Measuring laser diode electrical characteristics

Hi there!! :wave:

I'm hoping someone here can provide me with some guidance. I have a few laser diodes (blue-ray, red and IR) and i'm trying to measure the rise and fall times of each, and the I/P curve (Bought them off ebay, and datasheets look dodgy). So, Questions....

1. I am using an optical power meter with photodiode sensors to measure the optical power. Is there a general standard used when measuring the laser power? (i.e laser distance from the sensor). what would be a general rule of thumb?

2. I am trying to measure the rise and fall times of the laser diodes using a photodiode (rise time 47 ps, fall time 246 ps), and i'm a bit stumped. My instinct would be to apply a low frequency pulse to the laser diode and gradually increase this, while observing output of the photodiode, but the transients of my rudimentary frequency generator could blow the diode, and I would very very much rather avoid this. Any suggestions?

If you need me to clarify anything please let me know! All help would be greatly appreciated!

tuk2amy said:

Hi there!! :wave:

I'm hoping someone here can provide me with some guidance. I have a few laser diodes (blue-ray, red and IR) and i'm trying to measure the rise and fall times of each, and the I/P curve (Bought them off ebay, and datasheets look dodgy). So, Questions....

1. I am using an optical power meter with photodiode sensors to measure the optical power. Is there a general standard used when measuring the laser power? (i.e laser distance from the sensor). what would be a general rule of thumb?

2. I am trying to measure the rise and fall times of the laser diodes using a photodiode (rise time 47 ps, fall time 246 ps), and i'm a bit stumped. My instinct would be to apply a low frequency pulse to the laser diode and gradually increase this, while observing output of the photodiode, but the transients of my rudimentary frequency generator could blow the diode, and I would very very much rather avoid this. Any suggestions?

If you need me to clarify anything please let me know! All help would be greatly appreciated!

Click to expand...

You have your work cut out for you.

First and foremost, even just for the CW L-I curve, you need a calibration number for each wavelength you want to measure with the photodiode. You also need to make sure your photodiode will even work for all your wavelengths (it's not uncommon to use different photodiodes for different wavelength ranges in order to get the numbers out that you want, different materials perform differently over different wavelenght ranges). Then you need to do a lot of figuring to make sure your photodiode can measure the things you want it to measure. You want to measure transients very fast, so you need to bias your detector, but that messes with other things like noise. You need to do a lot of reading to make sure your photodiode will tell you what you want it to tell you. I have a hard time believing that your photodiode will truly have a 47ps rise time in the real world when measuring a laser diode. Even if it is true, a fast rise-time also means small area, so you're going to have a hard time getting all the light onto the detector, and the detector is going to saturate VERY easily, so you likely need ND filters and such.

And that's all just getting you optical measurement to work like you want, then you still have to deal with the problems of driving your laser diode. It's not trivial in the least.


You have a very complicated problem on your hands if you want to do it right. How precise are you trying to be here?

You have your work cut out for you

Click to expand...

Don't I know it.

Could you explain what you mean by a calibration number? What do I need it for?
I have made sure that the photodiode is suitable for the wavelengths I'm dealing with (according to its datasheet). I Intend to bias the photodiode, and will be using an RC filter to try to reduce the noise. I have attached an image of the circuit I intend to use


My main issue here, is driving the laser diode. I am currently using a circuit off Sam's Laser FAQ's to power them in CW mode. I'm trying to figure out a safe way to drive them in pulsed mode without frying the diodes.

How precise are you trying to be here?

Click to expand...

I'm trying to be reasonably precise. I'm taking these measurements to ensure I don't push the diodes above their limit, so any results that give a reasonable estimate in that area will do. I'm more interested in designing a set-up that will give consistent results for all diodes, to allow for comparison. (I have 2 of each diode).

I suppose there are several options for the rise/fall time measurement.

One way to do it, would be constructing a pusled current source, that sends pulses a bit longer that the expected rise and fall times, with a duty cycle of 10% or so.

In this case this would mean a pulse width of 1 nanosecond, with 9 nanosecond gaps in between. Constructing a pulse generator that does this will be a challenge in itself, as you are dealing with GHz speeds. To get decently square pulses that are just 1 nanosecond wide, your components would have to be rated 10 Ghz+.

And then there is the actual measurement: You would need a bloody fast scope, and monitor diode current over a sense resistor on one trace, and the phototransistor output at the other.

With all that set up, you still need to know the characteristics of your photodiode, so you can determine how much of the delay is due to the sensor diode, and how much is due to the laser diode.

Sounds like an expensive endeavor to me.

Benm said:

I suppose there are several options for the rise/fall time measurement.

One way to do it, would be constructing a pusled current source, that sends pulses a bit longer that the expected rise and fall times, with a duty cycle of 10% or so.

Click to expand...

Yeah, thats one way. Any ideas on the easiest/quickest way to achieve this? I considered using a 555 timer, which will allow me to modify the frequency easy enough, but it has slow response times ~100ns.

Benm said:

In this case this would mean a pulse width of 1 nanosecond, with 9 nanosecond gaps in between.

Click to expand...

I would be absolutely delighted if my laser diodes had rise times of 1ns! However I'm not that optimistic! I currently have a pulse generator (borrowed) that gives 5ns width pulses with varying duty cycles. (can't vary the pulse width.) Here is a sample of the pulse output on an oscilloscope.


Problems? Not a square wave. 2, I highly doubt the laser diodes are that fast, and I can't vary the pulse width of the pulse generator, just the repetition rate (and thus the duty cycle).


I should be able to source a fast oscilloscope, if the one I currently have isn't fast enough

With all that set up, you still need to know the characteristics of your photodiode

Click to expand...

Apart from those given in the datasheet? Any suggestions on how I could go about characterising the photodiode?

Benm said:

Sounds like an expensive endeavor to me.

Click to expand...

Not if you have access (albeit temporarily) to the right resources

Well you have put more into it than your first post led me to believe. Sorry for underestimating you at first, as most of the posts from newbies around here are truly from newbies, people who have no idea what they're doing.

For your photodiode, you need the conversion A/W, as in how much current comes out per light power in. This changes with wavelength. The datasheet for the photodiode will probably have a curve, but I don't put too much stock into those curves after some of the ones I've used previously.

As far as your drive source, you've got some major ringing there, and it'll only get worse with a laser. ESPECIALLY if you're going to be measuring different types of laser diodes. Is that a voltage or current in your image? I'm sure you already know you probably need inductive current probe or the like, that measuring current in series isn't a good idea with this. Your impedance is going to change with every laser, and very well could cause the whole thing to ring out of control. If you want short pulses, you're better off with a voltage-source driver than a current-source driver (I'll leave "why?" as an exercise to the reader). But you don't necessarily need very short pulses as far as I can tell, you need clean rise and fall, which is also hard to do. Your circuit needs to be carefully designed to get clean rise and fall out of a laser diode, and it'll be different when going from like a 405 diode to a red one, you'll need a whole different circuit.


And I still have trouble with the idea of your photodiode actually turning on that fast. I'm afraid you're going to be disappointed if you're expecting it to just work. Photodiode data sheets are kind of notorious for skewing how they present the information to you, like using log plots and the like, and things like giving you a single rise-time number as "the spec" even though that is only encountered under certain, perfect conditions, conditions which you're not likely to be using. Even wavelength can come into play with risetime.

pullbangdead said:

Well you have put more into it than your first post led me to believe. Sorry for underestimating you at first, as most of the posts from newbies around here are truly from newbies, people who have no idea what they're doing.

Click to expand...

Ah no worries pbd! As a female engineer i'm used to being underestimated . I do really appreciate your assistance though. I'm fairly new to lasers in general though, so must of the stuff i'm doing have been via a lot of research, and I've pretty much hit a dead end so far, so here I am! ( I have been a lurker for some time now though. :angel

pullbangdead said:

For your photodiode, you need the conversion A/W, as in how much current comes out per light power in. This changes with wavelength. The datasheet for the photodiode will probably have a curve, but I don't put too much stock into those curves after some of the ones I've used previously.

Click to expand...

Ah! The responsivity. Gotcha. It is in included in the datasheet, but I am sorely tempted to ask the manufacturers for the actual data (if they're willing to release it) as the curves supplied are.....lets say....unsatisfactory. but I'm not trying to re-invent the wheel here, because that will take far too much time, and effort. Presuming there is a margin of error in the A/W curve supplied, as long as this error is relatively acceptable, and consistent. Then it at the very least gives me something to work with, and I can tweak as required as soon as I can actually observe some sort of a signal from the photodiode.

pullbangdead said:

As far as your drive source, you've got some major ringing there, and it'll only get worse with a laser. ESPECIALLY if you're going to be measuring different types of laser diodes.

Click to expand...

I agree completely. Plan is to use a power diode, or something similar to counteract the ringing effect.

pullbangdead said:

Is that a voltage or current in your image?

Click to expand...

Voltage.

pullbangdead said:

I'm sure you already know you probably need inductive current probe or the like, that measuring current in series isn't a good idea with this.

Click to expand...

Instead of measuring current in series, I'll be measuring voltage across a load resistor and determining the current using ohms law. Its a round about method, but a less intrusive approach. Again, this will most probably have its drawbacks, and these are yet to be seen.

pullbangdead said:

Your impedance is going to change with every laser, and very well could cause the whole thing to ring out of control. If you want short pulses, you're better off with a voltage-source driver than a current-source driver (I'll leave "why?" as an exercise to the reader).

Click to expand...

Agreed. I initially went for a current source driver as a quick and easy way of limiting the current to the laser diode as to the best of my knowledge, its over-current, and not over-voltage that destroys laser diodes.

pullbangdead said:

But you don't necessarily need very short pulses as far as I can tell, you need clean rise and fall, which is also hard to do. Your circuit needs to be carefully designed to get clean rise and fall out of a laser diode, and it'll be different when going from like a 405 diode to a red one, you'll need a whole different circuit.

Click to expand...

Excellent. Any ideas/suggestions on how such a circuit would be designed? Even in its most basic form. My main issue at the moment is finding a way to drive the laser diodes in pulsed mode. I currently have two options:

1. A TTL frequency generator with clean(to a reasonable degree) square waves. Any suggestion on designing a circuit that will allow me to power the diode using this?

2. Using a 555 timer to generate square wave pulses. I have some experience with 555's so I should be able to tweak the circuit to obtain as clean a square wave as possible. Problems: Will the pulses be clean enough even then?, Fast enough pulses?, How to design circuit to drive laser diode adequately?

pullbangdead said:

And I still have trouble with the idea of your photodiode actually turning on that fast. I'm afraid you're going to be disappointed if you're expecting it to just work.

Click to expand...

If I was expecting it to just work, I wouldn't be here . I'm really looking for some advice that will send me in the right direction, as opposed to just stabbing in the dark. Nothing worse than realising that you've been putting a lot of effort doing the wrong thing. (Been there, done that.) I am also doubtful that the photodiode switches that quickly. But even a 470ps rise time should be fast enough? I have no idea what typical risetimes for laser diode are, but I'm presuming tens of ns? (any thoughts?)

pullbangdead said:

Photodiode data sheets are kind of notorious for skewing how they present the information to you, like using log plots and the like, and things like giving you a single rise-time number as "the spec" even though that is only encountered under certain, perfect conditions, conditions which you're not likely to be using. Even wavelength can come into play with risetime.

Click to expand...

In my experience, all datasheets in general are guilty of this, so I always allow for a good margin of error when reading datasheet values. I have attached specs of the photodiode, frequency generator and red laser diode, to give an idea of what I'm working with. (Now that I think of it, the red laser diode is a Sony-JL, and I've seen it pop around a lot here. If anyone has already measured specs of it, links would be awesome!)

Any and all help will be hugely appreciated! Thanks!!

Did I scare everyone off? Problem still persists ! I'm sure that there are lots of people around whose laser expertise would be very helpful! oke: