Pulsed lasers -- Any advantages??

Hi all,
 I just measured my red DVD LD power and Bluray LD power yesterday -- Luckily we've got a few Scientech Laser Calorimeters in my lab at MIT - We use them for measuring millimeter wave sources, "Masers".  I don't know when/if they are calibrated right, though, as the numbers came out disappointingly low (see plots below).  I know the Aixiz housing lens has a terrible obstruction in front of the lens, so I took it off and positioned the laser just 0.5" or so from the calorimeter surface so it collected all the light.  I was able to get reproducable precision to within 0.2mW, so accuracy is really my main concern.  Also my red DVD diodes are from Senkat (thanks!).

 Anyway, the purpose of the power measurements was to see if I could gain something by using PULSED power to the LD.  I'm planning a 2kHz or faster pulse rate so the "choppies" can't be see so easily (the TI TPS75105 current regulator I'll be using turns on in just 33us, so 2kHz is easy; careful, it's a 9-pin BGA only 1.2mm square!).  

 According to the red DVD laser plot below, I see the following two scenarios:
1. For a peak LD current of Ipk=250mW, 50% duty cycle pulsed, the average current is Iave=125mW.  For this, I would get about 68mW(peak) or 34mW(ave).  The input power would be Pin=3.2v*250mA=0.8W(peak) and 0.4W(ave), and the efficiency would be Pout/Pin=34mW/0.4W=8.5%.
2. For a CW (100% duty cycle) at I=125mA, V=2.7v and I would get 30mW out.  The efficiency would be Pout/Pin=30mW/(2.7*0.125)=30mW/0.34W=8.8%
 Appearantly, there's not much difference between the two for the DVD diode, which I think is due to the very linear slope of Pout vs. I.  The story is different for the blu-ray...

For the Blu-Ray diode measurements, same scenarios, different numbers:
1. Pulsed: Ipk=50mA at 50% D.C., Iave=25mA, V=4.8v.  Pout(pk)=8.2mW, Pout(ave)=4.1mW.  Efficiency=Pout/Pin=4.1mW/(4.8*0.025)=3.4%.
2. DC: I=25mA, V=4.3v, Pout=1.8mW, efficiency=1.8mW/(4.3*0.025)=1.8%.
 Clearly, I get MORE power and MORE efficiency by pulsing it in this case.  But why wouldn't I just run it at 50mA all the time (yes, lifetime issues, etc.).  Can I pulse it at 70mA and get higher average power than I could using DC at 50mA?  Has anyone tried this?  I think I remember seeing these are "rated" for something like 50mW (peak) output, but my measurement of 8mW is well below 50mW!  (again my calorimeter calibration could be bad, but I don't think it's off by more than a factor of 2).  So I wonder how we achieve the "ratings" of these diodes.

So can I gain anything by pulsing it besides efficiency?  Efficiency itself is a good reason to pulse for a high power laser, but we WANT all the average power we can get!!  The DVD diodes are somehow "rated" for about 250mW pulsed.  Has anyone tried pulsing them to 400mA or 500mA, for example?? How do we maximize the average output power??

I've been doing some pulsed experiments on the GB diode today. In CW, i usually keep current at round 200 mA, which works out fine so far. Therefor i aimed to keep average current at 200 mA, while playing a bit with peak current and duty cycle.

I set up a current source plus pulsed driver, configured the current source to supply 400 mA, and the pulse driver with a variable duty cycle between roughly 2 and 50%.

In terms of visibility (or average output) there is little to be gained, as was expected from the power curves (see attached graph too). 50% duty of 400 mA appears as bright as 200 mA CW. I tried both a low frequency (100 Hz or so), and a higher (10 kHz or so). With 100 Hz you can obviously see its pulsed when you move the dot around, at 10 kHz this isn't really visible anymore.

However, i also did some experiments with just smoldering things, which gave a little bit more interesting results. It's all a bit hard to quantify, but it seems to burn a bit better 50% pulsed at double current. Also, i could get some smoke from plastics on incense at very low duty cycles (5% or so). The average power in such setup is perhaps a bit over 10 mW... not enough to do this in CW.

It might be more thermally stressful to pulse the LD from 0 - x mA versus pulsing while providing some DC current. The 0 - x mA pulses can cause uneven stresses along the length of the cavity along with the high peak power loss on turn-on due to the forward recovery of the LD if the dI/dt is large. A simmering current ensures that the diode is mostly on and can reduce the peak thermal stress during the front porch of the pulse.

Well, i figured these diodes are intended to be used in dvd recorders, where they pulse from zero to full power in a matter of serveral nanoseconds. In that respect, the dI/dt is much lower in my setup, especially since i still have a 2n2 cap across the diode.

It would however be possible to make the zero not entirely zero, but just above the treshold (50 mA) or so, which would 'pre glow' the diode in the same way that's done for incadenscent lightbulbs in lightshows. I have no idea if there is any benefit to diode life though.

Using lower frequencies i guess there is some additional stress due to thermal expansion and contraction with the pulse pattern, but i doubt that will be an issue in the kHz range.

Thanks for the info. It's good to see that Benm got the same result I expected: That doubling the current at 50% d.c. doesn't really increase the apparent intensity for the DVD diodes due to the linear slope. The pulsed power being a better burner is interesting, though. Good to see the power should be more than 100mW rather than the 70mW I measured with my questionable calibration.

About the power levels: The graph above was meausred by Deadal using a power measured intended for lasers like this - i consider it reliable.

Pulsed operation can give good results lower in the curve, for example, 50% at 160 mA would yield 44.6 mW average power, versus only 28.1 mW for 80 mA CW.

Benm said:

Well, i figured these diodes are intended to be used in dvd recorders, where they pulse from zero to full power in a matter of serveral nanoseconds. In that respect, the dI/dt is much lower in my setup, especially since i still have a 2n2 cap across the diode.

Click to expand...

Good point about these LD being pulsed in application. Maybe the thermal stress isn't a big deal with these LDs.