- Joined
- Apr 26, 2010
- Messages
- 4,175
- Points
- 83
That frustrating moment when you're initially just thinking of a way to directly double 960nm to 480nm ... and you start to build an OPA in your head....
So for chagrins, I'll post about it
The biggest issue here is that in a directly doubled setup, you risk sending an enormous amount of light, back to the semiconductor in use. Resulting in a "fried" laser diode.
In 808 -> 1064 -> 532, or 808 -> 946 -> 473, there is one commonality. All optics are HT@808nm. Meaning that anything 808nm will just pass through, and not reflect back into the laser diode. All other coatings in question, reflect 1064 and 532/473 on it's way back to the diode, meaning no light can harm the diode.
This is the issue I am trying to solve here. Though I ended up making it much more complicated than I needed to, by mentally developing a 960 -> 480 specific OPA. Though it was fun to do, and I hope it sparks some nice discussion
So you're going to start with a 960nm pump source (I was looking at a 4W c-mount). You're also going to need:
polarizing beam splitter (PBS)
front surface mirror (M1)
Pockel's cell with fast rise driver (<7ns) (PC)
Quarter waveplate (WP)
Faraday rotator, continuous wave (FR)
1x 10cm RoC mirror HT@960nm & 480nm (M2)
LBO gain medium
1x 10cm RoC mirror HR@960nm & HT@480nm (M3)
The 960nm light will initially be in a horizontal, or vertical polarization (we'll start with horizontal for this demo). This light will pass through the polarizing beam splitter (PBS) (make sure to orientate it like so). Currently, the Pockel's cell (PC) is down, so it will have no effect on the light currently. This light will continue in horizontal polarization.
The light will pass through the quarter waveplate (WP), and rotate the polarization to the right 45*, then the faraday rotator (FR) will rotate the polarization another 45*. Now the polarization is vertical.
The, now vertically polarized, light will pass through the first mirror, then the LBO. the LBO will double the frequency of the 960 (rather inefficiently). All 480nm light created will pass through the next mirror. The 960 will reflect back through the gain medium (creating more 480) and pass through the next mirror.
Still vertically polarized, the light will now hit the faraday rotator again. Rotating it ro the right, 45*. However, this means it's flipping back the same direction as the first polarization. Then the waveplate flips it backwards another 45*, to vertical still.
To explain this, place your hand in front of you, flat, palm down. Imagine there is a clock through your hand, and your thumb is touching the 9, and your pinky, the 3. Rotating your hand 90* clockwise places your thumb on the 12, and your pinky on the 6. Ok?
So, initially, your hand is palm down, like the light, horizontal. It passes through the waveplate. Now your hand rotates to the right 45*. So, your thumb should be at 10:30, and your pinky is now at 4:30. The faraday rotator does another 45*, so now your hand is in the vertical polarization, thumb on 12, pinky on 6.
When passing back through, the faraday rotator can only rotate in one direction. So when you're coming back through, it's going to be clockwise again, relative to the light. So to you, since it's traveling towards you, it is flipped (counter clockwise). So if it flips 45* counter-clockwise, it reverses the last rotation. Thumb at 10:30, pinky at 4:30. However, the waveplate beyond the rotator acts normally. Rotating the polarization clockwise again. So your hand is now in the vertical polarization. Thumb at 12, pinky at 6 O'Clock.
Due to the flipped polarization from the original orientation (horizontal) the PBS will bounce the light out at a 90* angle. This light will then go to a front surface mirror (M1). Then the PBS, and 90*, back into the laser cavity. It will do the same thing again, however, this time, once it gets passed the pockel's cell initially (before the waveplate, and faraday rotator) the cell will be turned on. So the light goes through the cavity like normal, flips from vertical to horizontal, gain medium, comes back, 45* one way, then 45* back, still at horizontal. The Pockel's cell, however, flips polarization by 90*. This makes the light vertical again, so that it hits the PBS, then front surface mirror, comes back, pockel's flips it to horizontal, pockel's turns off, and the process cycles like this continuously.
So, in conclusion!
This will maintain a constant amplification throughout the cavity, keeping only one polarization inside, allowing a constant gain within the cavity, and constant gain on 480nm, and cycling of 960nm
Complicated, yes, as well as expensive.... Though it's a fun concept to think about
Later folks! Time to relax now
Pretty little diagram
So for chagrins, I'll post about it
The biggest issue here is that in a directly doubled setup, you risk sending an enormous amount of light, back to the semiconductor in use. Resulting in a "fried" laser diode.
In 808 -> 1064 -> 532, or 808 -> 946 -> 473, there is one commonality. All optics are HT@808nm. Meaning that anything 808nm will just pass through, and not reflect back into the laser diode. All other coatings in question, reflect 1064 and 532/473 on it's way back to the diode, meaning no light can harm the diode.
This is the issue I am trying to solve here. Though I ended up making it much more complicated than I needed to, by mentally developing a 960 -> 480 specific OPA. Though it was fun to do, and I hope it sparks some nice discussion
So you're going to start with a 960nm pump source (I was looking at a 4W c-mount). You're also going to need:
polarizing beam splitter (PBS)
front surface mirror (M1)
Pockel's cell with fast rise driver (<7ns) (PC)
Quarter waveplate (WP)
Faraday rotator, continuous wave (FR)
1x 10cm RoC mirror HT@960nm & 480nm (M2)
LBO gain medium
1x 10cm RoC mirror HR@960nm & HT@480nm (M3)
The 960nm light will initially be in a horizontal, or vertical polarization (we'll start with horizontal for this demo). This light will pass through the polarizing beam splitter (PBS) (make sure to orientate it like so). Currently, the Pockel's cell (PC) is down, so it will have no effect on the light currently. This light will continue in horizontal polarization.
The light will pass through the quarter waveplate (WP), and rotate the polarization to the right 45*, then the faraday rotator (FR) will rotate the polarization another 45*. Now the polarization is vertical.
The, now vertically polarized, light will pass through the first mirror, then the LBO. the LBO will double the frequency of the 960 (rather inefficiently). All 480nm light created will pass through the next mirror. The 960 will reflect back through the gain medium (creating more 480) and pass through the next mirror.
Still vertically polarized, the light will now hit the faraday rotator again. Rotating it ro the right, 45*. However, this means it's flipping back the same direction as the first polarization. Then the waveplate flips it backwards another 45*, to vertical still.
To explain this, place your hand in front of you, flat, palm down. Imagine there is a clock through your hand, and your thumb is touching the 9, and your pinky, the 3. Rotating your hand 90* clockwise places your thumb on the 12, and your pinky on the 6. Ok?
So, initially, your hand is palm down, like the light, horizontal. It passes through the waveplate. Now your hand rotates to the right 45*. So, your thumb should be at 10:30, and your pinky is now at 4:30. The faraday rotator does another 45*, so now your hand is in the vertical polarization, thumb on 12, pinky on 6.
When passing back through, the faraday rotator can only rotate in one direction. So when you're coming back through, it's going to be clockwise again, relative to the light. So to you, since it's traveling towards you, it is flipped (counter clockwise). So if it flips 45* counter-clockwise, it reverses the last rotation. Thumb at 10:30, pinky at 4:30. However, the waveplate beyond the rotator acts normally. Rotating the polarization clockwise again. So your hand is now in the vertical polarization. Thumb at 12, pinky at 6 O'Clock.
Due to the flipped polarization from the original orientation (horizontal) the PBS will bounce the light out at a 90* angle. This light will then go to a front surface mirror (M1). Then the PBS, and 90*, back into the laser cavity. It will do the same thing again, however, this time, once it gets passed the pockel's cell initially (before the waveplate, and faraday rotator) the cell will be turned on. So the light goes through the cavity like normal, flips from vertical to horizontal, gain medium, comes back, 45* one way, then 45* back, still at horizontal. The Pockel's cell, however, flips polarization by 90*. This makes the light vertical again, so that it hits the PBS, then front surface mirror, comes back, pockel's flips it to horizontal, pockel's turns off, and the process cycles like this continuously.
So, in conclusion!
This will maintain a constant amplification throughout the cavity, keeping only one polarization inside, allowing a constant gain within the cavity, and constant gain on 480nm, and cycling of 960nm
Complicated, yes, as well as expensive.... Though it's a fun concept to think about
Later folks! Time to relax now
Pretty little diagram
Last edited:
