Pages: [1]   Go Down
Author Topic: Vishay TEPT5700 - Ambient Light Sensor problem  (Read 1586 times)
0 Members and 1 Guest are viewing this topic.
Slovenia, EU
Offline Offline
Full Member
***
Karma: 0
Posts: 153
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

I am using Vishay TEPT5700 sensor for automatic turning outside light on and off between day and night. I wired the sensor with 10k resistor to analog pin. I made the whole circuit, test it and measure it. Everything works but it looks like sensor depends a lot of outside temperature although graph in datasheet is almost linear. This is the problem because when outside is -10 degrees, the sensor value is 185 in dusk (for example, when i want to turn the light on), and 195 when outside is +10 degrees. Because of that i can't set proper on/off turning border (i tried but difference is too big).

Have got anybody any idea, what could be wrong (maybe resistor value or wiring?) or recommend another sensor?

Here's my wiring scheme for sensor..
https://www.dropbox.com/s/1lnpmau4ah16jv3/vishay.png

Sensor datasheet
http://www.google.si/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&ved=0CEgQFjAB&url=http%3A%2F%2Fwww.vishay.com%2Fdocs%2F81321%2Ftept5700.pdf&ei=6NgSUZGvNpGM4gS9woF4&usg=AFQjCNHQRhqLQjioGuUq4yGjNVLnacavWA&bvm=bv.42080656,d.bGE
Logged

Dubuque, Iowa, USA
Offline Offline
Faraday Member
**
Karma: 49
Posts: 2544
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

The typical scheme is 5V -> 10K resistor -> analog pin -> phototransistor -> GND (a "common emitter" scheme).  This sets the pin high when it's dark and low when it's bright. I can't say I fully understand the math but it just seems to work better at typical light levels.

With your current wiring a larger resistor should give you more of a sensor swing.
Logged

Slovenia, EU
Offline Offline
Full Member
***
Karma: 0
Posts: 153
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

The typical scheme is 5V -> 10K resistor -> analog pin -> phototransistor -> GND (a "common emitter" scheme).  This sets the pin high when it's dark and low when it's bright. I can't say I fully understand the math but it just seems to work better at typical light levels.

With your current wiring a larger resistor should give you more of a sensor swing.

So this scheme is not ok? I don't understand how you mean that i should wire it.. Can you make a scheme please?

And you recommend bigger resistor value.. I will try that. Any starting point (1M, 10M?)

Can somebody recommend any other sensor, which works well for this purpose? Btw, what kind of sensors are in those automatic turning on/off light detectors?
Logged

Offline Offline
Sr. Member
****
Karma: 3
Posts: 259
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

For starters you setup is totally wrong you want to feed the ptoto transistor with a 100k resistor VDD to collector and tap the  ADC to the emittor with a 10k resistor

 


* lightsen.png (5.5 KB, 415x284 - viewed 103 times.)
Logged

Slovenia, EU
Offline Offline
Full Member
***
Karma: 0
Posts: 153
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

For starters you setup is totally wrong you want to feed the ptoto transistor with a 100k resistor VDD to collector and tap the  ADC to the emittor with a 10k resistor

Thank you for your explanation.. I will try that!

I used this scheme before.. http://img.techpowerup.org/090723/pt%20circuit.gif
So that mean that this sheme is wrong..

I don't understand why Vishay didn't put any "typical configuration" schemes in datasheet. I googled a lot of pages about phototransistor but there's so many different schemes.
Logged

Slovenia, EU
Offline Offline
Full Member
***
Karma: 0
Posts: 153
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

The typical scheme is 5V -> 10K resistor -> analog pin -> phototransistor -> GND (a "common emitter" scheme).  This sets the pin high when it's dark and low when it's bright. I can't say I fully understand the math but it just seems to work better at typical light levels.

With your current wiring a larger resistor should give you more of a sensor swing.

I changed my wiring like you said, with common emitter and 10k resistor..

Like this, just with phototransistor
http://upload.wikimedia.org/wikipedia/commons/thumb/a/a0/NPN_common_emitter.svg/220px-NPN_common_emitter.svg.png

I will test it outside but still, ratio between night and room with medium lighting is really small.. In the dark i get value about 833 while in room with medium lighting value is about 825. Could i make this "area" more accurate, with bigger ratio area if i change resistor value?
Logged

Offline Offline
Sr. Member
****
Karma: 3
Posts: 259
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

The way I posted was at the emitter side which should work fine the one you
Posted is not going to have the range your looking for
Logged

Slovenia, EU
Offline Offline
Full Member
***
Karma: 0
Posts: 153
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

The way I posted was at the emitter side which should work fine the one you
Posted is not going to have the range your looking for

I just find original scheme from VISHAY and it shows that phototransistor should be wired exactly the same as i wired it first time, just with 200k resistor.

https://www.dropbox.com/s/u735esrmkvcc2yb/vishay_datasheet.png

I also tried with your scheme be80be but with 100k and 10k resistor i get really small values range..

If VISHAY's original scheme is correct than, in that case, sensor really depend of temperature a lot.
Logged

Slovenia, EU
Offline Offline
Full Member
***
Karma: 0
Posts: 153
View Profile
 Bigger Bigger  Smaller Smaller  Reset Reset

I fixed it.. I used the same scheme as first time, just with 200k resistor. Now the whole thing is really sensitive and i can set turning border more accurate.
Logged

Pages: [1]   Go Up
Jump to: