Question about laser diode power
Peter Walsh wrote 08/07/2019 at 18:20 • 0 pointsI'm considering a speed-of-light experiment using the diode from a cheap laser pointer to generate ns pulses of light using a transistor in avalanche mode.
The experiment would benefit from having powerful pulses, and the avalanche mode circuit allows me to choose the energy of each pulse by appropriate choice of capacitor.
I was wondering if there is any reason this wouldn't work, *assuming* the average power is within the laser diode spec, and *assuming* no other parameters are exceeded such as spark-gap distance or capacitor voltage rating.
For example, I am considering 12,000 pulses/sec at a peak power of 100 watts per ns pulse, which averages 1.2 mW, which is well within the 5mW rating of the diode with some safety overhead. (The avalanche voltage needed to generate pulses is around 180 volts, so 12,000 times a second a capacitor charged to 180 volts would dump energy into the laser diode.)
Is there any reason why this wouldn't work using cheap laser-pointer diodes?
(I need Red, Green, and Blue pulses to verify that the results aren't wavelength dependent).
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.
I used to design switching power supplies for a living, so I dealt with these issues a lot. Even though the power averages out to a low value, the thermal mass of the diode may be such that even with a pulse width of 80uSec (~12,000 pps) the individual pulses cause sufficient thermal stress on the semiconductor die to overheat it to the point of damage. There is also the issue of thermal shock, as the die rapidly increases and decreases in temperature between pulses. None of this is good for semiconductors.
Consider this: as your pulse begins the semiconductor die starts to generate heat. The thermal impedance between the die and the part’s package is constant and so provides resistance to dissipating that heat, and the die temperature rapidly increases. The longer the pulse continues the higher the temperature of the die becomes, even while the part’s package is still almost at ambient. If the pulse goes on long enough it raises the temperature of the die beyond its design limit and causes damage or destruction. This phenomena is independent from the parts average power handling capability.
Power semiconductors (bipolar transistors,MOSFET, IGBT, diodes) are often specified with a transient thermal impedance curve (Zth) in the form of a graph which provides information as to how much power the device can dissipate given a pulse of a given width and a given duty cycle. It is a result of the thermal mass of the die, the thermal impedance of the die-to-package, and the maximum safe temperature of the die.
If you can find the data sheet for your diode it may include a transient thermal impedance graph. If it does not you can probably contact the manufacturer and ask for one. They would also be able to answer any questions about the saturation point of the laser, as Xasin mentioned previously.
Are you sure? yes | no
Thanks - exactly the information I was looking for. Will probably save me several tens of hours not having to chase the problem. I'll use high-power diodes and design for transient thermal impedance. (I swear - half of all research is figuring out the correct term to use!)
Are you sure? yes | no
Good article about avalanche transistors (rus lang). https://kit-e.ru/articles/elcomp/2010_08_49.php
Are you sure? yes | no
I'm not entirely sure where you would get the Laserpointer diode from, but it might not be internally set up to handle these kinds of pulses. For example, parasitic capacitance could be a problem.
Also, keep in mind that the physical phenomenon that generates laser pulses doesn't infinitely scale up. The emitting material that sends out the coherent photons might get saturated, limiting output power. The charging of the gas might take a few ns to us, limiting the rate at which the laser can increase or decrease its output. Most importantly, the pump photon source that gets the emitting material into a charged state will almost definitely have a maximum output, internal resistances, and other factors.
I'm just not that confident that a cheap laser will have the necessary internal construction to support such high power, very finely defined pulses.
Do try it out though - after all, as you said, it's just a cheap laser diode, so if it breaks it'd not be a great loss, but if it DOES work it'd make for a really awesome HackADay entry!!
Also, why do you need different color pulses?
Varying the distance the light would travel should be a very strong indicator of a speed-dependent measurement, and even though there may very well be a wavelength-dependent factor, that's acceptable (plus, this wavelength dependence will always be present unless working in a vacuum, so ... Yeah >.>)
Are you sure? yes | no
These are my feelings on the diode as well, but definitely agree that I'd love to see what comes of the attempt.
Are you sure? yes | no
I want different wavelength pulses to do the experiment simultaneously with three separate initial conditions using the same path. The pulses need to be combined to take the same path, and then split at the end for individual measurement. I figured different wavelengths would be an easy way to accomplish this.
Are you sure? yes | no
That would actually be pretty interesting. I have no idea if it would. If you feel like you've done enough research, indicating that there's no information about this, then go ahead! Even if it wouldn't work with diodes, you could still use the same device on neon lamps and get some finding in that respect.
Are you sure? yes | no