth-aachen.de)
Even if all the snow were burnt, ashes would remain.

John Larkin wrote:
> For a more serious challenge, use the live 7-seg display itself as the
> ambient light sensor. No ADC allowed. For advanced credit, no extra
> parts allowed at all, just firmware.

Time to call on the Keebler Elf Konsulting Kompany, huh?

BrunoG wrote:
> Hi,
> 
> Here is an idea about using a LED as a light sensor, to dim a 7-segments LED 
> display :

Reminds me when I needed a light sensor on the spur of the moment. I 
took a pair of dykes and peeled off the metal can of a transistor and 
used that.