Measure Wavelength of IR LED

[Phoxtane]

I'd like to be able to measure the wavelength of the IR LED in a remote control. Thing is, I don't know how to approach the problem since the typical DIY spectrophotometer seems to be more of a demonstration unit as opposed to a measuring device, and I only ever see them being used with visible light.

The main reason for wanting to do this is so I can pick out an IR LED that duplicates the one in the remote as closely as possible. However, I don't actually know if I need to worry about this; if I remember correctly, IR refers to a band of wavelengths, and in theory the receiver should respond to any properly encoded signal within this band, right? Unless, of course it only responds to a much narrower range of wavelengths within that of IR light...

[grouchobyte]

You dont need to know the wavelength for what you are trying to accomplish. More than lightly the wavelength is between 800-900 nm. We call that near IR. Anything you use in that band will work. The detector just sees all pulses of light in the spectral band it can see and cant really differentiate one wavelength from another.

That said, only Si and Ge detectors are going to have responsivity in that band of interest. Longer wavelengths are generally more expensive and reserved for higher bandwidth fiber optic applications

Hope that helps answer your question

@grouchobyte

[bktemp]

840nm and 940nm are common wavelength values used for IR remote controls. It is easy to differentiate between both, because you can see 840nm slightly glowing in the dark, but 940nm is virtually invisible.
You can verify the LED is active by using a digital camera: The LED should be visible as a white or purple spot.

[Augustus]

If it's a standard ir remote control, they most likely use a ~940nm emitter. For example, this is from the datasheet of a common tsop31238 ir-remote receiver module, it has its highest sensitivity at around that mark. So if you want to get the longest range out of your remote it is a good idea to use such emitters. Shorter wavelengths at around ~850nm are more commonly used for lighting purposes (night vision security cams etc.), at least to my knowledge 

[bobaruni]

Just use two different photo diodes and trans-impedance amplifiers, one centred at 850nM and the other at 940nM and compare the output ratio of the two.
If you have an 850nM remote control, the 850nM photo diode will register more signal and visa versa.

Using this technique, you can make it as simple or as elaborate as you like.
Also, a way of achieving very narrow response for the said wavelength is to use an actual IR emitter LED as a photo detector, just keep in mind LEDs won't be as sensitive and capacitance may be very high limiting the speed.

[Circlotron]

You can verify the LED is active by using a digital camera: The LED should be visible as a white or purple spot.

My iPhone 4S has an IR filter on the main camera but not on the selfie camera, so use that one. Other phones may be similar.

[grouchobyte]

Just use two different photo diodes and trans-impedance amplifiers, one centred at 850nM and the other at 940nM and compare the output ratio of the two.
If you have an 850nM remote control, the 850nM photo diode will register more signal and visa versa.

Using this technique, you can make it as simple or as elaborate as you like.
Also, a way of achieving very narrow response for the said wavelength is to use an actual IR emitter LED as a photo detector, just keep in mind LEDs won't be as sensitive and capacitance may be very high limiting the speed.

Bobaruni......with all due respect......

To differentiate between two relatively close wavelengths like 850 and 940nm you will need more than two PDs and two TIAs. More like narrow optical filters. Using an LED as a PD is like using a kitchen knife as a screwdriver. Ya, it works kinda.....But then again there is really no reason to do that. Measuring relative optical magnitudes to determine wavelength based on incident power and PD responsivity is gonna lead to erroneous results.

Just my two cents having designed fiber optic gear for over 29 years......

@grouchobyte

[Kleinstein]

Most of the detectors work with both 940 nm and 850 nm LEDs. The curve for the TSOP31238 is the first a saw with so much difference. Usually the sensitive range is more like 750 nm to 980 nm. So it is not that important to have the same wavelength as the original.

One possible way to distinguish between might be using a 850 nm LED as a detector - it should have rather low sensitivity to light from a 940 nm LED - mainly from the small amount of < 860 nm light coming from a 940 nm LED. You may still need two detectors to compare, so one with a normal Si photodiode and one with a 850 nm IR LED as a detector. The ratio of the signals from these two should be rather different.

I have not tested a 850 nm IR LED as a detector so far. Some IR diode use quite some tricks for better efficiency and thus these LEDs might not work that well as a detector at all. But chances are quite good for it to work.

DIY spectrometers should work in the NIR range as well. The main limit is the usually silicon based sensor. So sensitivity could be lower at 950 nm, but 950 nm have quite some intensity. The simple way is using a camera (preferably without IR filter - so more like cheap web cam). So getting one without a IR filter (or removing the filter) is the only challenge.

[bobaruni]

To differentiate between two relatively close wavelengths like 850 and 940nm you will need more than two PDs and two TIAs. More like narrow optical filters. Using an LED as a PD is like using a kitchen knife as a screwdriver. Ya, it works kinda.....But then again there is really no reason to do that. Measuring relative optical magnitudes to determine wavelength based on incident power and PD responsivity is gonna lead to erroneous results.

Just my two cents having designed fiber optic gear for over 29 years......

@grouchobyte

My suggestions are based on ease of availability and cost, the price and availability of narrow band optical filters exclude them from general hobby use.
This is why I suggested using IR emitter LEDs as PD's as this makes them behave somewhat like a PD with a narrow band optical filter in front of them.
You would be surprised at how well LEDs work as narrow band PDs, but don't take my word for it: http://makezine.com/projects/make-36-boards/how-to-use-leds-to-detect-light/

While I agree that there is great potential for getting erroneous results, especially when it comes to getting 2 PD's of different wavelengths to have a similar axial response, my idea will work.
Again, don't take my word for it:
lets do the math with 2 PD's
PD1, TIA1: 850NM lets call this PD850
PD2, TIA2: 940NM lets call this PD940
Firstly, both TIA's are calibrated for roughly 1V output when stimulated by it's centre wavelength of equal power.
850nM beam: PD850 output ~ 1V, PD940 output ~ 0.5V
940nM beam: PD850 output ~ 0.6v, PD940 output ~ 1V

To me, it looks like it would be pretty easy to differentiate between the incoming wavelengths.
If you need more differentiation, then spread the wavelengths of the PD's out more, eg 820nM and 950nM.
Or you can use narrow band optical filters and normal  PDs or a couple of LEDs acting as PDs with narrow band optical filters.

[grouchobyte]

.......filters exclude them from general hobby use.
This is why I suggested using IR emitter LEDs as PD's as this makes them behave somewhat like a PD with a narrow band optical filter in front of them.
You would be surprised at how well LEDs work as narrow band PDs, but don't take my word for it: http://makezine.com/projects/make-36-boards/how-to-use-leds-to-detect-light/

While I agree that there is great potential for getting erroneous results, especially when it comes to getting 2 PD's of different wavelengths to have a similar axial response, my idea will work.
Again, don't take my word for it:
lets do the math with 2 PD's
PD1, TIA1: 850NM lets call this PD850
PD2, TIA2: 940NM lets call this PD940
Firstly, both TIA's are calibrated for roughly 1V output when stimulated by it's centre wavelength of equal power.
850nM beam: PD850 output ~ 1V, PD940 output ~ 0.5V
940nM beam: PD850 output ~ 0.6v, PD940 output ~ 1V

To me, it looks like it would be pretty easy to differentiate between the incoming wavelengths.
If you need more differentiation, then spread the wavelengths of the PD's out more, eg 820nM and 950nM.
Or you can use narrow band optical filters and normal  PDs or a couple of LEDs acting as PDs with narrow band optical filters.

Bobaruni

While I see your point about the practicality of using an LED as a detector, I am not keen on the idea of assuming repeatability of the optical or incident power. Measuring incident power as a function of responsivity in order to determine wavelength can yield erroneous results in a practical application or enviroment. However in a controlled laboratory experiment or setup it may work fine.

I think we agree on these salient points

@grouchobyte

[Siwastaja]

As someone already mentioned, you can use your eye as a quick instrument to identify (or give a really good guess) between 850nm and 940nm. Turn your lights off or go to a dark room and you'll easily see the 850nm LED in the dark, and you won't see 940 nm (in most cases). No need to make this any more complex. Then you know which wavelength to buy if you want to really match it; even though it doesn't really matter.

Digital cameras see both wavelengths similarly; even the cameras with IR filters, because the filters are not perfect; they attenuate the light significantly, but these LEDs are so bright that when you directly aim them at the camera from a near distance, the one without a filter is totally blown away, while one with a filter still clearly shows the light. The light can show as any color, depending on the RGB filter materials used in the sensor, but purplish white is typical.

[Buriedcode]

The difference in wavelength between LED's in remotes will make little difference when compared to the carrier.  Although 36kHz will work with 38Khz Rx and 40kHz Rx the latter will have greatly reduced range as they tend to have fairly narrow filters built in.

Why not just socket the LED and test them in a remote?  If they both work at the range you require, then you could try reducing the LED current to see which one starts losing range.  The peak current in IR remote LED's is high due to the very low duty, but this sort of test is far easier, and perhaps quicker than just going by wavelength.

[janoc]

Just point the LED at a phone camera. 850nm will be bright white (even if the camera has IR filter - that does not totally block the IR, only attenuates it!), 940nm will be barely glowing, because the common CCD/CMOS camera sensors are much more sensitive in the 850nm range than around 940nm.

That should be more than enough to quickly identify the wavelength between the two values.

[LaserSteve]

Ebay seller bejomejag has a IR dichroic lowpass  filter with a cutoff below 900 got 19.95. That would settle this once and for all.

Steve