Welcome to Laser Pointer Forums - discuss green laser pointers, blue laser pointers, and all types of lasers

Buy Site Supporter Role (remove some ads) | LPF Donations

Links below open in new window

FrozenGate by Avery

pulse drivers

Benm said:
The experiment was intended to find out if it is possible to get more mW per mA on average using pulsed modulation. It turns out that it is, but only at fairly low power levels, mainly because of the diode threshold sort of 'taking the first 70 mA' without resulting in any light.

That said, it doesnt work at higher power levels because the mW/mA flattens off so it only results in a loss from a certain point (see the other thread for details).
Click to expand...

If the slope of the line mW/mA, ie the differential efficiency, is dropping with increasing power while pulsing, then you're not pulsing fast enough (I believe you were doing kHz, so indeed, you weren't pulsing fast enough). Heating effects in a laser take less than a microsecond, so any pulsewidth greater than that isn't helping with heating.

That drop in differential efficiency, the slope of the line, is due to heat. Pulsing a laser removes that effect of the heat, but only if you pulse for short enough times for the heating effects to not have time to happen (wow, what an ugly sentence).

Pulsing will allow you to get more mA/mW, ie increase the slope of the L-I curve, on the high end where roll-off is occurring, if you pulse fast enough to negate the heating effects.
 





Perhaps you are onto something there - i used CW based mW/mA graphs as the source for information, but i'm not sure the drop off in the curve is solely for thermal reasons. It would only work that way if the thermal resistance between die and can is the limiting factor.

Building precise current sources that can be switched in the megahertz range would be a bit of a challenge though. Most common opamps would be unusable to gain-bandwidth or slew rate limitations. I suppose a two-transistor based current source would be fast enough using the right transistors, but those are more difficult to adjust for exact current.

It would be interesting to see if these is something there though - say run a LOC at 200 mA CW or 400 mA 50% duty at a few megahertz.

The downside of it all is that we operate these diodes way beyond spec already. If the challenge were to produce maximum power without exceding manufacturer specs, the pulsed approach definitely has merit. Many people already run these diodes CW at the max current specified for short pulses as it is.
 
No good based on that spec.. thats the frequency it switches the inductor, not the frequency at which the whole thing can or will respond to a change in load.

Its probably slower that linear regulators in that regard, since it requires at the very least one switch cycle to make any adjustmnt.

A transistor based linear power source would be faster if you can use fast transistors. For limited currents, the bc337 series are suitable, with a >100 MHz gain-bandwidth product. A good old BD139 can be fast enough to follow that too. Those things should suffice into the low-MHz range.

If you want to go even faster than that, things get more tricky, especially if you need to be handle a decent amount of current. Using BFR91 and 2N2219/2N3866 transistors would push it to the 10 to 100 MHz range if required.

As a proof of principle giving this a go at 1 MHz would suffice i think.
 
Last edited:
oops! that was a dumb post (mine)... of course that critter doesn't pulse the output in any reasonable way after it goes through the inductor it's no where near square. Sorry 'bout that - my spine was screaming at me so loud this morning that I couldn't hear my brain. So you might say that made me dumb as a post!

DanQ
 
benm im not trying to say a 100mw pulsed at the right frequency is brighter than a 100mw cw but rather a 100mw pulsed being brighter than a 50mw cw
 
Chiming in a little late here:
Perceived brightness would definitely be higher per unit of power in a pulsed laser.

It's a very common practise in the LED back lights in LCDs in low power scenarios. You pulse the LED at 5 times the current rating, but only in a 20% duty cycle, and the perceived brightness is higher. This is scaled down of course, and a brighter back light for the power consumed is attained. It's to do with the eye "remembering" the period of brightness, and ignoring the dull stages. It's generally done in the 100-1000Hz bandwidth.

Explained at the bottom of this page, as well as many other places on the web.
 
THANKYOU AXIT thats exactly what i was trying to clarify. so why dont we use pulsed drivers if we can achieve brighter beams?
 
Its fine in the sub-kHz range really - dead ordinary transistors have gain-bandwidth products in the order of 100 MHz.

I suppose an experiment would be in order here: one diode running 50 mW CW next to an identical diode running 250 mW @ 20% duty cycle. It wouldn't be a problem to match them exactly using a thermal power meter. I seriously wonder if the pulsed one would appear brighter, and if so by how much.

Practical application for pointers is limited anyway: at such low pulse rates, you'd be drawing dots on the wall instead of lines when you move the laser. For laser shows its useless because the scans are so fast it would totally mess up any beamshow.

Personally i dread all those flickering lights. I dont think they appear brighter, but stand out more because they are darn annoying. Especially the retrofitted LED traffic lights here are a nightmare, happily blinking away at 100 Hz.
 
Apologies to all for an obvious misunderstanding on my part at the beginnings of this thread. I never mean to disseminate falsehoods, and always try to admit them if I find out I have.

I am educating myself further whilst reading this thread! :thanks: guys! :pop:

M
:)
 
Ignoring DPSS completely here, that's another challenge all together.

A major problem is that with say our LOC diodes, we're already running them CW at above the manufacturer's pulse rating.

I don't think they'd stand up too well to 5x their rating....
 
aXit said:
Ignoring DPSS completely here, that's another challenge all together.

A major problem is that with say our LOC diodes, we're already running them CW at above the manufacturer's pulse rating.

I don't think they'd stand up too well to 5x their rating....
Click to expand...

EXACTLY.

With an LED, that's no sweat.

But we'd probably very lucky to be able to truly push up to twice CW power, even if you're going down to 1% duty cycle.

And I can't imagine there's much to be gained, even if you can go to to 2x power and 50% duty, which I kind of doubt is you can only go up into kHz, because in kHz you're essentially running much closer to CW than to truly pulsed. With 50% duty in kHz, you're still getting PLENTY of heating, and losing all the other advantages of running pulsed, so I don't know that you'd even be able to do that consistently.
 
That seems to be a problem indeed.

With DPSS the story is entirely different - the doubling process is more efficient with greater power density, and if the pump diode dissipation is the limiting factor, pulsing makes total sense... and it used to be done in pointers.

I think most of them have reached the same limit as direct diode lasers though, the pump is just driven at its peak power continously, and the rest is pot luck with doubling crystal efficiency.
 
im sure someone here on lpf is inteligent to come up with something. id def buy a few, itd be insane to power a bluray at near a watt of pulse power.
 
I think the only possible gains here are higher perceived brightness for a certain power consumption. We're running our diodes hard enough already. And they're already very bright.

I just don't see a market.
 
You must log in or register to reply here.


Back
Top