# Will a 950 nm IR emitter work with a 870 nm phototransistor?

I'm planning to build a simple IR light barrier across a distance of 15 mm.

I plan to use this receiver with this emitter.

The reason is to have some SMD PNP capable components. The emitter and sensor have a slightly different IR wavelength. Will it work anyway?

• for the short distance of 15mm , you could have used the LED as a PD, reverse biased. Got any stray sunlight? – Tony Stewart Sunnyskyguy EE75 Jan 29 at 8:25

Side by side: -

The receiver will be slightly off centre compared with the transmitter (purple lines I added are sitting at 950 nm) but the receiver will still be above 80% of its optimum sensitivity.

You should also consider that as the emitter warms up, the peak centre emission can move by 0.3 nm per degree C so, if it warmed 40 degC, the peak wavelength could move 12 nm. This could lower the overall receiver sensitivity into the 75 % area.

Bear also in mind that the graph of receiver sensitivity has a centre point that is defined as “typical” in the data sheet so, there could be some temperature dependency or drift associated with the receiver too. The data sheet does not appear to give details on this.

• +1 If the emission spectrum were a lot narrower I'd worry a bit about the change in band gap and central emission wavelength with temperature, but it's not, so I won't. – uhoh Jan 28 at 7:55

Yes, your system will work. Consider the system spectrum (red line below), which is each sensitivity at each wavelength times the emission at that wavelength.

You can see that the transmitter is operating on a 0.84 efficiency portion of the receiver, which ought to be well within any error margin you might have, considering losses for distance, misalignment, focus, etc.

Compare with an imagined transmitter whose peak is at 875 nm to match the receiver (dashed blue line) and whose corresponding matching system spectrum is also shown (thin red).

This is engineering: how good is the fit in numbers? The area under the 950 nm pairing is 90% of the area under the 875 nm pairing, and this considered a small loss.

Of course, the receiver is also dropping compared to a perfect "white" receiver, and it's about 5% (see for example difference between dotted blue and thin red at 800 nm.)

The spectrum for the receiver is scraped from a graph on the datasheet. For the transmitter, Osram publishes it in "optical simulation" section, ASCII file download. (The staircasing on the graph is just an artifact from the graph drawing.)

• +1 If the emission spectrum were a lot narrower I'd worry a bit about the change in band gap and central emission wavelength with temperature, but it's not, so I won't. – uhoh Jan 28 at 7:56
• awesome! thx @jonathanjo – Chris Jan 28 at 20:19
• I get 84% but close enuf i.stack.imgur.com/SV7Sp.png – Tony Stewart Sunnyskyguy EE75 Jan 29 at 8:17
• @TonyStewartSunnyskyguyEE75 of course you're right, fixed it. Also added efficiency calculation for quantitative purposes. – jonathanjo Jan 29 at 11:05

Yes, it will work. Look at the sensitivity/emission vs wavelength curves. It is a spread and there is a point of overlap where both devices have reasonably high values.

• +1 If the emission spectrum were a lot narrower I'd worry a bit about the change in band gap and central emission wavelength with temperature, but it's not, so I won't. – uhoh Jan 28 at 7:56

Yes, it will but sensitivity will be far from optimal.

• This is an unsupported statement, and given the information in the other answers (which were submitted before yours, so you could have read them) it's also an incorrect statement. – Alexandre Aubrey Jan 28 at 14:52
• Sorry, but your answer just doesn't seem to add anything to what was posted earlier. – Dmitry Grigoryev Jan 28 at 15:16
• @ Alexandre Aubrey -- the question was The emitter and sensor have a slightly different IR wavelength. Will it work anyway? -- now please read my answer. What is the point to explain how devices work if the OP does not ask for it? You have a problem with understanding the question and the answer. – Polar Bear Jan 30 at 22:14