It's just not my week...

[Dave VanHorn]

Ok, what's the possibility that visible LEDs (Red, yellow, green, and Blue)
have any significant output in the near IR spectrum? This has been brought
up as a candidate explanation for low performance on an IRDA transciever,
with integral filter.

I'm in process of asking the LED manufacturer, but nobody's in, and none of
their data sheets have anything more than a center of emission number...

It's been a !@$#@$ week.

 

[John Larkin]

Ok, what's the possibility that visible LEDs (Red, yellow, green, and Blue)
have any significant output in the near IR spectrum? This has been brought
up as a candidate explanation for low performance on an IRDA transciever,
with integral filter.

I'm in process of asking the LED manufacturer, but nobody's in, and none of
their data sheets have anything more than a center of emission number...

It's been a !@$#@$ week.

Red, a tiny bit maybe. Spectral emission of LEDs is pretty narrow. A
635 nm red might have a couple per cent at 700, fading fast.

John

 

[Rene Tschaggelar]

Ok, what's the possibility that visible LEDs (Red, yellow, green, and Blue)
have any significant output in the near IR spectrum? This has been brought
up as a candidate explanation for low performance on an IRDA transciever,
with integral filter.

I'm in process of asking the LED manufacturer, but nobody's in, and none of
their data sheets have anything more than a center of emission number...

It's been a !@$#@$ week.

In addition to what John said :
you can use a black and white digital camera to detect
near IR. As IR filter you can use a CD as grating.
Just test whether somethinge comes (angular) beside
the red deviation.

Rene

 

[Dave VanHorn]

Red, a tiny bit maybe. Spectral emission of LEDs is pretty narrow. A
635 nm red might have a couple per cent at 700, fading fast.

That's what I thought. When you add the fact that the LEDs are not
modulated, it seems pretty unlikely. This is kind of like being sued, if
you fail to defend, you loose...

 

[James Meyer]

Ok, what's the possibility that visible LEDs (Red, yellow, green, and Blue)
have any significant output in the near IR spectrum? This has been brought
up as a candidate explanation for low performance on an IRDA transciever,
with integral filter.

If it were my problem and if the problem would show up long enough, I'd
try switching the LEDs off to see if the problem went away. If I couldn't
switch them off, I'd cover them with a couple of layers of black electrical
tape.

Jim

 

[John Popelish]

Ok, what's the possibility that visible LEDs (Red, yellow, green, and Blue)
have any significant output in the near IR spectrum?

(snip)

I don't have detailed experience with visible LEDs, but I found out
the hard way that 940 nm IR LEDs have a long wave IR tail, with a big
bump around 1200 nm, and a fade out past 1500 nm. They also have a
bit of short wave output down to about 700 to 800 nm. None of this
shows up on the data sheets, unfortunately. It is a shame they don't
use a log log (Bode) plot of amplitude verses frequency. I would not
be at all surprised if other types also produce some low level long
wave outputs.

 

[Fred Bloggs]

That's what I thought. When you add the fact that the LEDs are not
modulated, it seems pretty unlikely. This is kind of like being sued, if
you fail to defend, you loose...

That is a good idea- do modulate a sample visible LED and synchronously
detect the IR detector output to quantify the interference- something
like ambient N -60dB.

 

[Dave VanHorn]

That is a good idea- do modulate a sample visible LED and synchronously
detect the IR detector output to quantify the interference- something
like ambient N -60dB.

Good idea. That would pretty much decide the question.
If I can't get the detector's attention with it face on, then it likely
isn't causing any problem sitting in it's proper place, with most of it's
energy going up a light pipe.