Linearity of optoelectronics ?

I'm not entirely sure about LED's, probably pretty linear, but keep in mind that they convert *current* to light. If you drive them with an analogue voltage waveform, they have a turn-on voltage so there will be a step in the response. So if you want to control their brightness, do it with current.

Photodiodes can be linear, depending on how you use them. See this article for a simple introduction to their foibles:

I didn't see the OP so I'm responding to this snippet.

You can greatly linearize the transfer function(s) of opto devices by employing a feedback loop. See the datasheet for the Burr-Brown ISO100 for one way of doing this. Art

I like photodiodes. I'm not sure about LED's as sources. How 'bout a laser diode? What's the wavelength?

George H.

Laser diodes are nasty for linearity. They have gross mode jumps as current, temperature, interferance reflections, or, well, anything else changes.

Photodiodes are very linear at reasonable currents.

John

LEDs light output is linear vs current. Some years back I had to build a headamp for a photo sensor. Rather than attemp a simulation of the photo diode, I built a light source to test it. We used a high intensity HP LED because it was fast (10nS turn on / off) and drove it with a 'diamond transistor' from Burr Brown. We got linearity better than 1% and response -3dB at 30 MHz.

G=B2

You should be very carefull about that. Unless you are getting into the low or high currents, the LED output is reasonably linear (~1% ballpark). But beware of the strong nonlinearities at the low or high currents.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

• Not a "step", it is a log relationship.

Tim

Hmm Good point. I should have known better. I was thinking that laser diodes (run way above threshold) do a better job of turning each electron into a photon.

Anyone know the efficiency of LED's? How many electrons are wasted and how many turn into photons?

George H.