Strange light curve

Tom F

I'm doing the WAM photoresistor project. I set the stamp board in a window facing a dark backyard at dusk. I collected data every five minutes, as shown in the attachment.

I expected a nice flat curve during the dark of night. Instead, I got the noisy thing I attached. I can't figure it out. The dusk and dawn curves are beautiful. But the middle of the night is so darned noisy. And it's not just random noise. Over a fifteen minute period (three points) there is a progression down or up...

I am using a byte variable to collect the RC time from the photoresister, but my tests with a word variable showed values always lower than 255, so I didn't think I needed a word. Tonight I'll try a word variable and see if I get better results.

Any other suggestions?

Tom

Post Edited (Tom Field) : 5/29/2006 7:40:48 PM GMT

MikeK

Was there a moon? The variation could be caused by clouds passing in front of the moon on a windy night.
How much light is required to drive the reading from low (around 5?) to 250 on your scale?

DaveG

Tom,

I'm not sure what your problem is, but an experiment might help.

1. With room light falling on the detector, take data for several 5 minute periods. The graph should
be near zero, or at some low value.
2. Without moving anything, put a light-tight cover over the phototransistor and let it run for a while.
3. Do you see the same fluctuations?
4. Make sure the .01 or 0.1 mfd capacitor is the correct value and that it is actually connected across the
phototransistor.

Dave G

Tronic (Greece)

Perphaps the inteferences was an atmospheric fainomenal like radiation from outer space invisible to human eye but luminal enouph to excite your sensor... LDRs are known to be affected by infrared light which is invisible to human eyes...!

Just an idea...



▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


Greekbotics: Greek Robotics Forum

Post Edited (Tronic (Greece)) : 5/30/2006 2:09:51 PM GMT

Tracy Allen

I agree with DaveG's experimental approach. You can also attach the sensor by itself to an ohm meter, and walk around to see what kind of resistance you are getting at night. My guess is that the resistance is up in the multi-megaohm range. The the circuit with those high resistances at night might be picking up electrical noise from the power line and other sources. Point sampling without averaging can give wild fluctuations.

Does the sensor point to a dark area, not a road or something where there is direct or relected indoor light?

Full moonlight is less than 0.2 lux, while at dawn and dusk the light level quickly rises above 100 lux, and up to 100000 lux in full sunlight. That is 6 orders of magnitude.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Paul Baker

Its the aliens trying to communicate with you Tom .

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
1+1=10

Beau Schwabe

What kind of data results did you get when trying the word variable? To me, it looks like there is "byte clipping" going on.
If you hook an Ohm meter up to your photocell, what kind of readings does it produce in day vs. night?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Tracy Allen

Good call Beau. It does look like byte clipping. The values plotted never get above ~250.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com