Laser diode wont work with certain power supplies

[WarFreak131]

Hey guys,

I bought a 100mW green laser diode (no driver) from some no name company from china. The specs recommended driving it at 250mA. So I plugged it into my variable power supply, and slowly increased the voltage until the supply was delivering 250mA. It worked very well, a nice well defined green beam.

I also have an LPC 826 with a DDL driver circuit with a 25 ohm pot. I have successfully powered that circuit using a fixed power supply that I built (which has settings for 5,6,9, and 12V, which is several more volts than I'd need). The LPC 826 worked in that circuit with no problems. Then I plugged my green diode into the DDL driver and started it out at the max resistance to prevent too much current. I slowly adjusted the resistance down, but no beam ever came out.

Uh oh, I thought I pumped too much current through it and broke it.

But then I plugged the green diode back into the variable supply... and it works fine.

So I thought maybe my fixed supply was busted. I brought it to work with me, and used a high end variable power supply. I plugged the diode into my DDL circuit, and it didn't work again. So I thought my LM317 was broken. Nope, still working just fine, and yes, it has a heat sink on it. Then I decided to plug the laser straight into the high end power supply and make it deliver 250mA.

And it doesn't work....!

 

[LarryDFW]

Some power supplies can have a surge when 1st turned on.

Also green lasers are sensitive to temperature variations.

LarryDFW

 

[WarFreak131]

lol, scratch that, its completely busted. The beam is no longer green, its a very very weak red.

 

[WarFreak131]

Both the first supply and the supply I built are regulated. A spike shouldn't have happened. What are the odds that it's just some cheap Chinese p.o.s. and crapped out?

 

[Matrix]

that happens to green lasers alot and with the tempature problems with them they are probably the most sensitive lasers on the market

 

[Zeebit]

I hope you did not look at the faint red glow. That is the IR pump diode. You have busted either the crystals or the diode itself. Becareful, it could still be putting out several mW of invisible light that is just as dangerous as visible laser

 

[WarFreak131]

I didn't look straight into it. I looked at it from an angle.

On a related note, I hear people saying that they order diodes from CNI. I can get to the CNI website, but how do I order?

 

[Zeebit]

I don't think CNI sells in retail. They cater to companies in wholesale.

What diode do you want?

 

[WarFreak131]

Nothing specific, I was just browsing around, saw some cool orange and yellow diodes. I would check them out if the price was right.

 

[blasterman]

Fickle isn't the word for green lasers. I can drive 445's on LED drivers with a bit of care. Greens? Still haven't figured those out. Phase of moon needs to be calculated in there as well.

About a month ago I pulled apart Chinese projector where the green wasn't running and tossed the diode aside. On a whim I plugged it into another projector with a working green, and within a few moments noted the smoking wood paneling on a wall 24" away. The supposedly bad laser diode is *easily* stroking 250mW, if not more. I don't have testing gear, but I know from experience when the reflected dot on a white surface starts looking more white than green it's a sign it's cranking some serious mW.

I tested both green channels on the different projectors and they read identical under load (both current and voltage). Diode from working projector has same intensity on both drivers (WTH?). The super intense greenie however runs like a 'scalded dog' on the second projector, and it's been running happily for a dozen + hours with no signs of deterioration. You got me...

 

[Bob_Boyce]

Both the first supply and the supply I built are regulated. A spike shouldn't have happened. What are the odds that it's just some cheap Chinese p.o.s. and crapped out?

Chances are real good that a transient voltage spike killed it, especially if the power supply current was set very close to the current limit of the pump diode. Many lab supplies have been measured to output a very short spike at powerup prior to full regulation kicking in. This may not kill most electronics, but laser diodes are much more prone to damage, and even more so if run close to the point of destruction.

Fickle isn't the word for green lasers. I can drive 445's on LED drivers with a bit of care. Greens? Still haven't figured those out. Phase of moon needs to be calculated in there as well.

About a month ago I pulled apart Chinese projector where the green wasn't running and tossed the diode aside. On a whim I plugged it into another projector with a working green, and within a few moments noted the smoking wood paneling on a wall 24" away. The supposedly bad laser diode is *easily* stroking 250mW, if not more. I don't have testing gear, but I know from experience when the reflected dot on a white surface starts looking more white than green it's a sign it's cranking some serious mW.

I tested both green channels on the different projectors and they read identical under load (both current and voltage). Diode from working projector has same intensity on both drivers (WTH?). The super intense greenie however runs like a 'scalded dog' on the second projector, and it's been running happily for a dozen + hours with no signs of deterioration. You got me...

Even real high powered green DPSS lasers are a little unstable with temperature changes, especially as they warm up. With 25W of pump power, I typically get about 7W of green. I can fine-tune the crystals and tweak it to about 7.5W but it becomes even more sensitive to temperature fluctuation. I don't dare crank the pump up to full rated 50W CW because these large crystals are very hard to find on the surplus market.

Bob

 

[Cyparagon]

Maybe you had polarity backwards?

You should make connections with the power off when running things at constant current. Without a load, the supply will increase the output voltage to your set point. When you connect a load, it doesn't instantly regulate the current; it takes a few microseconds. It may not seem like a long time, but you're essentially discharging a capacitor into the diode and that can be bad for them.

 

[WarFreak131]

Many lab supplies have been measured to output a very short spike at powerup prior to full regulation kicking in. This may not kill most electronics, but laser diodes are much more prone to damage, and even more so if run close to the point of destruction.

But then how can you ever be safe in turning on a diode if the regulator doesn't regulate for enough time to possibly kill a diode?

 

[Bob_Boyce]

But then how can you ever be safe in turning on a diode if the regulator doesn't regulate for enough time to possibly kill a diode?

Using a bench supply that is rated for LED use is one way, they have OVP (overvoltage protection) and most tend to have soft start. Otherwise, build your own or buy a constant current regulator or driver designed for the purpose.

When I'm testing an unknown to me laser diode on the bench supply, I set the coarse voltage control to 1.5V with fine voltage control set to minimum, and set both current limit controls to minimum before connecting the laser diode. After connecting the laser diode, I slowly turn the fine voltage control up. If I see no visible response by the time I reach 1.9V then I stop and check connections, and try to find if I am connected correctly. Exceeding 2V in reverse polarity can destroy a laser diode. I use a CCD camera to check for possible IR emission. Most of the unknowns I have to test are IR. If I get no response with any connection combination at under 2V, then I'll put the laser diode under the microscope and examine the connections to the die and the die itself.

Once I verify emission, I slowly raise the voltage while watching the current and the output from the laser diode, raising my current limit as needed. When I hit threshold, I note the threshold current. Once I have established a good safe operating current above threshold, I can use that as a safe voltage setpoint for connecting and disconnecting the laser diode as often as I need to while the power supply is on. I never set the voltage higher than that required to push the maximum safe current. This prevents damage to the laser diode that could be caused by a capacitive discharge if the laser diode connection goes intermittent.

Do NOT do this for the typical constant current laser driver or LED driver though, as when the laser diode is disconnected, output voltage can rise and charge the output capacitor. When you reconnect the laser diode, that capacitor will discharge into the laser diode and chances are good that it will be destroyed.

I hope this helps.

Bob

 

[WarFreak131]

Using a bench supply that is rated for LED use is one way, they have OVP (overvoltage protection) and most tend to have soft start. Otherwise, build your own or buy a constant current regulator or driver designed for the purpose.

Couldn't this be achieved by throwing in an inductor into the output of a DDL driver? As soon as the voltage spike hits, the inductor will resist the change in voltage/current and slowly loosen the throttle on the v/i allowing it to slowly raise.

Once I verify emission, I slowly raise the voltage while watching the current and the output from the laser diode, raising my current limit as needed. When I hit threshold, I note the threshold current. Once I have established a good safe operating current above threshold, I can use that as a safe voltage setpoint for connecting and disconnecting the laser diode as often as I need to while the power supply is on. I never set the voltage higher than that required to push the maximum safe current. This prevents damage to the laser diode that could be caused by a capacitive discharge if the laser diode connection goes intermittent.

This is nearly the same procedure I take. Beforehand, I take a DMM with a diode checked, and check the voltage drops across the leads. The photodiode is usually a regular .6 - .7 V drop but the Laser is usually around 2.5-3

Then if I want to run the diode directly through the current adjustable supply I have, I connect the diode, turn the supply on with the voltage out at 0 and slowly raise it until the diode starts conducting and the current jumps, and I note that value.

Then I adjust the resistors accordingly in my DDL driver to make it match that value.

 

[robmobz]

I would power the supply up with a lead shorting out the output so that the start up spike dumped through that an then remove that once it has stabilized.