I am currently using the EPC9154 Eval Board to create quick pulses of light (around 2ns pulse width). So far, I’ve been using a blue LED just to get used to the board, but I am switching to a laser diode. I am trying to find one that emits blue light (preferably in the range of 400-415nm) and would produce on the order of 10**7 photons per nanosecond. I’m mostly concerned with the timing stability of the pulses. My grad mentor gave me the D405-20 laser diode, but I fear that may not work. Does anyone know if that diode should be okay or maybe have suggestions for other diodes?
Greetings,
How to go about finding answers to such questions is much of the point of graduate work; a large portion of the burden in engineering/design fields is determining whether or not a particular device is suitable for the purposes one intends.
Much of that chore involves articulating specifically what “be okay” means in context of your application. Criteria for wavelength, photon flux, and timing are mentioned but could use some elaboration. Does wavelength stability matter? How is the flux requirement to be defined/measured? What exactly is meant by “timing stability” and what’s the tie-in with the flux requirement?
A good sniff check on the flux requirement would be to translate the optical power output spec given for the device into photon flux terms (or vice-versa) for comparison. A decimal place’s worth of difference between spec and requirement can give a person good confidence one way or the other.
If one cares about wavelength, one can expect that it’s going to shift with device temperature and probably drive current. To what extent? Does it matter to you? Questions to be answered.
“Timing stability” can mean subtly different things. Is it rise time one’s concerned about? Pulse width? Turn-on/off delays? What quantitative limits apply? One can expect that the device isn’t going to be a perfect current-to-photon transducer at nanosecond time scales, but then again even the interconnect between driver and laser has a say in the matter when one starts looking closely enough.
Hi!
Thanks for the quick reply. I’d like to preface this by saying I am a physics undergrad and know basically nothing about technology, so sorry in advance I’m not providing enough detail, or it takes a bit to explain things to me. Wavelength can fluctuate a bit, but I really do need a diode that stays between 405-415ish nm. The diode I’m using so far is definitely good wavelength and power wise (although tbh the flux is less important although I’d prefer it to be decently consistent). As I said before, I’m mostly concerned with the timing. I need to know that if I send a pulse into the board, that the laser will produce light at some consistent time with like a nanosecond of error. My main concern with my current diode is that the inductance may be too high, and that could mess with my timing. I have not attached and tested this diode yet (that requires a whole process of me designing a dark box and other things so I want to be sure), but when using the LED, only about 40% of my light pulses hit at the same mark (about 30ns after sending a pulse through the input). I know that it being an LED and not a laser diode also probably makes it a bit worse, but I want to be reasonably sure about my diode before I put in the time and money into really testing it. If there are other factors I should be considering for the time stability, or if people have used other laser diodes for this board please let me know. I hope my rambling added some clarification.
Thanks!
Welcome to the club. Turns out none of us know anything if one views the matter as a rates and limits question…
It’d be helpful to quantify “ish” here. Exact interpretations can vary, but a “typical” value is often a mean value measured from the production distribution, with mins and maxes indicating something like a 3-sigma limit plus or minus from there. Under that reading you’d have a 50/50 chance of getting a device with your desired wavelength under the indicated conditions of 25°C case temperature and 20mW output.
If one doesn’t like the wavelength out of the box there’s some potential for tuning by tweaking device temperature, and to a lesser degree drive current. Temperature’s going to change depending on how one runs the diode; continuous wave will burn half a watt and warm things appreciably, 2ns pulses at 0.1% duty cycle, rather less so.
I assume you’re speaking of a time delay when using the terms “hit” and “mark” here but in the physics realm that sort of thing can’t be taken for granted.
That said, how much of the reported result is attributable to the test equipment versus the system under test?
Series inductance has a tendency to smear a waveform out across time; this effect should be repeatable to a very good first approximation. The above result suggests a variation in a measured value from one trial to the next, hinting that either the system’s changing over time or there’s some uncertainty in the measurement apparatus that’s giving rise to the observed result.