PWM drivers for lasers. Why/why not?

[LSRFAQ]

I still wonder why nobody makes curret regulation via µC
It's more precise than turning a pot and you can permanentyl regulate it by measuring a shunt over the ADC...

Because, for the most part, UCs need, in some cases, more time to react to a spike then a linear driver. Linear drivers are cheap, tend not to spike, and tend to start up 100% of the time, compared to a uC that does not always reset cleanly.

Steve

 

[pullbangdead]

There are ways to reduce ringing and ripple. It really just depends on how good is "good enough", but you can make this as much of an adventure as you want it to be. Transmission lines, impedance matching, actual laser behavior, it gets complicated. The test set up I use routinely is a multi-thousand dollar setup with a bench-top pulse generating supply, and we still battle these factors to reduce ringing constantly. It's a never-ending battle.

Also, laser diodes will behave differently, especially at turn-on, than the Si diodes. The Si diodes make a fine enough test load CW/steady state, but they won't behave quite the same under a transient. How different? Eh, who can say without some real work, but there's one sure way to find out. Just don't use an expensive diode for the first test.

 

[ossumguywill]

You let the blue smoke out? Careful, transistors are made out of blue smoke Also, does your driver have any way to determine how long it's been off, or does there have to be a constant power supply for mode switching?

 

[pullbangdead]

You let the blue smoke out? Careful, transistors are made out of blue smoke Also, does your driver have any way to determine how long it's been off, or does there have to be a constant power supply for mode switching?

Are you talking to me or rev0?

 

[ossumguywill]

rev0. Do you have a driver as well? :thinking:

 

[rev0]

You let the blue smoke out? Careful, transistors are made out of blue smoke Also, does your driver have any way to determine how long it's been off, or does there have to be a constant power supply for mode switching?

It doesn't. My driver has a pushbutton that works only when it's on, I'm going to have to drill a hole in my host for the pushbutton, and change modes only when it's on. Good thing is I can configure it to act any way I like since it'll be MCU controlled. I didn't go the click-to-change-modes route.

Thanks Steve and pullbangdead for the info on how lasers act. It's been sorta tough for me to figure that part out, couldn't really find any good spice simulation models of laser diodes either, and I'm pretty new to it so I don't know too well how they work. I think I might try some testing at lower currents with a bluray diode or my 445nm diode, all I have to do is change the feedback resistors. Hopefully the low current will provide enough of a buffer zone even with overshoot at power-up. Worst comes to worst I can make a linear driver on another PCB to stick in line with the switcher, or get myself a LASORB.

 

[ossumguywill]

Good luck
If you don't mind posting your code, I'd like to look at it (unless it's secret or something). I've done far too little with the attinyx4 series.

 

[rev0]

Good luck
If you don't mind posting your code, I'd like to look at it (unless it's secret or something). I've done far too little with the attinyx4 series.

Well there's no code to show yet.. I honestly sorta hate programming so I procrastinate with it. Depending on how this project goes, i.e. if the driver is safe enough to use with 445nm diodes, I'd like to sell a few, in which case I probably wouldn't post my code in its entirety, though I'd be happy to help you out, and I'm doing a writeup currently on an older project of mine which used an ATmega88PA (previously an ATtiny88). I got a steal on the 44A's, a little over $1 each for 25 of them from Digikey, but they've since raised the price. It's a great micro though, lots of features.

 

[ossumguywill]

Yeah I have a spare 24, the only reason I ask for code is because I never use it LOL. I would have done the project with a tinyx5 or a tinyx313 simply because I'm more familiar with them... but I want to get a hang of the registers on all of these things. Also, if you're using PWM on OC0B or OC1B there's a special trick you have to use FYI

 

[rev0]

Update: I'm abandoning my project, I forgot that my probe was in 1x mode, I changed it back and changed feedback resistors to an output of ~300mA, and it rings/ripples like heck, probably about 50-80% of the full scale current, and at high frequency. I really don't know what to trust (regarding using an oscilloscope to measure current indirectly), I don't have nearly enough electronics experience, but I'm just going to play it safe and say that using a direct switcher driver is a bad idea. At least I didn't have to kill any 445nm diodes or invest a bunch of money in parts/PCBs to figure it out. Thanks to those who helped anyways!

 

[ossumguywill]

Bummer.... you've tried a bigger cap across the output and stuff? How are you switching it? What do you mean by "direct switcher driver"?

 

[rev0]

I could try more caps on the output. Right now there's just the recommended ~20uF ceramic capacitor for the boost IC.

By direct switcher I mean I'm only using a constant current boost LED driver IC, no linear driver after it.

 

[rev0]

More caps?!


No change.

Pretty impressive tower of SMD caps, no? I got 18uF, 4.7uF, 1.0uF, 0.1uF, 10nF, 150pF, and 47pF.

 

[FireMyLaser]

Before the caps, try putting a schottky diode in series with the load and maybe also put on a little large electrolytic cap?

 

[rev0]

Before the caps, try putting a schottky diode in series with the load and maybe also put on a little large electrolytic cap?

Another schottky diode?

(In case you didn't know what a boost driver schematic looks like, this is the prototype PCB's):

Interestingly, a 330uF electrolytic cap made things much worse:

It would probably help if I had a bit more education on switching power supplies and such.

 

[ossumguywill]

I still haven't figured out how an inductor works lol, I have no idea how it boosts voltage, but considering that more caps made it worse have you considered taking off all caps? Lol it's worth a shot...