this afternoon, we saw some electric cables badly connected to their power supply.
we fixed it by soldering them together and screwing the result to the power supply.
(see 1st image)

we lift the plate of the 1st stage and we check for optical leakage in the fibers (see 2nd image + picture of the top part of the cassette).
(Aurélien took several images)

without the 1st stage amplification, we saw some lealage only in the bottom part of the "optical cassette".
light was scattered mostly from one side (2 spots) and we saw also very weak scattering in the other directions.

with the 1st stage amplification, we clearly saw the losses from the bent fibers inside the top part of the cassette => it's a good sign.
but after the 5%-10% coupler (the one used for the diagnostic of the power to allow the use of the 2nd stage), we don't see any losses, which means there is no light in this part !
the fiber break could be in between...
Aurélien should send the images to Jérome to get a diagnostic.
the old schematic is attached but it has been modified in the Loic Thesis (p. 165)

we identified the black optics components as 2 isolators (AFW-PISO-30-1W-FB) and 1 circulator (AFW-CIR-PM-30) from AFW technologies.

this morning with Alice, we sent the Menhir 160MHz injected in a fiber (with 6mW at the end of a long fiber) into the laser amplifier, to look for leakage or damage in the first stages of the amplifier (the amplifier is totally off).

first of all, we checked for light scattering around the laser crate with a sensitive optical card => nothing

and then, we checked for light scattering inside the laser crate with an optical viewer => we just saw 1 or 2 small spots located at the end of an optical element at the 2nd stage level.
but it's difficult to understand the optical path and know the different elements with the 1st stage still in place.

we think it is mandatory to open the top of the crate and lift the 1st stage to have a better look inside the optical parts which are at the 2nd and 3rd stage levels:
we could remove the front side of the crate without any damage to any fibers in the crate.

we just saw 1 fiber, glued to an optical element on the 2nd stage, and going to the 3rd stage.
the 1st stage is just an electronics parts stage which seems easy to be removed.

... to be discussed...

summary:

- the motors used to move the D-shape are the Newport Picomotors 8303-V
the sensitivity is roughly 30nm/step
the range is 1 600 000 steps or 50mm

- the 4 axis controller used ot move these motors is the Newport 8742.
channel 1 is for the vertical D-shape
channel 2 is for the horizontal D-shape
+N steps on the controller, you retract the D-shape mirror from the beam
-N steps on the controller, you push the D-shape mirror to the beam

the 0 position, vertically and horizontally is close to the beam.
the stand position is at ~ +200 000 steps in both directions.

Here's a summary of our experiment last week:

The initial telescope position: 920 mm (f=+250mm) and 1148 mm (f=-150mm) from the amplifier output.

Mon May 6: We moved the concave lens 0.5mm closer to the cavity.

Tue May 7: We moved the D-shaped mirror position at high power, and the intracavity power reached a maximum of 566 kW at 7 A (as Fig 1). The telescopes are the same as on May 6.

Mon May 13: We moved the two lenses closer to the cavity by 12 cm with the two lenses 20 cm apart. At 5A and 6A, we tried several times to move the concave lens slightly to get higher power. CEP and alignment were optimized after each movement. The best power is shown in Fig. 2 and the table.

Tue May 14: We moved the two lenses far from the cavity ((in the middle of May 13 and before). We tried several times to move the concave lens slightly to get higher power. CEP and alignment were optimized after each movement. The best power is shown in Fig. 2 and the table.

We find a small peak in the transmission at high power when the cavity is just locked (as shown in Figure 4-6 at different powers).

here is a Matlab code to try to optimize the telescope for a hot cavity,
taking into account the thermal lens in the coupling mirror.

from that code, one can deduce using the "Gaussian Beam" software (using the attached xml file) an optimized telescope with 100% geometrical coupling @ Pcav = 700kW and absorption in the coatings = 0.6ppm

Today, Ronic, Daniele and I redo the high-power 2-mirror cavity experiments, and the results are shown in the table (Figure 1 and Excel 2 ).

- The intracavity power ~500kW can be obtained at 47W injection, but we then have no increase or even a decrease in intracavity power when increasing the injection power, and the coupling is decreasing. It looks like the saturation power of the current device.

- We moved the telescope last week at 2A by moving the concave lens 0.5cm closer to the cavity but almost no change in intracavity power (195kW to 193kW). The telescopes for today's experiment are in the new locations from last week, and we didn't move them today.

- Figure 3 shows the locking curve at 500kW with some thermal effect changes.

- Figure 4 shows the de-lock and to-lock curves at 14kW.

- The current results may be due to two causes, the thermal lensing effect and the physical change in the mirror coating. It is possible that the transmission of the two mirrors changes with temperature.

- The next plan is to adjust the telescope at 4A to see if we can increase the intracavity power. Meanwhile, do some simulations about dynamic locking, coupling rate, and transmittance.

Today, Ronic and I recorded some intracavity power and cavity mode size as shown in Fig. 1.

Coupling was calculated using the locking curve of this overcoupled cavity. Pr/Pi = 1-Cgeo*Cimp, Cimp = 1-|1-2T1/RTL|^2

We can see that the effective gain, coupling, and mode size decrease with increasing power. And the beam is constantly moving.

Tomorrow we will try to optimize the telescope for the high-power hot cavity.

Today, Daniele and I injected the laser into the fiber, installed the telescope, connected the second stage of the amplifier, and obtained resonances.
- The output power of the menhir laser @ 216MHz is 150mW, after CVBG is 28mW , 9.6mW injected into the fiber, and 1.6mW via AOM and EOM. This is not far from the minimum 1mW seed power required by the amplifier.
- The spectrum after CVBG is shown in Figure 1.
- The waist of this 2-mirror cavity is 0.583 mm, and the position is on the M1. A set of telescopes is designed and installed as in Figure 2.
- We injected the second stage of the amplifier into the cavity and obtained fundamental mode. Aurélien and I are trying to lock it.

- Today Daniele and I cleaned the spherical mirror by wiping it with alcohol, and the finesse increased to 47k in air.

- After vacuuming, the final finesse is about 45k. The enhancement factor is expected to be 23k.

- Then we tuned the cavity length, FSR = 216.666 MHz. Aurélien helped us to install the menhir laser of 216 MHz.

- Tomorrow we will optimize the optical path and inject the laser into the fiber.

Today Viktor and I completed the installation of the two-mirror cavity and managed to lock and measure the finesse.

- The finesse is 36k now (see figure 1). For the designed value of the mirror, the expected finesse is ~50k.

- The diameter of M2 transmission is 1.67 mm,1.65 mm (see figure 2).

- The installation process took a lot of time in orienting the PBS. In addition, we found that the cavity reflected beam and the window reflected beam would interfere (see figure 3). The small spot in the lower right corner is the window reflected light.

- We need to discuss whether the next step is to clean the mirrors or vacuum and move on. 

Today, Viktor and I started installing the two-mirror cavity.
- Firstly, we cleaned the environment and the dust counter showed good cleanliness
- After opening the cavity we tried to determine the source of the strange spot with a laser detection card and found that the beam was very close to the front edge of the longitudinal D-shaped mirror. In addition there was nothing else strange.
- The setup of the two-mirror cavity is shown in Figure 1. We have to use the menhir laser of 216MHz. The mirrors used are shown in Figure 2.
- We have installed the M2 and will continue the installation tomorrow.

These days, Ronic, Aurélien and I use OEwaves CW laser to measure the finesse of SBOX. We made 5 measurements at 100kHz / 4s sweeps.

The finesse is around 35k (see Figure 1), corresponding to an enhancement factor of 14k.

In our experiments, we only saw up to 9k gain with 70% coupling, corresponding to an enhancement factor of 12.8k.

It could be because of the additional losses introduced by the high power, or the mirror became cleaner after the experiment......

Additionally, we found that the output of the OEwaves CW laser was not a perfect circle, with a depression at the edge of the circle.

Additional information:

The pulse duration has been performed in RF on a UPD-70-IR2-P photodiode from Alphalas GmbH by carefully deconvoluting the response function of the photodiode measured directly with the sub-picosecond laser beam.

Figure 1 shows the pulse width through the CVBG. Figure 2 is the pulse width when amplified to 10W.

- Last week, we obtained three curves of the variation of different cavity modes with power (Fig. 1). By comparing the gain for similar cavity mode sizes, we found that the gain always drops with increasing power.

- We measured the pulse width. The pulse width of the seed laser, after CVBG, amplified at 2A was measured by UPD (rise time < 70ps). Code filtering was performed by comparing the data to reduce the effect of rise time. The final result was t= 186 ps after CVBG and t=162 ps for the amplified at 2A.

- Today we will measure finesse using CW laser.

- We re-measured the gain before moving the mirror. Gain ~9000 was achieved at 3A, but as the power increased, the gain dropped and was difficult to optimize. In fact, we found that each day the gain was a little higher than the previous day.

- We then moved the M3 spherical mirror 1.7mm to make the beam size larger and measured the variation in cavity mode size at different powers. (Figure 1, red is the original result and blue is the result for a larger cavity mode). It is clear that the larger the cavity mode, the larger the slope. The new slope of w_y is 7.9mm/MW. Tomorrow we will make the cavity mode smaller (like in Carstens' paper) and compare the three curves.

- It is not simple to compare the gain variations of different cavity modes because it takes more time to optimize the telescope and alignment. Ronic suggested that we could compensate for the cavity mode variation by moving the spherical mirror to see how the gain changes at different powers while keeping the cavity mode unchanged.

- In addition, we measured the spectrum of the menhir laser, after cvgb, amplifier output at 3A (Figure 2). We found that the peak changed from 1031 nm to 1032 nm after CVBG, probably because of the imperfect alignment of CVBG.

- Today we moved the position of the D-shaped mirror at 6A. When motor1 (vertical) is 0.2mm away from the spot, the power in the cavity rises from 457kW to 483kW. Gain=8407 is similar to that at low power (Gain=8511). So the D-shaped mirror lost some of the gain in the previous experiments. At 4A and 5A we did not move the D-shaped mirror. (Figure 1)

- At 8A, we got 553 kW inside the cavity for one minute (Figure 2). The pump temperature is higher than yesterday (up to 34°C).

- At 7.5A and 8A, the cavity can remain stably locked, but the power fluctuation in the cavity is so large that it is difficult to optimize the alignment. This may be due to the short time the amplifier was on, the pump temperature, amplifier pointing and power fluctuations, and thermal effects in the cavity....... The amplifier operated differently at different moments.

- We measured the spectrum of the amplified laser. (Figure 3) The peak is 1032.2 nm. We will optimize the alignment and increase the power to optimize this measurement.

- Next arrangement
   Thursday: larger laser beam size
   Friday: smaller laser beam size
   Monday: finesse measurement with CW laser (Firstly check the possibility of measuring with pulsed laser)

Yesterday, Ronic, Xing, Qili and I achieved a more stable 520kW power at 7.5A (71W injection) by optimizing the alignment and locking parameters. (Figure 1)

- The cavity can be stable locked when airflow is on. At 7.5A, the pump temperature is about 28℃. The chiller temperature didn't change, to the same 23 ℃ setting. We can try 8A later (75W injection) for a short time;

- Figure 2 demonstrates the cavity mode variation, wy/Pc ~ 1.7 mm/MW, half that of the OL paper (3.3 mm/MW). The thermal deformation of our device is much smaller.

- The experimental data are shown in Figure 3. Figure 4 shows the injection power vs circulating power.

- There are some tests that can be done at the moment. I'll update on the elog after discussing the necessity today. ^_^

Last week, we achieved a stable intracavity average power of 500kW, limited by amplifier power. The experimental data are shown in Figure 1.

- We measured the transmitted laser with a power meter in the windows behind M2 and M4 respectively, and the results were consistent, so the measurements were credible.

- There is only one transmitted laser spot behind both M2 and M4.

- We measured 10-minute locking data at different powers (Figure 2). 480 kW data was not optimized, and we will add 500 kW locking data later.

- We compared cavity modes at different powers (Figure 3). There are fluctuations because we only saved one data at one power. More data will be collected for averaging later.

- After finishing the high-power experiments, we will measure the finesse and the transmission of the mirrors used. As well as the pulse duration, spectrum, phase noise, and repetition rate of the laser.

We measured again the 100W CELIA laser amplifier with a pump current until 8A.

as the first current pump of the amplifier has a Peltier issue, we don't exceeded 6A on this stage and we compensated with the 3 other stages.

7A average current is obtained with 6A / 7.3A / 7.3A / 7.4A
7.5A average current is obtained with 6A / 8A / 8A / 8A
8A average current is obtained with 6A / 8.6A / 8.7A / 8.7A

we did the power measured either with the "big" powermeter which is able to handle 100W
and with a smaller powermeter after a wedge, in the reflection path, which is multiplied by 39 to match the big powermeter measurement.

a fit a 12W/A from the cut-off current of 2A is a good estimation until 5A.

Today, Ronic, Daniele, Aurélien and I measured the amplifier power and mirror transmission.

For transmission measurements, we used the same new mirrors as Sbox and ThomX, and installed an iris and a 2-inch mount to block the scattering laser.

The angle of incidence during the measurement was about 0.5°. We changed the angle and the measurements remained the same.

If the mirror being used also has a transmission of 1.75 ppm, the original 270kW is actually 463kW!!! The gain is 6549 and the finesse is 28585 (70% coupling).

We will do more tests to check it.

• Redo the experiment and check the spot behind the window at high power.
• Move the power meter to the plane mirror M2 window. It was previously behind the curved mirror M4 window.
• Compare locking curves, cavity mode sizes, and coupling efficiency at different powers.
• After finishing the high-power experiments, we will measure the finesse using CW laser and the transmission of the mirrors used.