How fast can an lpc-826 laser pulse?

[laserbright]

I would like to try using a red diode laser to create very fast laser pulses, with pulse widths of 1 to 5 nanoseconds or less, with pulses occurring at a low duty cycle, like 10 KHz. I need at least 200 mW of power during the pulses. I understand that the average power will be very low because of the low duty cycle. Does anyone know if an lpc-826 laser can give laser pulses this fast? If not, will some other red diode laser give this much power during such short pulses? (My application is to measure the speed of light using the time of flight technique, with a fast photodiode detector circuit that I am separately going to build.) Thanks for any help!
laserbright

 

[mhemling33]

Interesting I know you can get 200mw from an 826 but I have never dealt with pulses. This is something I am going to enjoy reading what others say! Welcome by the way :beer:

 

[Cyparagon]

They were originally used in DVD drives, as you may know. 1x DVD speed is 10.5Mbit/s which is roughly 10.5MHz square wave, ideally. There are 20x burners out there, so there's no reason the diode itself shouldn't be capable of 210MHz which is coincidentally just under 5ns. The tough part is building the supporting circuitry. What project are you working on?

 

[laserbright]

Thanks for the reply. As I mentioned, I am interested in a basic physics experiment - measuring the speed of light. I could use an infrared diode laser or LED, but then it's more difficult to see where the beam is, focus it, etc. Since light travels very fast (about 1 foot per nanosecond or 30 cm/nsec), it is best to use very short pulses for this measurement. From the way that LEDs operate, I would guess that their response time tends to be rather long for a powerful one, perhaps 10 to 20 nsec. Most diode lasers are faster, but it depends on how they are made. I wish I could find a spec sheet that gave a modulation speed or response time, but I haven't found one yet. I may just buy one and try it.

 

[Cyparagon]

One important thing to know: they respond much faster if they have a bias current. This current should be set to just below lasing threshold.

 

[laserbright]

See if this makes sense… As mentioned above by Cyparagon, some diode lasers are capable of 210 MHz operation in a 20X DVD burner. If this is the -3dB bandwidth of the diode laser, then by standard theory, this frequency corresponds to a (1/e, or 63%) response time of 1/(2*pi*f), or 0.76 nsec, with a 10% to 90% response time of about double that, or 1.5 nsec. In this case, at 210 MHz, the laser output would not look like a square wave, but would have quite rounded edges. However, if the DVD burner design is more conservative, then the output of the diode laser at 210 MHz is more like a square wave. In this case, the -3dB frequency is somewhat higher, with a corresponding 10% to 90% response time which is shorter, around 1 nsec or less. My impression is that these LPC-826 diode lasers are being used in current production DVD burners. Thus, it seems likely that their 10% to 90% response time is about 1 nanosecond or perhaps less, as hoped for. Regarding Cyparagon’s other comment about biasing the laser just below threshold, this sounds like a useful thing to do. However, it also means that heat sinking begins to matter. If the threshold current is around 80 mA (?) for operation near room temperature, then biasing it at 60 mA with roughly 2 volts gives 120 mW, which is still fairly modest. It also requires a slightly more complex biasing/pulsing circuit. Comments?

 

[scumbagatheist]

I think we have the winner of the Most Informed Noob award!

 

[ApexProxy]

AFAIK, if you were to bias the diode at below threshold current then you may not even get any output other than an extremely faint glow. Laser diodes exponentially gain output as you hit their threshold current so dipping below it even slightly will hack the output to near nothing and result in a loss of coherency. Also, every diode has a slightly different lasing threshold so testing your chose diode to find its exact threshold would go a long way in helping tune your rig.

This is based off of my observations entirely BTW.

 

[Cyparagon]

You've misunderstood. Bias current is "standby" current. On-current will still be whatever he decides to set it at - 400mA or whatever.

 

[ApexProxy]

I didn't intend to mean anything regarding the normal operational current. The bias current only being the current that's below threshold current in order to speed up the diode's reaction time.

Mainly what I meant was that he won't even attain 120mW of output at below threshold currents since laser diodes emit a very tiny fraction of their full power at anything below threshold. Its an exponential gain of output right as you pass the threshold current so holding just below it would result in nearly no output which as I understand is a good trait for this specific experiment.

 

[Sigurthr]

His 120mW figure was the standby dissipation prediction; used to determine heatsinking requirements.

 

[laserbright]

First, I appreciate the feedback. Secondly, I understand that, even a little below the threshold current, the optical output is at extremely low power, and the device is no longer lasing - it acts like an LED, with a relatively broad optical output spectrum. The 120 mW is only referring to thermal dissipation, not to optical output. Only a tiny fraction of this would go into LED-like optical emission. For an edge-emitting diode laser like this operating below threshold current, most of the LED-like optical output is reabsorbed, and turns into heat. If this were a VCSEL (vertical cavity surface emitting laser), then the LED-like emission below threshold would be more efficiently emitted. I'm frankly unsure how large the benefit is to biasing the diode laser a little below threshold. For one thing, the pulse generating circuitry then has to charge up a smaller amount of capacitance, which may help. This may be a little confusing to some, but here goes: The diode laser has a certain capacitance at zero bias, like any diode. As the diode is forward biased to a level just below threshold, some charge from the pulse circuit goes into charging this added capacitance. From that point to the desired lasing current, the capacitance continues to increase as the space charge region of the diode decreases in thickness, and more current from the pulse circuit is needed to charge up this added capacitance. The other benefit to biasing the device a little below threshold is that the diode laser active layer would already have a large electron and hole concentration, just below the point of population inversion which is needed for lasing. Thus, a smaller density of electrons and holes need to be injected into the laser active region by the pulse current to begin the lasing process. However, what really matters is the response time of the laser from the point that it begins to lase until the laser is emitting the desired optical power, and it is not clear to me that this time would be substantially reduced by biasing the device just below threshold. Such biasing would decrease the time from the beginning of the electrical pulse to the beginning of the laser output pulse, but that time does not matter in the experiment I am trying to do. I would be interested to hear reactions to these thoughts... Thanks!