Weird POV effect with 589nm.

[Sta]

all dpss beside 532nm from CNI are usually pulsed for longer duty cycle.

My Spartan 589 isn't pulsed - perhaps it's used to reduce the power.
Is your PGL 561 pulsed?

 

[tomfire1735]

My Spartan 589 isn't pulsed - perhaps it's used to reduce the power.
Is your PGL 561 pulsed?

it only applied to the glp pens.

the pgl is CW.

 

[joeyss]

My Spartan 589 isn't pulsed - perhaps it's used to reduce the power.
Is your PGL 561 pulsed?

Mines a pen and I just checked and it sure as hell is I can see it easily. I'd say 70-90hz. I can even see led night lights at 60hz in the corner of my eye pulsing.


another thing when it's mode hoping or w/e it's doing I swear when it dims and gets bright ther color changes i tried this in my room which is lit by 2 35W daylight CFLs. I hold it on in the styrofoam and I close the box a bit and I see it. Still eats my 200 mah rechargeable CR2 batteries like crazy.

 

[Alaskan]

Any idea how yellow is produced inside that pointer?

I found this on Wikipedia regarding 594:

Yellow DPSSLs use an even more complicated process: An 808 nm pump diode is used to generate 1,064 nm and 1,342 nm light, which are summed in parallel to become 593.5 nm. Due to their complexity, most yellow DPSSLs are only around 1% efficient, and usually more expensive per unit of power.

Another method is to generate 1,064 and 1,319 nm light, which are summed to 589 nm.[4] This process is more efficient, with about 3% of the pump diode's power being converted to yellow light.[5]

 

[Cyparagon]

Any idea how yellow is produced inside that pointer?
...
Another method is to generate 1,064 and 1,319 nm light, which are summed to 589 nm.

https://en.wikipedia.org/wiki/Sum-frequency_generation

The formula for sum-frequency generation (translated into more familiar terms) is 1/λ1 + 1/λ2 = 1/λfinal - since conservation of energy is required, and energy is inversely proportional to wavelength.
1/(1/1064+1/1319) = 589

Mix and match Nd lines for weird new colors.

Second harmonic generation is actually a subset of sum-frequency generation. You'll notice:
1/(1/1064+1/1064) = 532

 

[joeyss]

Whats a formula that has 808>1064>532

Just wondering?

 

[Alaskan]

Different in that case due to using a Neodymium-doped yttrium orthovanadate (Nd:YVO4) crystal to convert 808nm to 1064nm first. I've been looking into how that works just the other day, I mean the physics behind it. Of course doubling (halving the wavelength size) 1064nm produces 532nm, that one is very straight forward, summing was my confusion, now I see what they are doing, thank you.

 

[dubious speculation]

It's more likely that pulsing gives longer duty cycles and reduce the heat.

I fail to see how this would reduce heat if CNI is using a form of PWM to pulse their pumping diodes. An LD running at
2A CW would produce the same heat as an LD running at 2A PWM controlled (regardless of the frequency with
which they are being pulsed, since even the total sum of heat not generated between pulses would be negligible).

Perhaps you could clarify?

 

[CurtisOliver]

Whats a formula that has 808>1064>532

Just wondering?

The conversion of 808 to 1064 isn't a harmonic process, so isn't important to calculate. It is purely based on absorption and emission of the ND YAG/YVO4 crystal.

Cypargons's equation is a simply put example of one that can be used for SHG and SFG calculations.

1/(1/1064+1/1319) = 589

The true equation is λΣ = c/((c/λ1)+(c/λ2)) as f=c/λ and λ=c/f, but Cyparagon's gives the same result.

With a bit of modification it can also be used for 3rd and 4th harmonic generations.

For example with a 1064nm YAG.
2HG = λΣ = c/((c/1064e-9m)+(c/1064e-9m)) = 532e-9m (532nm)
3HG = λΣ = c/((c/1064e-9m)+(c/(1064e-9m/2))) = 354.66-9m (354.66nm)
4HG = λΣ = c/((c/(1064e-9m/2))+(c/(1064e-9m/2))) = 266-9m (266nm)

 

[Alaskan]

Thank you Mr. Oliver, I have a practical question regarding YVO4 and a SHG crystal for doubling, how far off wavelength can the pump diode go from 808nm before reducing the green output from a NL crystal by an appreciable amount?

 

[CurtisOliver]

The more the pump central wavelength meets the highest absorption peak in the ND:YVO4, the more 1064 output you can expect. This will in turn allow more 532nm to be generated, depending on phase matching. You have approximately +/- 2nm before you see serious drop off. I hope this graph helps.

 

[Alaskan]

Thank you, much appreciated, that number just rattled an old memory, I had forgotten. Can you tell me what the major cause is with many of our laser 532nm DPSS pointers to be so temperature temperamental; the laser diode is going too far from center frequency or the NL crystal being outside its optimum operating temperature?

I am guessing the answer is both, but if one is more of a problem than the other, which one? The diode being off wavelength or the NL crystal being too cold or warm?

My understanding is many of these NL crystals will double IR sources to any wavelength which is capable of falling within the visual spectrum, but we are limited due to the IR pump diode wavelengths we can currently choose from, that we either haven't been able to cheaply build them or have not figured out how to do so yet. Do I have that right?

 

[CurtisOliver]

Thank you, much appreciated, that number just rattled an old memory, I had forgotten. Can you tell me what the major cause is with many of our laser 532nm DPSS pointers to be so temperature temperamental; the laser diode is going too far from center frequency or the NL crystal being outside its optimum operating temperature?

With temperature the diodes wavelength will shift slightly, so that is one reason. If the wavelength drops or rises by 1nm, the 1064nm output will change. Also KTP is a crystal that likes to be warm, and its conversion efficiency changes with temperature. In some crystals, temperature affects the phase matching angles. I believe the NLO temperature has the biggest effect. On a cold day, I often have to hold the top end of the laser pointer in my hands for a bit. this gives me a more output power.

I am guessing the answer is both, but if one is more of a problem than the other, which one? The diode being off wavelength or the NL crystal being too cold or warm? My understanding is many of these NL crystals will double any wavelength within the visual spectrum, but we are limited due to the availability of different wavelength coherent light sources, that we just haven't been able to cheaply build them, especially so in regard to diodes.

Most NLO crystals will double lines within its optical transparency. But this again is down to coatings, absorption coefficients, phase matching angles and polarization. This is why exotic wavelengths are hard to come by, due to the complexity and limitations for some DPSS systems. Please bare in mind, most of this is from self research. I could be interpreted some things wrong.

 

[Alaskan]

Thank you again, the reason I am asking is I am trying to determine if purchasing a edit: wavelength locked C-Mount 808.4nm laser diode (from 20 to 30C no more than .5nm error) is worth the extra cost or not, for 4 watts of rated output (two of them PBS cubed to 8 watts), or whether I should use my FAC corrected 15W 808nm C-Mount laser diode instead. I have a 18mm long heat sinked 3mm diameter YVO4 rod I was thinking might be something I can use with a large 9x9x4mm KTP cube I also have to produce 532nm output, but the vanadate rod might be too long, or does that matter? I don't know why this YVO4 rod was built so long.

I'm not meaning to turn the thread into a 532nm discussion, but thought I'd ask here.

 

[mojo_1234]

Hi all,
do have an older CNI 473nm pointer. The 808nm diode is pulsed with several kHz. Surely has to do with thermal issues. The conversion of 946nm of the YAG is also more critical as doubling 1064nm using a KTP and also highly temperature dependent. And to produce the 589nm output to conversion is even more complex.

 

[CurtisOliver]

What specification does your 15W C-Mount have. What is the spectral tolerance? I have a Coherent FAP-800 50W that is rated +/- 3nm. This was used in their Coherent Verdi systems and would of been likely to of produced 10-15W of 532nm. If this was +/- 3nm, then I gather your 15W would be acceptable. I can only imagine a 20-30% increase in 1064, for the phase matched 4W. If we presume that 50% of the 808 is converted to 1064, and 60% is converted to 532 then the figures will probably be like:
808 +/- 3nm(15W) --> 1064 (7.5W) --> 532 (4.5W) This is quite an efficient system you would probably expect to have a system efficiency of around 20-30% (3-4.5W)
808.4 (4W) --> 1064 (2.8W) --> 532 (1.68W) The 1064 conversion was 70% in this model.

This is pure speculation. I could look at the graph and try to approximate a true increase in absorption, but it wouldn't be necessary.