Creating <5mW's of 532nm question

[NightExplorer96]

Hey guys,

I have a question in the back of mind I wanted to ask the experts on(LPF). I heard awhile back that creating <5mW of 532nm light is actually more difficult then one might think. My question is as follows:

How come creating <5mW of 532nm light is more difficult then let's say 10mW's? Also, what kind of pump diode is needed to make a 5mW 532nm pointer?

Thank you.

Edit: 2,500th post!!

-Alex

 

[Pman]

Don't know the answers but I'm thinking that if you actually wanted a <5mW 532nm you could probably get it with the modules that have the pots on them such as the ones I have purchased from FastTech that are way overspec. Found that if you adjust the pot for the FastTech ones to about .42mA with 3.2V with a variable DC power supply you get about the max output from them which I've seen as high as 141mW. Never tried going the other way.
I have a few modules left so I could try adjusting the pot down if you ultimately wanted to make your own for a potential cat toy or give away.

 

[Sigurthr]

It's difficult because you're going for a target power, which is not how DPSS pointers are manufactured. They simply make all the pointers the same and due to inconsistency of variables you get different output levels, which are later binned off.

It can certainly be done though. The hard part is also getting a NLO MCA that responds well at low pump powers. Some of those crystals like to be run hot, and the ~80mW of 808nm isn't going to warm anything. You can muck up alignment to get lower output from a higher pump but then you get cruddy beam specs.

 

[upaa27]

Take a 1064 mm diode(uses nd yag as gain) then use a ktp, lbo, etc crystal to put it into its second harmonic(532nm).

This is difficult because of loss from the frequency doubler. Only around 20% is actually 532nm depending on your crystal and the rest is 1064nm. Also these crystals are VERY wavelength dependent. If your frequency doubler says it take 1064nm it means EXACTLY 1064nm. So the pump diode has to be temp. stabilized.

For a pump source for a 10 mw lets say you get 20% out that is 532nm so you need a 50mw 1064nm pump. THIS IS NOT COUNTING IN OTHER FACTORS THOUGH it is just an example

 

[Cyparagon]

It's only because DPSS is unstable, and it is even more unstable in pointer form where there is no temperature regulation. This means even if you design the pointer for 5mW, it can wander from 0-15mW or more. The only way around this is to use optical feedback. The wickedlasers core does this, for example. All green DPSS pointers in the range of about 5-30mW use a 200-300mW pump.

Take a 1064 mm diode...

If your frequency doubler says it take 1064nm it means EXACTLY 1064nm. So the pump diode has to be temp. stabilized.

...so you need a 50mw 1064nm pump.

Almost everything you said is wrong.

1) virtually no green pointers have temperature regulation.
2) the pump is 808nm, not 1064.
3) 20% is kind of an ideal upper limit. Not the case for low powers.
4) KTP does not need exactly 1064 to double. It will take several tens of nm in either direction.
5) In pointers, the 1064nm is typically generated by Nd:YVO4, not a laser diode, and not Nd:YAG.
6) the loss is mostly in the form of heat, not 1064nm light.
7) LBO is not used for green because it is more expensive and less efficient than KTP.
8) wavelength dependence is not as important as you think. Peak absorption is at 808nm, but I've seem pump wavelengths vary from ~800-820nm and still have decent green output.

 

[upaa27]

It's only because DPSS is unstable, and it is even more unstable in pointer form where there is no temperature regulation. This means even if you design the pointer for 5mW, it can wander from 0-15mW or more. The only way around this is to use optical feedback. The wickedlasers core does this, for example. All green DPSS pointers in the range of about 5-30mW use a 200-300mW pump.



Wrong on several accounts.

1) virtually no green pointers have temperature regulation.
2) the pump is 808nm, not 1064.
3) 20% is kind of an ideal upper limit. Not the case for low powers.
4) KTP does not need exactly 1064 to double. It will take several tens of nm in either direction.
5) In pointers, the 1064nm is typically generated by Nd:YVO4, not a laser diode, and not Nd:YAG.
6) the loss is mostly in the form of heat, not 1064nm light.
7) LBO is not used for green because it is more expensive and less efficient than KTP.
8) wavelength dependence is not as important as you think. Peak absorption is at 808nm, but I've seem pump wavelengths vary from ~800-820nm and still have decent green output.

I had said that the pump was an nd yag I never said anything about an electrically pumped gain medium other than the one pumping the yag. Also pretty much anything with neodymium and ytterbium will get 1064 I was just giving an example.

Green pointers don't generate too much heat so the heat sinks compensate for that

I was referring to most crystals when talking about dependency and for max efficiency. Not "hey a small bit of green is coming out."

Didn't know about the heat part. Thanks for letting me know.

 

[NightExplorer96]

Got it, thank you very much for taking the time to answer my question!

-Alex

 

[The Lightning Stalker]

The problem with DPSS is that it is so
temperature dependent. You could set it up
for 5mW at one temperature, and when the
temperature changes, the output drops or
increases. It will need to be temperature
controlled, or like Cyp said, optical
feedback, or both. That is why the Chinese
pens just blast the crystals with as much
light as they can reasonably get away
with. That way it is guaranteed to get at
least SOME output at room temperature.

 

[Pman]

I wonder if someone is actually setting the Chinese pens or not. It doesn't really seem like the modules are as if I remember correctly I've measured them drawing anywhere from .18 up to .38 from likely over 200 of the 3.2V and 5V modules that I had purchased over the last couple years to modify and throw into a wide assortment of hosts. Of course I am just talking about FastTech so it is also likely that their source isn't bothering to bin them.
When they used to sell the 50mW rated ones I'm pretty sure they were more stable than the ones sold now even though maybe 1/3 to 1/4 of the "now" ones (especially the 5V ones) can be set over 100mW. Don't remember a single one that couldn't be set for more output.
The only cheap Chinese laser pens I ever come across now that are not rated above <5mW (but are actually still way over-spec) don't come with a pot anymore. Only the ones rated above the 5mW in the ads have modules that look like the "potted" ones. Couple years ago I swear I remember most had the pot. All about making them as cheaply as possible now.
Will throw in that I honestly don't believe I've ever ran one over 20 seconds straight even when setting the mA so even though the life span has to be reduced none that I can think of have actually died. It's not in me to deliberately kill any laser unless I'm going for pieces/parts to make something else.