Home Return to Main Page

RW MOPA Version 6 Application Notes. Updated 7/18/01 (Below)

rwmopa6.gif Diagram of Version 6

rwmopa6.jpg First Photo of Version 6

r6osc.jpgPhoto of Oscillator Installed

ruby.jpg Version 6 at 1.5 to 2 joules. 13" of power amps

ruby.jpg  3/8" diameter amps. Version 6.

rw6vo.jpg Final Photo of Version 6 corrected for elliptical beam and l2" of power amps (1 to 1.5 joules).

oc.jpg OC End of Laser Head.

hr.jpg HR End of Laser Head.

cavray.htm Ray trace of dual flashlamp cavity.

pw65.jpg Old Picture of ver 6 power supply. The new one (modified) is 8kjoules

Lessions learned have pointed the way for this version: more stable resonator with higher resolution optic mounts, Higher gain by using a beam expander and larger diameter ruby rod amplifiers, tilting all optical axis to reduce backreflections, dual pump the amps to get more uniform pump distributions. The 3/8" ruby rods that I will be using for amplifiers were from various auction sales and suppliers and have frosted sides to help diffuse the input pump light

A 3 rod (3/4 inch diameter Invar x 36" long) resonator mount is used with a coaxial 6 mw HENE installed. Although for the purpose of a pulse laser like ruby, stainless steel rods would work okay as Invar is very expensive. The laser's basic design is like RW-MOPA-5 but the HR/dye cell is replaced with a standard HR and a newly designed two window dye cell. These 694 nm vcoat single side AR windows were purchased from CVI laser. This feature allows me to change HR out as future testing allows. The oscillator is a 1/4"x3" rod in a single flashlamp arrangement. Amp1 is a dual pumped 3/8" by 6" ruby rod, Amp2 is a 3/8" x 4" rod and the Amp3 is 3/8" by 3" rod. The amps were arranged like this to yield max aperture when having to tilt the rods to avoid back reflections intercepting the previous stages. The shorter rods can be tilted more than the longer ones so when the beam is at it's minsize the longer rods are used first. Extra care is being taken to eliminate non uniform pumping to achieve even a higher quality TEM00 mode beam by the use of dual cavities. But what was found out was that the beam shape did become elliptical due to larger divergence in the plane of where the lamps are. The output beam became 6.8mm wide by 4.5mm high due to this difference in divergence. This also required that the tilting of the rods in the vertical direction to maximize the available aperture to beam size to reduce diffraction effects caused by limited aperture size of the tilted rods. Diffraction was calculated to be no more than 4% peak to peak per AMP2 and AMP3. Diffraction was very low for amp 1 since it had the smallest beam size and the least tilt (and reduced aperture due to tilt). The use of ruby AR prisms in RW-MOPA-5 have been replaced with ruby-dielectric coat front surface flat mirrors 1"x 0.25" set to 45 degree incidence to help eliminate extra surface reflections. The main problem with RW-MOPA 5 which used the 1/4x3" rods (1 for the oscillator and 4 for the amplifier) was the power density grew upwards of 11 to 19 joules/cm2 which contributed to AR coating burns or pitting problems of the ruby rod ends. Inaddition the maximum gain was around 180 to 360 millijoules depending on the number of pulses. But RW-MOPA-3 to RW-MOPA-5 showed the potential for these surplus items. Calculations show you could use a beam expander of around 1.5x magnification on my earlier versions and this would reduce the high power density problem. These calcs reveal that RW-MOPA-5 would have produced 500mjoules and kept the power density down to 8joules/cm2. Therefore the use of a BEAM EXPANDER IS A MUST FOR THE LASER DESIGNER. Technically you would want to keep the beam diameter at 1/2 the radius of the rod for the lowest diffraction effect. Later I would find by using larger rods and lowering the beam density that much greater gains could be achieved per rod length (due to more rod volume) and RW-MOPA-6 is designed to yield 1.0 to 1.5 joules at a power density of 7j/cm2 which will protect the rod and optics. Larger rods demand more power so the power supply increased from 5kj to 8kj).

Experiments with using two 3" rods in series in the resonator (master oscillator) proved that this idea gave more gain but even at 0.8mm aperture I couldn't kill the other spatial modes and dropped that idea for the new laser. Currently the new resonator has the following specs: 1"x0.375" 7meter concave HR, a resonant reflector. A resonator length of 77cm, G of .889, an exit beam waist of 1.39 mm, 6% mode size increase at HR with a 1.4mm aperture for tem00 mode yielding a Fresnel number 0.91 and a loss of 16.5% for TEM00 mode and 45% loss for the TEM01 mode per pass. I changed from a 3 meter HR used in RW-MOPA 5 to gain more mode volume through the ruby rod for higher output since I now use 4" gimbals mounts for alignments. RW-MOPA-5 used 1.5" mounts for fine tuning the resonator. A 3m HR gave a lower G value for a wider tolerance of mirror misalignment, but at a sacrifice to mode volume size of the active region in the rod and the corresponding lower power output of the oscillator. In addition, a higher tolerance of alignment also requires a more stable structure to hold those mirrors.

I tried using some solid saturable absorbers set at brewster angle. Test proved that I would have to limit the output of the oscillator to 8mj in order to get good q switch modes with the solid type as it was not adjustable. I decided that the DOTCI/ethanol for it drawbacks on maintenance issue was still superior for pulse control on setups of 2, 3, or 4 pulses. Also the liquid dye method allowed adjustments to give outputs of up to 32 millijoules. Overall I found the Q switch mode can be 1/2 of the free running mode of either saturable absorber. You adjust the solid one by input power or aperture setting. The liquid type is adjusted by setting desired input power and aperture and then adjusting dye concentration to get desired q switch behavior. As you increase the dye the laser goes from free running to multiple pulses, to single pulse and finally to no pulse. The new flow through cell design allows me to take apart the cell and clean or replace items when necessary over the HR cell design. Also I plan on giving the new cell a slight tilt even with AR coats to eliminate resonance or reflections in the systems.

For the best quality for your holographic laser, always use optics in the resonator portion of at least lambda 1/10 and surface quality of 10/5. As hundreds and some thousand passes may be done through these. There is always more tolerance later downstream as wavefront errors are accumulative overall in the entire holographic setup.

3/18/01 - Machined a mount for the resonant Output Coupler. Installed a 7meter concave HR from CVI laser. 1"x 0.375" optic, too bad I found out too late that it also reflected too much of the HENE laser as I got poor transmission through it (I should have specified AR for 632.8nm P/n SWP-0-R694-T632-SMCC-1037-7.00CC $596 ouch! vs $225 for the regular one P/n R1-1037-0-7.00CC) Installed XY translation stage with Aperture. Found out that the aperture mounted on the OC side gave beautifully round and clean TEM01 and TEM10 modes! Couldn't kill either even down to 1mm aperture. At 1.5mm multimode began. When I switched the stage and aperture to the HR side, I got TEM00 mode even at 1.4 mm. Installed Dye cell and Photodetector and got 54mj output with a 1.4mm aperture and free running. When DOTCI dye was added, I got a 19mj single pulse with 35nanoseconds of pulse width. Dye concentration was measured at 11.8% loss per pass. By the way the oscillator output was measured after the beam expander and this was found to only reduce output by less than 8%. The beam expander was made from Thorlabs parts: SM1 lens tubes, holders and two AR coated lens. I think the -B AR coat lens from Thorlabs may have a low energy limit. The expander increased the beam size by 2.5X (a 30mm plano concave and a 75mm plano convex). This decreases the 1.4 joule/cm2 power density down to 0.23 j/cm2 for the entrance into the amp stages with a beam size of 3.9mm.

Flashlamp pulse was 750 usecs due to 1160ufd and 58uh circuit at 1110volts. Approximate 700 joule input to oscillator. Updated the RW-MOPA 5 power supply by adding two more circuits from 5 to now 7 flashlamp circuits. (1 for osc, 2 for amp1, 2 for amp2 and 2 for amp3). Changed cable design from quick disconnect to direct lug to lug attachment to minimize possible arc overs that had occurred on some plugs. Also beefed up the 12volt supply circuit to drive the HENE in the laser head and also to supply regulated power for the photodetector and timing circuit also in the head. The laser head contains the timing circuit, trigger circuits, and trigger transformers. Oscillator uses an EG&G 35uh series trigger trans with a 23uh air coil. Amps will use Laser module's LM640-2 trigger trans at 56uh for each flashtube. Trigger circuit is SCR triggered (ECG 5450 - 25amp 800v) with 300volt supply.

3/18/01 - Tested LM640-2 trigger trans and they needed 600v to at 200nfd to fire the flashtubes. Updated Power supply and provide 600volts in addition to the 300v needed for the osc trigger trans.

6/16/01 - Performance of the dual cavity looked dismal as only 1.35x energy boost to the oscillator signal. I was concerned that the cavity might not be as efficient as an elliptical. Traced problem to the timing electronics. Problem was the oscillator was firing at the same time the amp was. Normally you want the amp to fire first and store the flashlamp pump before the oscillator fires. The problem was corrected by using a shielded coax for the osc trigger wire and by addition of optoisolators, separate ground for NE556 timers and separate 12volt supply with some wiring run changes to minimize pickup of trigger pulse from the trigger circuits.

With the ability to adjust the delay again it was found that I could obtain 4.3x amplification in a 3" amp! The delay was set to 426 usec. The flashlamp pulse width is 1msec for oscillators and amps. Oscillator pulse arrives 300 usecs later from its flashlamp pulse . Therefore the total delay for the amp is 726 microsecs. These timing circuit fixes have proved critical to the laser's performance! RW-MOPA 6 would put out 1100 mjoules properly timed but if the oscillator fired when the amps did the same laser would only put out 330 to 450 mjoules! The final overall system performance showed a signal boost of 100x or around 3.7 joule/cm3 pumping. As a final note a ray trace of the cavity showed generally good performance. The cavity dimensions are 1.45 inch wide and 5/8" wide. The flashlamps and ruby rod are 3/8" inch. The flashlamps are placed at their widest and next to the reflector. The cavity was milled flat in the center and had a 5/16" radii on each end next to the flashlamps as this provides a bit larger reflector to allow the flashlamp to achieve a greater than 180 degree radiation pattern without being absorbed by the flashlamp blackbody.

Also updated the power supply. Added another variac and microwave transformer for the 6" amp(7/16/01). 2025v and 1600 joules per lamp from 5 resistor divided 3900ufd caps (total is 780ufd) and 110uh trigger trans and inductor yield a pulse of 900 microseconds. The inductor was an additional 54uh air coil to properly adjust the circuit to 110uh. At 1189 volts for osc, amp2, and amp3 the supply provides 819, 1837 and 1837 joules. Amp 1 at 2025 volts receives 3200 joules.

7/17/01 to 8/28/01- Installed plano concave lens in a SM1 tube and mount for the external output port. Tilted rods vertically, due to elliptical TEM00 mode output of 6.8mm by 4.5mm. Each amp further causes more divergence in the horizontal plane due to thermal lensing from the absorption of the pump energy. In order to correct this problem the last two amps have to be reoriented vertically to increase the divergence in the vertical direction. Output of the laser is from 1 joules to over 2 joules depending on the number of pulses (1 to 5 pulses).