Following on my previous post I have Photodiode connected to OpAmp in inverting transimpedance configuration. (Circuit diagram can be found in previous post).
Excitation source is 4000K LED with unknown bandwidth. datasheet
Photodiode in operation is BPX61 datasheet.
OpAmp in operation is LTC6228...
4000K means it's a white LED. Does it have yellow phosphorus of unknown decay time? The phosphorus gets excited from the blue LED die and emits yellow, making a white LED. The phosphorus keeps glowing after the power is off. Which is why white LEDs make poor data transmitters with low bandwidth.
A more interesting question would be "what is the rise time of the LED excitation source"? Your LED is a luminescence low-pass filter after a few kilohertz. Might actually be less - depends on phosphor luminescence's decay time.
@ LifeisBeutiful The photodiode has a bandwidth for a given reverse bias. The LED, especially a white one is dicey - it will have a pseudo-bandwidth that is wavelength-dependent.
@LifeIsBeatiful What exactly are you trying to measure. And don't say "bandwidth". It doesn't really make sense in this context. Maybe you mean "risetime" or "decay time"? Possibly "emission half-life"?
@LifeIsBeatiful There's a communication problem here: you're asking for a property that doesn't really apply in the application. It would be like asking: "what speed is the color red?" How about explaining why you are trying to measure what you're measuring? AFAIK, that waveform could be an artifact from your excitation source.
@MOSFET It's for optical communication purposes. I'd like to find out what is the modulation bandwidth of LED. Photodiode circuitry has more than 10Mhz bandwidth however these LEDs are only known for up to couple of Mhz. that is the point that I'm trying to measure.
@LifeIsBeatiful That's an old 1976 article. It even predates when I was working on 2 GHz optical communications over fiber in 1982 as a Tek employee. It's ancient news. Systems are so much more advanced now. We weren't even considering multiplexing fiber Bragg gratings back then. But today? What is the real question?
@LifeIsBeatiful First, measure your excitation source to make sure you are starting with a clean, sharp LED drive signal. Second, your LED of choice is inappropriate for what you are trying to do. It's like passing light through a "glow-in-the-dark" medium and then trying to measure data on the other side.
@LifeIsBeatiful You must be aware that the photodiode is not very sensitive to the blue LED with narrow spectrum, it is more sensitive to the yellow phosphorus with wide spectrum. Even if the blue LED would be capable of even 1 MHz communication, it takes time for the phosphorus to charge up and it will keep emitting for a while after LED is off. This is low pass frequency or DC bias in the yellow spectrum. The blue may still work at high frequency so it stands out from the yellow "ambient".
@LifeIsBeatiful I kind of figured it was something more like that. And I see an understanding of Bragg is needed to read it. At least I feel as though you have narrowed down the question, though there are still several options to investigate. I'm definitely leaving this one for others. No time to re-educate myself and pull up to date to even suggest something useful. But I do think your question needs significant improvements, now.
@MOSFET it's how people measured the modulation bandwidth of their LEDs. and also some well known academic instituions suggested the same setup. make sure photodiode has high enough 3dB bandwidth. then measure modulation bandiwdth of led in the same way by sweeping the frequency. scirp.org/pdf/OPJ_2013071911271300.pdf
@LifeIsBeatiful post the oscillogram of the transimpedance output, the voltage across the function generator (exciter?), and, if possible, the voltage across the LED all at the same time. Let's do this at a super low speed. 10Hz, 20Hz. even 100Hz will be fine.
As soon as I saw the white LED I knew the comments would be full of people who think the "phosphor" on a white LED is phosphorescent and therefore will glow for a long time after excitation. They're not actually phosphorescent (they would quickly saturate if they were), they're fluorescent, and they have a lifetime of a few tens to maybe 50 or 60 ns. 1 or 2 MHz modulation of a white LED is absolutely possible. Beyond that becomes tricky depending on the exact diode. If you're getting KHz modulation bandwidth, that is an electrical problem, not optical.
From the asymmetric waveform, something is wrong with your circuit. Since you have a negative rail on your opamp, my guess is that you are not driving the LED correctly. What is your drive waveform? Are you applying a DC offset to keep the LED forward biased as you modulate it?
Discussion on question by LifeIsBeati…
Discussion on question by LifeIsBeati…