I am attempting to turn white light LEDs on and off at a very high speed. Here is a link to the LED being used.


To measure the rise time of the signal, I am using a New Focus Model 2001 photodiode connected to an oscilloscope. The photodiode is supposed to be able to adequately receive signals of up to 100 KHz+. However, when using a square wave function generator (HP 33120A) on my LED, the signal "shark fins" and takes 30 to 50 microseconds to reach peak value. At about 10 KHz. the signal begins to flatten out, seemingly indicating that the rise and fall time of the LED are possibly slow.

My question is three-fold:

1) Is this LED exhibiting a slow rise / fall time?

2) Do you have any suggestion for a fast (low rise / fall time) LED?

3) Is there a better way to measure the rise time of an LED?

Thank you in advance!

Edit: Some additional info.

Function Generator: HP 33120A

Oscilloscope: Tektronix TDS 210

  • 3
    \$\begingroup\$ White LEDs often operate by having a blue LED illuminate a phosphor. The phosphor will switch much slower than an LED. Do you have a datasheet for your LED? \$\endgroup\$
    – The Photon
    Apr 16, 2018 at 16:32
  • 2
    \$\begingroup\$ "white light LEDs that I have" doesn't really tell us much. Give us a link to the datasheet or we can only make wild guesses. \$\endgroup\$
    – Finbarr
    Apr 16, 2018 at 16:34
  • \$\begingroup\$ Your signal generator does not have very low output impedance. If you measure the voltage across the LED you will probably see the same. \$\endgroup\$
    – τεκ
    Apr 16, 2018 at 16:41
  • \$\begingroup\$ Show oscillogram. Sounds like excessive inductance. \$\endgroup\$
    – winny
    Apr 16, 2018 at 16:46
  • \$\begingroup\$ I have now added a link to the website for the LED. I think most of the specifications you could need are provided there. Thanks for the quick responses. \$\endgroup\$
    – EAC
    Apr 16, 2018 at 16:55

2 Answers 2


The response time is strongly dependent on the phosphor type used.

The YAG type yellow phosphor is pretty fast (should be sub-microsecond) but they may be using a different type of phosphor in that particular LED. There's also no guarantee they will stick with that particular phosphor.

I suggest using a blue filter and compare the response time of the blue driving LED die. If it's still very slow you may have to improve the LED drive circuit. If the speed is much faster you might want to try another type of white LED or use an RGB LED.

  • \$\begingroup\$ Unfortunately, I don't have a blue filter available. However, it seems many of the comments are agreeing that this LED is most likely phosphor based and intended for room lighting, and thus slow. That being said, are there any specific LEDs or types of LED that you can suggest for low switching times? \$\endgroup\$
    – EAC
    Apr 16, 2018 at 17:53
  • \$\begingroup\$ Try a similar blue LED to prove out the system. I have tested generic cool white LEDs and they were quite fast. The warm ones have another phosphor besides the yellow. \$\endgroup\$ Apr 16, 2018 at 17:59
  • \$\begingroup\$ Don't assume you want to use a white OR blue LED, it's not the most efficiently received wavelength for a photodiode. IR LEDs, like in a remote control, are probably a better choice. 850 nm is where the 2001 model peaks. \$\endgroup\$
    – Whit3rd
    Apr 16, 2018 at 20:10
  • \$\begingroup\$ @Whit3rd Blue is the base LED of the white LED so I suggest that. I presume OP wants white because they need visible light. If they don't then IR is probably fine and is more likely to be specified for rise/fall times. Blue sensitivity tends to be low for silicon, as you say. \$\endgroup\$ Apr 16, 2018 at 20:12

This is more of a question than an answer. If my eyes watching TV and I am watching a 30 frames per second up to 120 frames per second than what is the application with an LED switching at 100KHz. What you have selected is a LED and the 10 K Hz is at the high end. Just a hint, you don't see a frequency rating on your LED specification.

To drive it at a higher frequency you need to pick the best operation point. I imagine that the forward voltage have to have to be less than 3.5 volts. At 3.5 volts + you need to apply a negative voltage to remove the residual charges from the diode.

We normally do this with opto couplers where we isolate a signal and we have to use it in a special configuration to get it pass 10 KHZ. (diode to diode) until they came out with photo optoisolators (Sorry this is outside scope of question)

  • 2
    \$\begingroup\$ So post it as your own question. This is a Q&A site, not a discussion forum. Answers belong in the answer section, clarification questions belong in the comments section. Apart from that, ask your own question and if needed, link to this for context. \$\endgroup\$
    – MrGerber
    Apr 17, 2018 at 5:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.