IR sensors too noisy

[shehry]

Hi,

i am designing a line tracking robot system. i was trying to detect the
difference b/w black and white using an IR sensor. Now the problem is
that the sensors are fairly noisy. and the voltage difference is very
minute...about .1V. Now, i dont want to use an amplifier since that
would amplify the noise too. so i want a way to reduce this noise and
so be able to detect the voltage by then using an amplifier

can anyone help me in this regard.

thanks

 

[John Popelish]

Hi,

i am designing a line tracking robot system. i was trying to detect the
difference b/w black and white using an IR sensor. Now the problem is
that the sensors are fairly noisy. and the voltage difference is very
minute...about .1V. Now, i dont want to use an amplifier since that
would amplify the noise too. so i want a way to reduce this noise and
so be able to detect the voltage by then using an amplifier

can anyone help me in this regard.

thanks

What light source are you using to illuminate the B/W object?

 

[shehry]

What light source are you using to illuminate the B/W object?
I was using an IR transmitter for illumination

 

[Luhan]

Hi,

i am designing a line tracking robot system. i was trying to detect the
difference b/w black and white using an IR sensor. Now the problem is
that the sensors are fairly noisy. and the voltage difference is very
minute...about .1V. Now, i dont want to use an amplifier since that
would amplify the noise too. so i want a way to reduce this noise and
so be able to detect the voltage by then using an amplifier

can anyone help me in this regard.

thanks

That 'black/white' distinction may not exist in the IR wavelength.
Hence, your poor signal to noise ratio. Better to use visable light here.

Luhan

 

[Dave]

I was using an IR transmitter for illumination

What is the room lighting - fluorescent tubes are bad news, for instance.

Try the system in a darkened room. If most of the interference disappears
you know where its coming from...

Make sure that your IR sensor has a built in filter (typically they are
moulded in what looks like black plastic)

There is a small possibility that the B/W target you are using may not be
Black and White when seen at IR wavelengths. Try different materials.

If the response is still poor, then you need some or all of:-

Reduce interference by shielding from direct and reflected room light

Stronger transmitted signal (=>higher current into IR LED, but watch
dissipation.)

Optics to concentrate signal onto a smallish area

Reduce distance between transmitter and receiver (inverse square law
applies)

Increase signal into detector (use a lens to gather more IR)

In the long term this kind of thing is best done by pulsing the transmitter
and using an AC amplifier with synchronous detection.

Have fun !
Dave

 

[John Popelish]

I was using an IR transmitter for illumination

Not nearly enough information.

By IR transmitter, do you mean some prepackaged device that might
include modulation, or just an IR LED fed by DC?

Modulating the source, and amplifying the detector signal through a
narrow bandpass filter is a common way to reject interfering signals,
like room lighting.

Do you have control over the surface of the black and white objects?
If so, replacing the white with retro reflective tape can increase the
contrast dramatically. This is especially true of the surfaces are a
significant distance from the source and detector. If the distance is
very small, it is difficult to get the source and detector close
enough that the retroreflected light ends up on the detector, instead
of on the source.

Is your detector filtered to block most visible light? IR LEDs come
in two common types, 880 nm and 940 nm. The 880's often have higher
output, but the 950's are close to the long wave limit for silicon
detectors. So the 950's allow slightly better short wave blocking
with only a narrow band of wavelengths that get through that can be
detected by silicon between the filter's passband, and silicon's long
wave limit.

 

[Rene Tschaggelar]

Hi,

i am designing a line tracking robot system. i was trying to detect the
difference b/w black and white using an IR sensor. Now the problem is
that the sensors are fairly noisy. and the voltage difference is very
minute...about .1V. Now, i dont want to use an amplifier since that
would amplify the noise too. so i want a way to reduce this noise and
so be able to detect the voltage by then using an amplifier

can anyone help me in this regard.

Yes, you should modulate your light source to get
rid of the ambient light.

Rene