Today: Manar, Ronic and Aurlien started the study of the beam profile of the alphanov amplifier at high power.

The setup shown in the image shows how the power is reduced by using Anti-reflective mirrors,

pick up 1 : Anti-reflective coating on both sides , pick up 2 : Anti-reflective coating on one side and High reflectivity on the other

using this method, we avoid saturation and damage to the beam profiler from the high power amplified laser

in addition, an OD3 filter is placed in front of the beam profiler. and a reflective mirror is placed close to deflect the reflection from the mirror(pick up 1) second surface.

The beam observed is relatively well shaped and fitted up to 50% of amplification is where the shape starts to deform a little and is not well-fitted by a Gaussian. (image attached shows the fit for 60% amplification)

The cause is yet to be determined, as it could only be related to the reflections that occur from the mirrors (pick up 1 and pick up 2)

*** Note be always careful at High Power :)

answers from Guillaume:

- the chiller temperature has to be set to 25°C
- the power measured with the software can change a little from what Alphanov measured with their laser.
=> one can set power tolerances to 30% in the "caracterisation.csv" file.

This morning, with Manar and Aurélien, we measure the power directly at the output of the Alphanov amplifier with the 2 compressors CVBG (seeding laser at Frep = 133MHz)
(we had to remove the base plate where the deflector mirrors were mounted to put the large powermeter).

with the previously described procedure, it seems that the Alphanov software is now working and we were able to start immediately the amplifier.

the chiller temperature was set at 25°C but we saw in the Alphanov documentation that the chiller temperature was closer to 20°C.
we did several measurements at 25°C and 20°C and it doesn't change a lot the output power.
so, we set the temperature to 23°C to avoid condensation (if too cold).
we will ask Guillaume what is the best temperature for the chiller.

power measurements:
power ratio     =>   measured output power (with external powermeter, not with the software)
10%               =>               1W
20%               =>               9W
30%               =>               17.6W
40%               =>               26.4W
50%               =>               34.5W
60%               =>               42.5W
70%               =>               49.3W
80%               =>               56.6W
90%               =>               63.6W
100%             =>               ~70W (expected value, not measured)

we stopped the measurements at 90% as we observed a difference in the software between the expected values (previously recorded by Alphanov) and the present ones.
the present values are quite bigger than the ones measured by Alphanov.
Aurélien will call Guillaume to check if it is a problem or not.

Attached is the current setup in ThomX

Additional information related to the injected power into the amplifier fibers.

power = 4.45 mW (as shown from the software)

the minimum to inject into amplifier is 2 mW

Aurélien, Manar, and I spent more than 2h trying to start the software communicating with the Alphanov amplifier controller.

each time, we had a problem with the software, asking to switch OFF and ON the controller before being able to switch the amplifier diodes ON.
we switched OFF and ON many times without any success.

in the end, Aurélien called Guillaume from Alphanov... and without changing anything, it worked... strange!

one possible problem could be the correct detection of Frep of the seed laser (OneFive).
as we didn't check the signal coming from the seed laser, it could be the reason... to be confirmed.

the present status for the controller is:
- the power connector (on the rear side) is ON
- the green relay (on the rear side) is ACTIVATED
- the key (on the front side) is OFF
- the emission button (on the front side) is OFF

the normal procedure to start the controller is:
- switch the front side key ON
- start the software (possible error msg asking to switch OFF and ON the power button: don't do that)
- switch the emission button on the front side (which is red) ON
- switch the preamplifier button ON
=> all the software LEDs should be green and the PD_PULSE window should indicate 133.33MHz
otherwise, try a RESET on the software and restart the procedure (and pray).

at the end of the day, we successfully switched ON the preamp and increased the Power adjustment around 20% to get something about 10W on the big PowerMeter placed at the output of the amplifier.
=> we need another day of practice to be more confident with the software!

Updated beam image after change on the mode axis.

Taken after P4.

this morning with Manar, we did a cavity mode axis change.

we checked the cavity mode centroid position on the beam profiler (placed behind P4) and changed it by roughly 1mm on the X-axis and by 500µm on the Y-axis.

for doing that, we played only on the S2 mirror while the cavity is locked and we slightly changed X and Y motors, step by step, and we realigned the laser beam with external injection mirrors when needed.

in the end, we were able to improve the transmission power by roughly 25% (810mV to 1020mV on the Transmission photodiode on the scope).

we did several Finesse measurements after that: 4950, 5190, 5100, 5120, 5200 => ~ 5100

compare to the previous Finesse value, around 4200, this is also roughly 25% better, in agreement with the transmission power increase.

conclusion: it seems the Finesse, thus the losses are almost the same "everywhere" on the mirrors => we need to clean them or replace them.

one thing which could reduce the Finesse is the L-shape metal piece if it is slightly inserted in the FP-cavity mode path.
understanding where this L-shape effectively is is not easy... some pictures are attached.

After changing the way of injecting modulation for PDH and modulation for FSR scanning (we split the modulations to 2 different EOMs), the locking is very stable and we can measure the Finesse.

FSR = 33.34MHz

hereafter, the transmission power during an FSR scan and its fit (sweep of 300kHz of FSR in 2s)

we took 5 acquisitions which give a Finesse of :

4236
4254
4400
4045
4177

=> roughly Finesse = 4200 ! far from the 17000 previously obtained........

to measure the Finesse, instead of having 2x EOM for the PDH and for the modulation sweep, I simply used 2x generators coupled with DC-blocks to a T connector (SMA) screwed directly on the EOM input.

as the transmission signal is fluctuating, it is not easy to have a good fit of the Airy peak.

if I measure the width at half of the maximum of the peak, I found roughly 10kHz instead of the awaited 2kHz... :-(

one needs a better evaluation with a more stable transmission signal and also to be sure that the L-shape metal piece (used to remove high order modes) does not introduce some losses and then reduce the Finesse...

The fluctuation problem has been solved.
It was simply the new scheme to inject 2 RF frequencies in a single EOM.
It maybe produces some standing waves in the EOM RF input and creates some phase noise.

we went back to the standard solution with 2x EOM and the problem vanished.
Now we have a very good lock and we can measure the Finesse.

This morning, me and Ronic managed to obtain the FSR and image of the finesse which is to be analyzed.

Adding:

• Beam Image at P4 propagation, no lens is added.
• Image of the Finesse on Oscilloscope

more information to be included later

Will continue in the afternoon .

As it can be seen on the first plot, even with a good locking (good reduction of "high" frequencies noise: we had better locking than on the picture) we still have very low frequency (~ 1Hz) fluctuations

these fluctuations prevent having a good measurement of the Finesse and they need to be understood.

they can come from fluctuations due to :

- input power
- input or feedback polarization
- phase noise
- alignment
- mode matching

1) input power:
we looked at the direct reflected power from the cavity without locking as an image of the input power.
=> we don't see these fluctuations

2) input polarization:
as there are many unconstrained fibers after the NKT (EOM/AOM) it could produce some polarization fluctuations.
we put a PBS and half and quarter waveplates in front of the reflected photodiode when the cavity is not locked to detect a change in the input polarization
=> we don't see these fluctuations

3) feedback polarization:
the beam on the PDH box is coming from a wedge which can change the relative gains between different polarizations.
we put half and quarter waveplates in the injection path to adapt the input polarization with the cavity mode polarization axis
and we put half and quarter waveplates and a PBS in front of the PDH box to select the right polarization for the feedback.
=> it didn't change the power fluctuations effect.

4) phase noise
we adjusted the feedback parameters (PID gains, AOM gain, locking offset, digital and analog low pass filters) to have a clean signal without high frequencies noise.
the transmission and coupling signals exhibit quite narrow lines at the millisecond level but we see 10-15% transmission change at the second level.
as the PID has a higher gain at low frequencies, one should not see more fluctuations at these frequencies.
or if it comes from external noise, one should see a correction signal on the PZT which is the image of these fluctuations => we don't see that.
we also stopped the cavity motors controllers without any effect on the transmission stability.

5) alignment
the alignment cannot change except if some vibrations are present, which should be seen also on the PZT correction signal => we don't see that.

6) mode matching
as the coupling is only 20% and the alignment has been already optimized, the mode matching is quite bad for sure.
could it be the source of the problem?
from experience, we know that a bad mode matching implies a bad locking but the reason is not clear.
=> to be discussed with Viktor: can we improve the mode matching with a simple lens?

.

Ending alignment series !!

Adding Oscilloscope images

• delocked + locked intervals showing the 20% coupling
• lock when the housing panels are closed + airflow on .
• lock when the housing panels are closed + air flow at lowest setting

an additional factor to the PZT voltage sensitivity is the housing panels, we see a decrease in the voltage when closing them.

This morning with Manar, we installed the AOM+RF amplifier and the associated fast feedback loop.
Now the locking with the Koheras is good with a coupling of 20%.
Tomorrow is dedicated to the measurement of the Finesse.... we will have to add the 2nd EOM.

We observed a very stable lock if the airflow is OFF.
when it is ON, the lock is much less stable... maybe a problem of optimization of the feedback... we will see that tomorrow if we have time enough.

We also observed a quite important sensititvity of the PZT voltage when slightly pushing on the housing with the finger: we clearly see the compensation on the PZT voltage.
I didn't calibrate this voltage but it seems to be an important fraction of 1µm... I would say around 100nm
 

This afternoon, I did some alignment of the injection mirrors with a fast scan on the LaseLock to get a regular transmission photodiode signal.
the coupling increased to 10-15%

I got a first lock of the cavity only with the PZT.
there is some ringing on the error signal and the locking is quite noisy, which means the cavity LW seems to be more narrow than the Koheras LW, which is a good sign.
tomorrow, I will add an AOM to improve the locking.

I did some alignment after locking.
it was difficult because of the outside noise (engines producing loud and low-frequency noises).
the coupling is now about 20% (position saved in the injection motors software).
I added a half waveplate which has to be optimized.

Installed and inputs:

• CW laser ; power = 101 mW
• PDH ;
  • at output
    • Low pass filter 50 ohm  DC-1.9 MHz
  • at input
    • freq = 8.4 MHz , Ampl = 1.2 V  , phase = 160 /
    • voltage =   ~ 6 V
• EOM ; freq = 8.4 MHz , Ampl = 100 mV , phase = 90
• Photodiodes
  • reflection
  • transmission
• beam profiler

Observed during this morning:

• coupling :   below 10% .... approximately ~ 8%
• Oscilloscope : 
  • yellow : transmission
  • purple : reflection
  • Blue : error signal
  • green : scan signal

Note : We observe a lot of higher order modes, and they are not occurring regularly.   

A schematic of the current setup is attached.

An image from the oscilloscope show a low coupling but clean error signal.