RoGeorgeI was demonstrating the Constant Current capabilities of my new power supply, Rigol DP832:
NOTE: This picture is from another project, please ignore the DP832 settings from this picture.
My intention was to demonstrate that, if the current is set for (limited to) a 20 mA, the voltage can be set for up to 32 V (which is the maximum for this power supply), without demaging the LED. So I did like this:
- Set the maximum current to 20 mA
- Set the maximum voltage to 32 V
- Hook a LED to alligator clips
- Enable power
The LED started to lit. The power supply indicated a drawn current of 20 mA and a 1.7V voltage drop on the LED. All good.
Then I wanted to hook a blue LED, in order to demonstrate that the voltage drop is dependent of the LED color. I did like this:
- extract the first LED from the alligator clips (while the LED was on)
- pick another LED from the table and attach the first terminal to the alligator clip
- attach the second LED terminal to the other alligator clip
LED FLASHED & POPPED OUT DEAD!!!
The LED was brand new, without any scratches. The strips that looks like scratches on the LEDs plastic body are from the stress during the pop.
WTF? Can a LED pop like that with the power supply set for 20 mA?
The lower side of the junction looks like it was flipped through the transparent material. I have no idea how it was possible. Maybe the plastic has melted during the flash, then the lower side of the junction mirror was flipped with an angle of about 30 degrees from its normal position, then the plastic solidified back. I don't know.
But why did it happened?
Any power supply have a big electrolytic capacitor at it's output. While I was changing the LED for a blue one, the source kept sourcing the 20 mA, but without the LED, all the current charged the power supply output capacitor up to the maximum allowed voltage set, which was 32 V.
When I connected the blue LED, the 32 V rushed through the LED, and all the energy stored in the capacitor passed through the LED junction, destroying it.
Later, I measured the current for the same setup, except this time I just shorted the alligator clips. A spark appeared, but look at the voltage drop on a 0.94 ohms wire:
For a few milliseconds, a huge current passed through the wires, with a peak of more then 18 Amps. That's about a thousand times more than the normal average current for a LED.
3D Printing AVR Arduino Art Audio Automation BeagleBone Bluetooth Cameras Clock Drones Environment Hardware IoT LED Medical Music Radio Raspberry Pi Remote Control Robotics Rockets Satellites Science Security Software Virtual Reality Wearables
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.
Was wondering about this recently... Thanks for doing the research ;)
Are you sure? yes | no
Thanks for the thanks, but I prefer to read your comment like this:
"Thanks for sharing the stupid mistake!"
:o)
Are you sure? yes | no
Read it as you will! But it does seem a bit like you've done the research I've been too lazy to do (and document) into why I keep blowing out laser-diodes. The results came at a good time, I have few remaining to spare!
Are you sure? yes | no
esot.eric, please be advised that the LASER diodes needs a tight _optical_ feedback loop when powering them. Usually, they have 3 terminals: 1 ground, 1 input to power the LASER diode, and 1 output from a _fotodiode_ that reads the actual optical light intensity emitted by the LASER diode. This 3rd output pin is used as a sensor in a current control loop that powers the LASER (2nd pin). Usually, the fotodiode is built under the LASER diode, so it can read the LASER intensity without obturating the LASER output path.
The fotodiode is needed because, in a LASER diode, the emitted light power is very strongly affected by the junction temperature. It is very easy to destroy a LASER diode if you just limit the current, without adjusting this current according to the internal feedback fotodiode.
Of course, cheap and low power LASER diodes might have only 2 pins and no optical feedback. But if you have a 3 pins LASER diode, and you want to achieve the nominal light intensity according to the specs, it's impossible without a very fast optical feedback control loop.
Are you sure? yes | no
Sounds like you've done a bit with laser diodes! Thanks for the heads-up. I had read the photodiodes exist for that purpose, but most circuits I've found are merely current-limiters, ignoring the photodiode (and feedback loop) entirely. I think a big reason for that is that most laser-diodes people tend to mess with are unmarked parts with no documentation... so what setting do we set that feedback loop for? With your commentary, things I've read elsewhere, and the fact those photodiodes seem to be in most systems (even just DVD readers, where they needn't have such tight control over not overpowering it during a read-session of a burnt disc), the whole "use your old DVD laser diode" phenomenon almost seems like a shock that so many people claim to have built working systems at all, let alone long-running ones... actually haven't seen many claims on their systems' longevity... maybe that's why :)
But, I still attribute most of my (much-too-soon) burn-outs to poor-handling, (aka "stupidity") on my part... The first I'm sure was due to the same phenomenon you ran into here... I plugged that diode into a powered-up current-source. The last, well, I was driving it off too powerful a current-source, with an easily-bumped potentiometer.
That's great info you've given. Do you mind if I paste it in my project-logs for future reference?
Are you sure? yes | no
No problem, but keep in mind that I did not experimented with LASER diodes. The info was quoted from memory. I red about it long time ago, in a datasheet (or application note - don't remember for sure) of some dedicated ICs for powering LASER diodes.
Are you sure? yes | no