1. Every day, turn on the lasers, measure the powers on the laser table and the experiment. Make sure everything checks out from day-to-day. If everything matches to the day before (within a percent), then go to step 2, otherwise go to step 3
2. Decide if you want to run and take data or not. Turn on the oven depending on what you are going to do.
3. Turn on the oven and wait for it to warm up. Measure the blue MOT fluorescence and the red MOT.
4. Only after measuring the blue MOT fluorescence and red MOT, re-optimized/re-fiber couple back to the previous day measurements (within a few percent since it will drop a little bit over time)
5. Ensure the red MOT is back to where it was the day before

Holzworth 332,107 → 332,121 kHz

rmot OD~ 1.58

Beginning of the day:

red_down input 1.18mW red_down output 0.78

After adjustment:

Inputs(Single freq/ stir): 1.15mW/73uW

Outputs(Single freq/ stir): 0.440mW/30.5uW

Efficencies: 38%/42%

Holzworth: 332,105kHz

Zeeaman power at window: 56mW

RMOT OD 1.67

DFG OD: 0.97

Downward red:

Input to fiber: 1.2mW

Output of fiber: 0.7mW

For the 698 temp tuning: +/-2C in 0.03 step increments and waiting for it to stabilize.

Holzworth 332,110 kHz → 332,107kHz

rMOT OD ~ 1.85

DFG OD ~ 1.05

Holzworth 332,106 kHz → 332,110kHz

rMOT OD ~ 1.7

DFG OD ~ 1.04

Holzworth 332,094 kHz → 332,101kHz

Holzworth 332,097 kHz → 332,094kHz

rMOT OD ~ 1.6-1.7

DFG OD ~ 1.05

Holzworth 332,107 kHz → 332,1kHz

rMOT OD ~

DFG OD ~

Holzworth 332,111 kHz → 332,107kHz

rMOT OD ~ 1.85

DFG OD ~ 1.15

Holzworth 332,147 kHz → 332,111kHz

rMOT OD ~1.87

DFG OD ~ 1.2

Holzworth 332,173 kHz → 332,147kHz

rMOT OD ~1.88

DFG OD ~ 1.23

RMOT OD 1.7, at holzworth 291,537kHz at 4:50pm (today)

Holzworth started at 319,837kHz re-found at 309,477kHz

rMOT OD ~1.6

bmot Output PD: 0.750mV

ECD Output: 1.054V

Holzworth Starting 330,400kHz

320,114kHz

On thursday and friday of the previous week( 17th& 18th) I was task to get the effecinacy of the AOM and collimators up for the entire 461 laser.

• I started with measuring all the powers.
• zeeman, where i only breifly adjusted vertical knob of the last mirror and adjust he AOM tip/tilt
• then adjust the next aom the spilts to the 2D and 3D path.
• 3D AOM, this on
• the 1-3 collimator
• the downward coliimator
• then moved to the 2D aom, This was messed up because i coupled in the the 0th order. The other orders were being blocked and they could only be seen immediately after the aom, not after the iris
• still need to do the Double pass and the imaging

In the meeting grady said we are all ready spin polarized, and it is happening in the MOT.

Pauls meausred of the Sigma and Pi pulses and it tracks with the Clebsch–Gordan coefficients.

When Grady tried to measure them, he notice he couldnt resolve the Higher negative spin states.

He then noticed the following graph from , and how it looked like what he produced experimentally.

Turns out, since we are using a 5 beam red mot, we are spin polarizing the atoms with a Sigma+ through the upward beam, But no doing a Sigma- through the downward beam( like Paul was doing). So we are only driving to the Plus side of the spin states.

the bmot was good, after grady aligned the 1st order to the collimator, the Double pass, and the imaging.

I looked up more info on Double Passes. We a use a CAT Eye setup. Where we place a lens 1 focallength away from the aom in order to collimate the different orders of of the diffraction, and then place a flat mirror 1 focal length after in order to collimate the 1st order back.

If the aom and lens were not 1 focal length away, when we'd scan the freq, the 1 st order would move alignment

If the lens and the mirror werent 1 focal length away, the 1st order would not be collimated when it goes back through the lens and the rest of the optics.

The RMOT wasnt there. The reason was the Ion Pump of the cavity shut off at some point and the holzworth drifts by 200kHz. the pump was arcing at 7kV, so grady brought it down to 6kV. It reads 1E-9Torr.

I should be able to run at 5kV to preserve longevity, but we will need to track the drift, and re-find the clock transition

holzworth 331,704 → 331,708 kHz

new 698 laser temp 20.520C

Holzworth 331,679 → 331,699 kHz

rMOT OD 2.3 DFG OD was 1.37 (1.5 Last time grady ran)

I coupled up the Red stir fiber. It was necessary b/c the stir power were lower, but are no back to normal. I check the bmot powers, they were slightly lower, but i noticed the bmot PID was lower. i relocked the laser to get to setpoint and the powers were where they were before. So all good. but i did notice the setpoint drifted to 0.25 and hover there. The potential reasoning are humidity in the laser or elaton cavity needs to be aligned, b/c the Output PD was 755mV yesterday, but shifted to 735mV. And today I can get it to 740mV, but no more.

I drew up a laser layout of the entire 689 Laser table.

Holzworth 331,668 → 331,679 kHz

the 922 iceblock LAN light was off again.

The laser powers were the same, the Rmot OD was ~2.3. I added in a Laser table log page, and a complete 461 layout.

Towards the end of the night, the bmot Output PD was 755mV but went to 735mV,I could lock it to 740mV, but im unsure why it changed, i didnt bump the table. the pid very slowly drifted down, i the Rmot OD stayed between 2.2 and 2.3

did an extinction ratio on the Abs fiber as well

I noticed the bmot pid wasnt staying at Setpoint. i tried to increase the coupling to no effect. But i notice the power would drift at the pid fiber. I did and excition ratio and it was 20dBm. Increase it to 40dBm.But the power on the PID is still at ~0.245

Date	Oven Temp(C)	Absorption(%)	Background(mV)	Image
10/17/2024	455	~28	61
10/17/2024	425	~11	50
10/17/2024	400	~6	50

I wasnt able to make it into lab today.

need to measure the Oven absorption at 400C, 425C, and 455C. and dont forget background from Oscillascope.

Holzworth 331,648 → 331,667 kHz

got a 2.2OD rmot, i updated the lasers powers. will create a log page for laser table.

ended the day with a 2.3 OD RMOT

• unblock the abs fiber from laser table. It the other Pbs path from the 2Dmot
• get an intensity of 1.0mW/cm^2 as best you can. I had a 70uW beam at w~1mm
• set the bmot pid's I and Ival to Zero
• scan laser for 3s at 3GHz
• i set the oscilloscope time-division to 500ms, because i was scanning the laser for 3sec.
• find ratio of the dip to the top. Mine comes out to ~27.8% absorption.
• Adsense told us this drop to ~30% was expected. Back in July.

grady decreased the TEC on 698 0.03, from 19.980 to 19.950

gradys run, 497.436000latt SP =0.6 for 578ms( shown on the Qcontrol), 698 SP 0.02 for 300ms, 10 averages from -100Hz to +100 Hz, steps of 10Hz

ID(8.43s) TOF=0

increased the 461 laser power from 12.5W to 12.75W

Downward bmot: 1.2mW input and 0.55mW output, 45.8% Efficiency , this is only fiber to fiber, not at the post for normal power measurement

Holzworth 331,618 → 331,638 kHz

I tried to change the 813 lattice back to 497. I struggled with coupling the 994 light to the wavemeter. II was able to get it , but it was sudden and i dont know what i did differently but i accepted it either way. I tried to look at the lattice, but the DFG OD was too low. I believe it was the ODT1 beam, so i tried to get the OD of the DFG higher with no luck. The the noise of the wavemeter at 994 looked abnormal, so i took an image of it. Knowing that i had to get the 813 lattice back, i stopped with the 497 and went back. Since i had to re couple the wavemeter light, i know the 813 light would be slightly off, but Luckily it was still on the mot( both forward and retro beams). I aligned the retro and got an OD ~0.58. I very slightly changed the forward beam to appear in the same location as an image of a previous day, and it helped and i got to ~0.61. This is still not the 0.65 were it was before.

holzworth 331,614 → 331,618 kHz

I spent the last couple days convincing myself that I was running the experiement correctly. And I was making sure I understood the 698PID, b.c the scope wasn't showing what I expected.

I did several clock measurement at 813.428nm in order to see if we were on magic. I got some interesting results. this was 4 averages and 698 setpoint=0.44

Before I did some of these measurements, I did adjust the 698 Set point to 0.22 and did smaller steps. I noticed 2 peaks , at ~0.9kHz and ~1.4kHz. This is weird, but when I was doing a measurement at Latt_SP=0.4, the 813 mode hopped during the run, and I couldn't get it back. Could this mode hopping explain the 2 peaks and the weird kHz/SP I got? At this point it was late.

Holzworth 331,608 → 331,614 kHz

Holzworth 331,599 → 331,608 kHz

Helped Grady get more data from the 698 Laser.

Redid yesterday's measurements, but at a sampling rate of 500k, instead of 1.25M

The button, that is directly above the Scan Amp Knob, is the menu button. In the Menu → Scan → Scan Frequency, we change the the scan speed from 10 Hz values of 0.5Hz, 0.05Hz, and 0.02Hz(Smallest), and looked at the noninv(located on the Falc) through the Oscilloscope.

Then i change the bnc cable from noninv to FastIn3(DLC), set the Scan Speed to to 0.02Hz, and adjusted the Scan Amp from (0.01V to 0.1V) to see the change in separation of peaks on the oscilloscope.

Holzworth 331,570 → 331,599 kHz

Dont forget to change the tofoffset in RedMOT.pys in Fedora to 80*ms for rMOT and 813Latt, and 0*ms when doing the 1064

If you experience the Slow integrator on the FALC shifting MHz, when flipping the switch from Reset to RUN. You might need to adjust the OFFSET Potentiometer on the FALC. The Red Line on the DLC should be close to 0V, adjust this OFFSET until its as close to 0V as possible( move in small turns).

813.428030nm

to do:

• take down rf 813 cable
• cable management on the Sr_table.
• No more low hanging cables
• relabel ofl1 to 698 clock in sequence

when aligning the 813 lattice, couple notes

• retro mirror is extremely sensitive!, especially the vertical knob.
• use the rmot to how much the retro changes compared to the forward beam.
• Image delay was set to -80ms for initial alignment
• forward lens was 200mm and retro is 175mm
• the clock repumper has a new wavelength for the 813Latt due to the shift. Its 679.285000 and scan knob at 2
• when going between rmot and dgf/latt, go inside RedMOT.py and change tofoffset from 0ms to 80ms

made a 813latt code folder , and measure the decay constant of the lattice while holding it. I took 2 averages at [0,2,4,6,8,10,15,20,30]seconds

Holzworth: 331,561 → 331,566 kHz

the 461 icebloc LAN light was not On. this was due to the emergeny power shutdown this morning. and the key needed to be switched Off then back ON.

RMOT was at 0.8 OD, got it to 1.2OD.

• got bmot florence higher then normal
• zeeman slower was at 48.5mW, instead of 49.5mW
• notice the vertical 689 power was low
• I increased Redstir coupling. No effect to rMOT OD
• was about to increase 689 vertical couple but notice there wasnt a HWP between the fiber and PBS.
• initial 679 coupling is 6.9mW / 12.7mW
• placed in the 679 HWP and QWP for the excintion ratio. new coupling is 6.6mW / 12.7mW
• placed a 689 hwp for the vertical Red. set the red power from 0.175mW to 0.2mW
• rMOT OD became 1.6
• did some power rebalancing for the Red and the new rMOT OD is 1.7
• logged the new powers

813 power:

• Start 190mW
• through beam sampler:179mW
• before wavemeter fiber:5.2mW
• wavemeter:0.35mW
• after optical isolator
• after aom

Holzworth 331,523 → 331,526 kHz

Found out the 698 was drifting in power, more noticeable at low power, because of polarization drift, due to the aom being turned on all the time. This is the same thing for the 497 lattice. we are changing up the beam sampler and PD placement of the 698. the best way to laser stabilize would be to have 2 aoms. in the path to use.

updated the part of the 698 path. Before it went through 2 waveplates → through beamsampler → mirror, where the reflection of the beamsampler would go into the PD.

now the path is 2 waveplates → reflect off the beamsampler → mirror, where the through beam goes into the PD.

Grady adding in the OFL1PIDTrigger and PID for the 698. Instead of just having it be a single switch(Sw698) that we just turn On/Off

An issue with the 698 PID is the pid setpoint is holding 0 as the last value to ramped to when the pid it turned off, but the value of the of the input light is at -1. Grady added in another channel in dio2.dio15 called Trig698HoldLast. This goes from the the adwin to the dio6-pin on the red pitaya and Gnd-pin.

this Holds the last value of the pid setpoint so it doesn't go back to 0 every time and ramp too far.

Scenarios:

1. if OFL1PIDTrigger is off: input light at -1
2. if OFL1PIDTrigger is on and Trig698HoldLast is off:
3. if OFL1PIDTrigger is on and Trig698HoldLast is on: Input light is set to last “ival” value

to turn on the 698 beam, to check on it:

• sw698 :1
• OFL1PIDTrigger: 1
• setpoint 698 : 1V
• sbench: continuous
• SpectrumTrig: 1

Things to check:

• See if you can code away how the angle of the probe beam and lattice will give you different pi pulses similar to the power broadening graph
• See if you can code away how a radial offset of a perfectly copropagating probe beam and lattice will give you different pi pulses similar to the power broadening graph

• our effects might be do to the interactions between the atoms and not saturations
• Eg loss and EE losses, EG are bad, EE are not as bad
• gg turn into eg then into ee. Energy needed for each steps is different?
• try to excited only 1 of the 2 atoms to the excited state
• the rabi freq dropping off might be due to the interactions. if w in hbar w is very strong the interactions are not as important. But is w is not very strong, microscopic self trapping, and cant drive the transition. Where interaction energy is too big compared to trapping freq
• Reference Paper
• multiplying the scattering length with more atoms.
• need to multiply our Kfa vector by 10(for spins states) going from 0.1 to 1. making us in a strong atom interaction regime
• should measure the uniformity of the spin states in the tweezers
• if we are in (eg+ge) we can use the singlet (up*down -down*up),
• if we are in (eg-ge) we can use the triplet (up*up+down down), (up*down + down*up), (down*up+up*down)
• Spin polarizing will make this all easier
• could be miss judging were are atoms are. 2 atoms in the second motional band is energy identical as 1 atoms in the first band and 1 in the third band
• characterizing a DFG: use TOF, measuring the compressability,
• people probe their mott insulator by seeing the atom- atoms interactions. Maybe we can bringing 2 tweezers together and seeing how the energy transfers between them. this could tell us how atoms in each tweezer

• focus on getting magic
• use spin polarization
• decrease probe power and longer pulse time
• measure the life of time of probing the 497 repumping
• make sure we have deeply deg fermi gas in tweezers
• work on lattice to help us with the tweezer

Took some Power Broadening measurements to see how for off they are.

I set the 698 polarization to Horizontal Linear. It was close, but not there. I had to get a little creative for the current setup. The optics near the before the window are extremely sensitive to alignment( 497 retro lattice and clock). Here is a picture of my set up. I measured a vertical Polarization, because the polarimeter was shifts 90 deg.

Grady switched out the MOUNT for the lattice focusing lens for a adjustable one in the axial direction, so we can move the focus to lign with the DFG

Grady realign the lattice, to make sure that the forward beam and the retro beam were more comparable, the forward beam was off the optics and the beam was bad. We were only really trapping in with the retro. the lattice is back to to where we expect it.

we did multiple more power broadening scans. And i did a radial sideband scan. all these scans were with 1 average

Description(from image):

1. Mirror 1 of a 4-mirror cavity
2. Mirror 2 of a 4-mirror cavity
3. Mirror 3 of a 4-mirror cavity
4. Mirror 4 of a 4-mirror cavity
5. Doubler Crystal
6. mirror for 497 output
7. mirror for 497 output
8. ??? for 497 output

Alignment Process:

1. Connect the 2 BNC cable to an Oscillascope. You should see 1 peak and the derivative of that peak.
2. Adjust mirrors 1 and 2, going two rounds of vertical-vertical to horizontal-horizontal KNOBS, until Peak is MAXIMUM
3. Adjust mirrors 3 and 4, going two rounds of vertical-vertical to horizontal-horizontal KNOBS, until Peak is MAXIMUM
4. Adjust Doubler Crystal knob in one-direction, then REPEAT Steps 2 and 3, until Peak is MAXIMUM
5. Continuing adjusting Doubler Crystal until HIGHEST possible peak is reached. If peak goes down, reverse direction of adjustment

497 ECD-X Layout

There are more details but you aligned this by:

1. plugging in the 2 BNC cable from the 497laser
2. scan the laser 6GHz at 40us
3. optimize the 994.86nm laser through the 4 mirrors and the doubler crystal
4. There is a lot more, but that the VERY brief run down

461 ECD-X Diagram, from Manual

Earlier this Grady coupled in the high efficiency to all the 461 light in the experiment. He is deciding to couple high power of the from the laser, where I will completely re align the laser table and measure all Efficiencies of the AOMs and PBSs. So, we can find the highest power possible.

I use the bmot florescence to get a higher bmot intensity. I went through all 461 paths except for the Retro mirrors and got the rMOT OD to 1.7. which is great from what i did. Grady changed the LOAD time to 6sec and the new OD is 1.6. We have kept it there.

Grady got higher power out of the 497 laser. He did this be realigning the doubler cavity to the 497.43 wavelength. Before, he aligned it to 497.00m, but not we went from 400mW to 1.1W. There are more details but you aligned this by:

1. plugging in the 2 BNC cable from the 497laser
2. scan the laser 6GHz at 40us
3. optimize the 994.86nm laser through the 4 mirrors and the doubler crystal
4. There is a lot more, but that the VERY brief run down

Because the laser alignment was changed, we need to realign the whole 497 lath and lattice.

Grady immediately went to aligning the forward beam with ODT1, after locating the DFG. this was at I.D.=1sec and TOF=1ms.

then went straight to the retro mirror, with the same set up.

Be Careful now, that the 497 and so much power! ECD output is ~1.9V

1. seed diode laser at 1064nm
2. in doubler to 532nm
3. into TiSaph for tuning to 994.86nm
4. then doubler to 497.43nm

Powers after beaming bumped:

• 461ECD 707mW
• 922Solstis 1.95W
• input for downward bmot: 2.8mW
• input for 1→3 bmot: 28mW

ECD 461 is 707mW

Soltis 922 is 1.95W

Holzworth 331,478 → 331, 493 kHz

Got the lattice back, The OD is 3x higher now, from 0.2 to 0.65. i aligned it better, and this is at >600ms of holding.

Grady went more in depths on PIDS and how our interface works. I know the basic componetns such as the input signal, PD, Red pitaya, feedback, RF, and aom. But we went over different sencerios of when each component fails. I should play around with it and create a new PID to get more familar.

from changing tweezers to lattice in the code, ther is more i need to change. I need to change Initialization.py;

• Sw497
• L497PIDTrigger
• vstart
• vstop

found the Radial sidebands for the lattice to be ~200Hz

for the Axial sidebands, we do see them at 100kHz currently, but we see a increase in atoms between 40 → 80kHz. I tried decrease the 698 power and increasing the Pulse length, but currently cant resolve this.

Last couple of days i have been working on understanding and reproducing the DFG code. Simply using the code wont get me any where, and my goal is to know each part of it. I started with getting the Fermi energy $T_F$ and the Degernacy Parameter $T/T_F$. The code uses the equations $Li_3(-fugacity)= -(1/(6T/T_F)^3)$, but i wanted to use $T_F= \hbar^2/2mkb *(6\pi^2N/(Vg_0))^ {2/3}$ to find $T_F$. With my data from the lattice, Paul's code gets me $T/T_F = 0.51$, but with my code and $T = 140nk$, i get $T/T_F = 20,000$ I beleive im not converting to proper units, but im not sure where. I know im already accountting for the pixels to um.

Went over more polarizabilty with Grady. In Summary, from Grimm's OPTICAL DIPOLE TRAPS FOR NEUTRAL ATOMS, $U_{dip}$ proportional to $-Re(\alpha) * I$ and that normally is approx to $(1/2) m\omega^2 x^2$, but the can also be a plane-wave for a large beam(compared to the atoms), and a $cos^2$ for a lattice. The $ \alpha$ controls the attraction and repulsion forces to the the positive and negative values. Also, for our red detune traps, power needed and heating are inversely propertional, meaning as the looking at a $\alpha$ vs $\lambda$ graph, as your trapping $\lambda$ gets closer(475nm) to your Transition light(lets say 461nm). you will need less power to trap, but the atoms will start to see the trapping light as resonant light, and heat up. But if your trapping light is far detuned(1064nm), you will need more power(6W) , but wont heat the atoms of the trap as easily.

Holzworth 331,449 → 331, 453 kHz

took a wide scan of the lattice with the Clock laser.

• The Green dots are -100kHz → +100kHz, 5kHz Steps, 2Averages
• The Blue dots are -102.5kHz → +102.5kHz, 5kHz Steps, 2Averages
• The Red dots are -10kHz → +19.5kHz, 500Hz Steps, 3Averages

what i was showing, during the meeting, was a sideband. I believe since I see 1 prominent axial sideband, and i know the atoms at near 170-270nK. maybe they are in the lower trap energy levels. Matlab stop before i could start the -20kHz to -10kHz, 3avrg

what i got is ~36.5kHz for the axial and ~13.5kHz for the radial. i believe the radial is wrong becasue the waist would be ~302nm, and normally,for lattices the radial sideband is in the hundreds of Hz range

Holzworth 331,444 → 331, 449 kHz

the clock laser was locking properly, i had to lock at 0.000300nm above the normal wavelength because when you flip the switch for FALC110 to Run, is creates that offset, so i just planned for it. and it stayed locked.

the clock transition is close, but not dead on, probably due to drift.

changed asg0 waveform ofl1 offset from 0.7 → 0.3 for 3mW clock

Shutters est. 08/23/2024

Holzworth 331,439 → 331, 444 kHz

What if the AOM drift is also causing the the foward beam to move after long run times. but when i trap the atoms for >100ms, it makes the horizontal slit of atoms.

got the lattice, and it can hold atoms for >600ms

going to take a Temp Meausrement

the H_temp= (270 +/- 30)nK

the V_temp= (174 +/- 9)nk

i took 25 avrgs at TOF=22ms, i want to find the degeneracy of the atoms

Also, the 1064nm laser did not turn off once today

Holzworth 331,434 → 331,439 kHz

started the day with the virtual machine being extremely slow , so i restart Fedora twice. I thought it had froze on my but it was just very slow. In the start instructions, if you see an error for the virtual machine you can use the sudo setenforce 0 command. I decided to do this even tho, i didnt get an error. but it ended up helping. Its currently working, but Grady advised to restart the whole computer, if it happens again.

went to check ont he foward beam and it was all good. maybe it just needed a day to relax. it was kinda close to being aligned but it need to be better. I was able to hold the the atoms in the foward beam for ~200ms. but the atoms were in the shape of a hortizontial slit.

Grady mention that i could use each indiviual ODT beam with the forward beam to get a ball of atoms. and better align it. It helped, because the hortizial slit of atoms turned in a ball when combined with an ODT beam, and was aligned to the DFG location.

I moved to aligning the retro beam., i was able to get the lattice to last >300ms, but i have moved it too much and lost it. i tried looking earlier in the sequence, and found it, bu ti will continue to align it tomorrow

Also, the 1064 laser didnt turn off once.

Holzworth 331,429 → 331,434 kHz

the wavemeter froze, i thought it was the keyboard, but switching that out didnt change anything. and no buttons work. I had to force restart it. I saw an error on the wavemeter program. the wavemeter wasnt ON, Khang, just unplug the USB cable and it started working again. I know for next time, that its okay to plug it if it happens again.

the 689 stable wasnt locking to a low enough noise, i got it as low as possible, but the RMOT is 1.45 instead of 1.65. I will focus on lattice alignment instead of RMOT OD.

the 497 laser had low power today, i couldnt get the ECD higher then 520mV, barely reach set point but made it work. i went to align the retro mirror but noticed the forward mirror was off. So i went to align that. I couldnt get the atoms to condense where the DFG was. the cloud looked very weird (see image below). I did bump anything. and the image made me believe one of the optics got something on it, but the next image is the beam profiler 2inches before the chamber. and the beam is pretty clean, not perfectly circular, but not smudged.

Getting better at aligning the lattice, due to my system from yesterday. Either way, i will trouble shot this issue tomorrow.

Holzworth 331,424 → 331,429 kHz

i noticed in the the swODT1 turns offf before the ODT1 reach 0, ( about half way during the final decline) This correlates to +0.03s to the image delay. I.e> if you want to measure the DFG at TOF=0MS and NO time in the lattice, if you ODTLattice Crossover is 12s, then your image delay is 12.03s.

I took a reference of the DFG at 6.30s Image delay, and then increase the the time but steps of 0.003s and overlapped the lattice. I notice the the beam was hitting the Iris, but since i never touched it, i didnt think anything of it. I then notice i couldnt get the Lattice and the orginal spot of the DFG,only a intensity falling stream of atoms underneath where the DFG was. I decided to change the iris and good thing i did, it was block the lattice from moving side ways. and they incorporated the iris into my adjustment of the mirror.

The forward lattice only is currently some few pixels to the right of DFG and it can hold the atoms for 70ms, but it slightly stretchs out the atoms in that direction . I tried to straighten out the forward lattice beam but no luck. I move to the retro and its too far away to effect the forward beam. Tomorrow i plan to put the Thorlabs Card on a stand and back reflect through the whole path.

The lattice 497 pid wasnt not following the red set path,it was just stepping straight to 0.9 setpoint. I adjust the I in PID and that got to the follow the Red path better

Holzworth 331,415 → 331,424 kHz

tried to get the lattice back, had to adjust some of the sequence. Here is everything i adjusted :

self.ImagingDelay: 7.925*s → 12*s → 6s

self.ODTLatticeCrossover: 7.90*s → 12*s → 6s

self.EvaporationTime: 2.5*s → 6.57*s → 0.57s

commented out all self.ODT related code inside the Lattice497.pys

i Increase the stirring power to reach setpoint.

the 497 laser kept losing power and i have to keep relocking it

So i dont need to relock the 497 but the setpoint consistently goes down while i run, and goes up when i check it.

My current guess is polarization drift becuase;

• the ECD output doesnt change from 0.53V
• the aom should keep the PID at setpoint, unless this has to do with the aom not being ON all the time, like a warm up period

The 1064 laser is dying more everyday. The specs the laser is 1064nm Continous wave, 10mW output, and 10kHz, its the NKT photonics Basik seed laser inside a Adjustik Frame.

USing sidebands, Grady found the tweezer wasit to by ~645nm. This is with the telescope lens being changes from a 450mm to 350mm. We believe there could be aberiations going on, and potential the Large dichrotic lens is cause some issues

the sidebands were near 26.56kHz and 4.86kHz

Got farther with the table set ups

Work on the documenting all the new optics since Paul's thesis. and plan to create a diagram for the tweezer board

Looked into other papers and our they document and display their experimental setup

Start added to Paul's figure for the experimental set up. the location is in Strontium/pictures

Holzworth: 331,364 → 331,377 kHz

Holzworth: 331,361 → 331,364 kHz

Checked the steady state because the rmot OD was ~1.25. The bmot powers werent changed, but the location of the bmot on the CMOS camera had changed. Even though i didnt touch alignment yesterday, and i had 1.6.

The bmot PID was railing to get to setup, but it was there. I tried to get the coupling better but no luck.

Because the location of steady state was at a different location, i changed only the vertical mirrors to get it back to the proper location on the Andor software and at similar intensity. The rmot had jumped to 1.55 OD and with the holzworth the OD is 1.65. I can work with this OD, but I need to keep an one on alignment of the bmot, to make its on a dramatic alignment shift

I went to get the lattice and noticed it has moved(bumped)> I restarted alignment of the forward beam. The atoms were being pulled to the right of the screen but i could get them to be pulled to the left. I ended up spending too much time a specific imagine delay time, and had to leave. I need to next early times and the the mirror, then go to later times.

Had the meeting with the NKT Photonics sales rep today to discuss the 1064 laser and possible fixes and replacements

Holzworth: 331,357 → 331,361 kHz

Holzworth: 331,353 → 331,357 kHz

Holzworth: 331,350 → 331,353 kHz

Watched Grady try to show Pauli suppression of our fermions

Bmot SP: -0.080 for freq range of 739 - 740 MHz

Holzworth: 331,333 → 331,348 kHz

The spring for the the Flipper0 broke. We first noticed that the mount in the middle position( in between up and down). And when we trigger the mount to move, it wouldnt. We checked that the flipper mound was get 9V to it. Grady found out that the mount needs to go all the way up to reset and without the spring it couldnt. HE took off the spring, cut 1 loop off, and bent a new loop to be the end. So it is the same, minus 1 loop. and it work. We need to keep an eye on it, just in case it breaks again

Holzworth 331,328 → 331,333 kHz

the OD was 1.6 after updated holzworth and changing bmot sp (which is lower then expected). I choose to go the the lattice and it was slightly offset from the the dfg, but it could be held for 50ms. so becuase it was so close i decided to get it overlapping the dfg.

Here is the how change the forward kbos change the lattice focus position from the image.

Tightening the Horizontal knob, moves the lattice to the left

Loosening the Horizontal knob, moves the lattice to the right

Tightening the Vertical knob, moves the lattice down

Loosening the Vertical knob, moves the lattice up

Got the lattice to OD 0.8 at 12.02s image delay and 5ms overlapping with DFG

and close at 5s image delay

the lattice can hold the atoms for 50ms

Added a image to to Experimental log, showing what the bmot floresnce looks like at 1.7 OD

Watched grady look at the sidebands. THe calculated waist of the tweezers is 760nm, which is 3x what we measure with the imaging set up on the tweezer board.

He notice the blue and red axial sidebands are not the same size. If we only had all 10 spins states occupying the motional ground state, then we would see the the red sideband at all.

Grady tried different ways to load into the tweezers, and decided that we should get the lowest signal possible in the tweezers, to actually measure sidebands and maybe get a different waist measurement.

When trying to to got sideband cooling and imaging in the tweezer, we want to cool in the same direction as we image. Meaning if you image axially then we cool axially. If we image radially, then we cool radially.

Holzworth 331,320 → 331,324 kHz

the Rmot was low (1.3) tried a alot of different ways to get up higher. Grady noticed the the OD was significantly better with the doors open. Maybe its a temperature issue. But i remembers, when i coupled up a high Eff through the Bmot fibers, I improper did a extinction ratios. I did them after a pbs, which would cause the polarization to look better then is actual is. This time i did it correctly; I took out the fiber for the downward beam and i put a 1/2“ mirror in front of the upward beam collimator.

While this made a big improvement. It still drops in OD when the doors are close, but now its a lot slower. Current theory is that the temp of the table and the temp of the room are too different for the fibers.

Got the OD to 1.8

Holzworth 331,315 → 331,319 kHz

got the lattice, then grady switched too tweezers

Looked for the sidebands in the a 3×3 tweezers. changing the clock in 0.01MHz steps

the axial side bands is ~90kHz, the radial sidebands are ~15kHz,

current waist is est. 500nm

Holzworth 331,000 → 331,315 kHz

Holzworth : 331,297 → 331,299 kHz

background voltage 920mV

min: 0.764mV

max: 1.03V

Watched Grady align the lattice again. I took more detailed notes and these notes couple with the figures of the atoms taken on this date.

When aligning the lattice, you have a good chance of not seeing the same images every time you align. So a safe bet when aligning is to find the complete physical range of the lattice as is effects the DFG and move the knob to middle. then move to the other direction and do the same. Go back to each knob and fining tune until you know you on the atoms.

This next part actually couples of the figures on this date.

• Imaging Delay(ID) and TOF will be written as (ID,TOF) i.e.(10s,5ms)
• Start at (5.015s,0ms)
• pick a knob and find the middle then switch knobs
• Adjsuted the retro until OD was 1.20
• changed to (5.05s, 0ms)
• changed to (5.05s, 5ms)
• changed to (8.05s, 5ms)
• changed to (10.05s, 5ms)
• changed to (12.05s, 5ms)
• changed to (12.012s, 0ms)
• OD was at 0.504
• w/o lattice, Changed to (12s,0ms), Marked center of DFG OD 0.455
• W/ latt ON, change to (12.01s,0ms),
• Block Retro mirror
• for forward aligning,mirror to adjust is the one immediately after the hwp
• Align until atoms are overlapped, until OD is 0.474
• Align Retro, until OD was 0.388
• changed to (12.02s, 5ms)
• align retro a bit
• change to (12.012s,0ms)
• until OD 0.998
• change to (12.02s,0ms)
• until OD 0.471
• checked (12.05s,0ms)
• checked (12.05s,5ms)
• checked (5.05s,5ms)
• checked (5.05s,0ms)
• Checked the Rmot, same changed in the sequence, but it was fixed
• cHanged to (5.016s,0ms),align foward beam, block retro
• until OD 0.559
• checked (5.002s,0ms), align foward until OD 0.671
• Check(5.012s,0ms)
• check(5.016s,0ms)
• change color scale
• check(5.012s,0ms)
• align foward, overlap with previous until OD 1.57
• change to(12.012s,0ms)
• change to (12.014s,0ms)
• change to (12.015s,0ms)
• Without lattice, check (12.015s,0ms), OD 1.09
• With Lattice, align retro,
• marked locatation at OD 1.09
• check (12s, 0ms)
• checked (12.015s,0ms)
• align retro until OD 0.389 (denser atoms)
• still good at (12.02s,0ms) and at (12.05s,0ms)
• checked (12.05s,5ms)
• checked “ofters” (maybe this said “afters”, but idk what i wrote down)
• back to (12.05s,5ms)
• align retro until OD 0.636, Sensitive in vertical direction
• change to (5.1s,0ms), until 0.926 OD
• check (12.1s,0ms)
• check (12.1s,5ms)
• check (12.2s,5ms)
• add more add load time from DFG to lattice

Holzworth 331,275 → 331,289 kHz

• change the bmot sp to -0.102
• 7s load OD was 1.20 +/- 0.05
• 3s load OD is 0.66 +/- 0.02
• Went to bmot floresence( with Grady) and tried to get the bmot in the locations as 09/28/2023.
• We werent able to get it back to correct intensity, only the same location and size.
• went to absorption on the oven and saw that it was at down , and almost by the same amount as our OD throughout the past week. SO our oven could be a problem( might be due to being dropping went going through transport)
• Im increasing the oven until it hits 50% absorp or until 480C, which ever comes first
• Went to 470C, and didnt reach 50%. but saw a noticeable increase in bmot florescence. Re-align bmot. and went to rmot. New OD is 1.5
• after rmot alignment, OD is now 1.8
• decrease oven to 465C, and OD is ~1.55
• volt across the oven is 19.4V and 50 ohms for a Pwr of 7.5Watts
• brought back down to 460C and OD is 1.4
• realigned to 1.55 +/- 0.07 OD
• this large uncertainty is due to the drift of 461 freq from 39-40MHz to 40-41MHz
• change bmot sp -0.095, the OD is now 1.65 +/- 0.05

hwp for b_transverse 2 : 259

hwp for b_transverse 1 : 140

hwp for b_downward : 188

hwp for b_upward : 28

transverse 1: it not retro reflected well

Went through steps with grady to increase the rmot od.

• went to steady state and change the powers of the bmot
• added the 481 and was able to increase the OD with 481 alignment
• there was a cable in front of the red trans1 beam path, increase OD further to 1.05
• tried chaning the bmot setpoint which changes the freq, increased it even more
• then went to relocking the 689 to different moise level it chnages the OD but not better
• adjust the 481 freq , but no change
• looked at the stirring timing in the sequence,
• Now alignment
• adjusta vertical bmot adn powers
• red retros and upward
• b retros and verticals
• zeeman
• 2dmot
• repeat
• then add r foward trans
• and b forward trans

Had the Lab shutdown this morning and turned everything back on, I will details the steps in the Ectricalshutdown.

Go the the MOT back but it was off. yesterday it was at 1.27 OD, and was slowly going down through out the week. I should had got it back up as much as possible prior to the lab shutdown. TOday the OD was .8 when the table werent floating and now 1.05 when the tables are floating. I measured the Bmot power because it didnt look so great.

I got the following powers into and out of these fibers.

Input for 3 way bmot:32mW

bmot_transverse2: 3.59mW

bmot_transverse1: 6.2mW

bmot_upward: 6.6mW

Input for downward: 3.1mW

out for downward: 0.53mW

After coupling:

bmot_transverse2: 3.67mW

out for downward: 0.80mW

I preform Extintion ratios on both paths, and they are both >40dBm, but Downward blue fluctuates a lot (0.75,0.85)

THe mot went to 0.55 OD, this could be to alignment(i bumped somthing), Power unbalance, or a mixture

Holzworth 331,256 → 331,260 kHz

Holzworth 331,244 → 331,256 kHz

Tired to align the lattice more;

I add all the lab shutdown details to this page: Lab Shutdown

Holzworth 331,244 → 331,244 kHz

Holzworth 331,242 → 331,244 kHz

Spent the day trying to align the forward lattice beam because it was bumped. But the images i see arent what i expect.

Holzworth 331,237 → 331,242 kHz

found the issue with the Sw497 being ON. It was set to 1 in Initialization's.py I switched it back to 0.

in additon, the set point for 497 light was too low ~< 0.4. Grady told me that i can switch the first Waveplate, infront of the isolater to get more power, in additon to the Waveplate infront of the PBS before the tweezer fiber.

The Marking on the HWP infront of the PBS for the Tweezer's was 172.

started the aligntment of the foward lattice be and the DFG:

• Grady implemented self.ODTHoldOffset. He had it at 50ms, but i changed it to 0ms for aligning the forward beam
• When looking at how well the frwd lattice and dfg are overlapped, you only need to change the Imaged delay and the second input of self.Lattice497HoldDurations
• Tthe atoms were too close to the bottom left corner of the screen, where it was hard to see. So, I change AOILeft_hImg: 590 and AOITop_hImg: 1110 in order to center it.
• I took a series of images of the atoms transferring from the DFG to the frwd Lattice beam. After the atoms were fully in the lattice, as time progressed the atoms leaked out periodical. This made me think the focus of the lattice was not on the DFG.( the radial position of the lattice focus seems align because the location of the cloud didnt move .
• The images i taken were dated from 7394356650 to 7394356715
• Tried to set up side imaging, but only saw Steady State. I had accidently commented-out bImage(0*ms,1) in AbsImg.pys.

Holzworth 331,232 → 331,237 kHz

Had to change the Detuning of the bimage in the DDS from 172 to 162

New imaging delay is 1.2s, with OD of ~1.2

I had a slightly lower OD when i came back from lunch, and i found the the Sw497 is turned ON through the entire sequence. Its something with the Main.py

Last couple days, i have bee working on Spin-Pol. The code to generate Spectrum Files is called SpinPolarizationSpectrum.ipynb, but i have been editing a copied version called SpinPolarizationSpectrum_05-23-24.ipynb. Both are located in Code/OFL/Waveforms/SpinPolarizations

The code generated, goes over each of the 8 freq kicks in a single trigger.

I trace all the Switchs and cables need to run spin-pol.

I looked at all the change made to Seq when we are doing Spin-Pol. Since we want to use the upwards Stir Beam, instead of the downwards one. I will make sense to leave the MOTZShutter open for longer, becuase currently its closes too early

If we want to ADD the Downward Stirring beam, we would need to move the fiber for RedGuideBeam to the tweezerboard RMOT, and add in the OFLShutter to the shutters that open.

Holzworth 331,135 → 331,150 kHz

Holzworth 331,130 → 331,135 kHz

check the clock transition again

497.431700nm 679.292500nm Lattice SP 0.5 clock SP 0.02

Start offset 0.003 → 0.004 583.075689265 → 583.076748926

50Hz steps, 5 avrgs

I checked the ODT at I.D.-14s, TOF=5ms, with the vertical florescent imaging , I wasnt able to optimize it, due to lack of time

I tried testing the Spin-pol file with the spectrum analyzer. I place it in channel 2 of the spectrum card. Its hard to make out was happened in Continue run mode, but when i do Single Run and at 1Seconds Sweep time. I got the following.

I believe i see 2 too many peaks, but its centered about the correct freq.

Holzworth689 331,125 → 331,130kHz

After aligning the Lattice(steps below); I change for the clock transition. I will not change the drifting, and keep the absolute offset.

I change the clock SP to 0.5, and lattice SP to 0.7. I found the transition near +2.8kHz from 0.

I attempted to measure the detuning of the clock transition at 497.431700nm lattice.

For each SP of lattice, i went from offset 0.0021 → 0.0033MHz

starting at 583.0713166MHz

i used values as: 679.292565(25) 50Hz steps 5 averages 497.431700(15)

the ODT and Lattice didnt change positions since yesterday, which is good, but that means the retro beam for the lattice was very slightly misaligned. I attempted to correct it very carefully, but lost the atoms for a bit. i turn on the lattice PID , and I went back and forth btwn to the retro mirror and the beam up stream. Here is some reference for the future

There are visible beams coming from the Cylindrical lens. Put the lattice at >0.4SP, to see them. They are vertical the same but horizontally offset. The beam closer to the U is brighter. I original aligned the retro beam to the brighter one, then i looked for the atoms. I change the horizontal knob for the retro mirror and i saw the atoms after so small changes. I went back to see where the retro-beam was aligned too. And it was aligned to the dimmer beam that was closer to the lab computers. this is good to note, now i should be able to get the lattice quicker if i misaligned it again.

I was able to get the lattice back and have it be at a higher OD, more stable, and most important overlapped with the ODT. Its about OD= 1.00 +/- 0.05

its good that i checked, but at I.D.=14.65 sec and TOF=5ms, the lattice expands quite a bit compared to the DFG at 14s,5ms=TOF. The most likely means the lattice is miss-align enough to closer heating. i will attempt to aligned the Retro beam further to better this signal.

I tried getting the lattice to be smaller size at 5ms=TOF, but it didnt work. maybe the forward beam is miss-algined or the focus arent aligned axially, which could cause a oscillation. I got it the best i can, which is making sure the center the ODT and lattice are overlapped at everywhere in time. i.e.) from I.D.=(5s,14s) and TOF=(0ms,5ms) Make sure to chang eth Max_color scale, in the imaging config script, to ~1.0 to know where the actual center is/

Holzworth 331,105 → 331,125kHz

With the issues of the 461 on Monday, the most likely reason if it not to lock to the freq of the atoms would be the humidity in the laser. I dont fully understand it, but there is a cartridge that would need to be baked. THis time it wasnt bake, because it started working again on Tuesday and Khang was using it today.

Still, a brief instruction guide to baking the cartridge is as follows:

1. Fill the laser with Argon some how.
2. take out the cartridge
3. attach a thermocouple to it
4. set it in the oven
5. Probably power the oven with a killswitch
6. bake the cartridge at for 100-120C 3-10hours

I have not done this nor have i seen it done. and this was just told to me as a general way.

I spent the day looking optimizing the ODT and lattice within the Sequences. The power of the Lattice was not reaching 0.9, i increased it via the waveplate and slightly changing the the Beam Alignment in the SOlsTis, because sometimes it 1 step off from optimal. I was getting very lower OD's and unfilled DFGs. I redid Grady series of images from transforming the atoms from the ODT to the lattice, and i notice the location the atoms didnt move, which was good, but the OD was horrible. THen i check at different times such as 5s anf 14 Imaging delay. I remember the Filter in the imaging script was at 1. I changed it to 0 and saw a change, but i know the Filter should be at 1 for imaging. I believe my biggest issue was i was in [400,400] for size_hImg (zoomout), b/c my OD get a lot better at [100,100]. I changed the Filter back to 1, and redid the overlapped of the ODT and Lattice and different times. I noticed they were very close to being overlapped, fully overlapped vertical and no more than 5 pixels off horizontally. in addition when changing the TOF=5ms, the size of the Atom clouds were the same, and small. I adjusted the ODT beams slightly to improve the overlapped and i get the OD of the lattice at 5.65=I.D. to be ~8.8. I was also comparing the Lattice at this exact time, to the images we saw on friday (05/31), and the its its the same locations and very similar size. I saw the OD fluctuate quite a bit, and i remember this might be do to the Retro lattice beam being slightly off. I didnt change the lattice alignment via the mirrors. I know they are very sensitive, and i plan to better align the retro mirror tomorrow.

When i went from seeing the MOT to the ODT, i saw nothing. Last week we had problems witht he ODT beams not being coming out of alignment. and when i image the ODT beams individually, they were misalgined.

ODT1:OFF, ODT2:ON

ODT1:ON, ODT2:OFF

Soooooo, Immediately after write out that the ODTs were not aligned, i saw the DFG.(like the very next run). I can think for 2 reasons for this: 1. I never had the HIGH power ON and hence no ODT beams. THis is likely. but I should have caught that. 2. the 1064 is acting up, and i will need to keep an eye one it and see if this happened again.

I dont believe the beams would magically come aligned, but we will see.

Ii turns out the 461 was was coming unlock. It came unlocked 4 times now within 2 hours. and i could see the MOT degrade as the bmot scope becomes unstablized. It might have trouble locking to the Lamp or maybe the power

holzworth: 331,105 → 331,1 kHz

Grady and I Starting taking data for the clock transition. The data can be found in 497-DSSC-2024-05-30.ipynb

Looking at wavelengths:

• 497.420000
• 497.450000
• 497.480000
• 497.380000
• 497.431670 (only one that was 2 averages per point, we dont include this one, just thought it was magic)

we scanned each wavelength at setpoints (0.5,0.7,0.9): over the entire transition. at 5 averages for each point.

The ODT gave some trouble at the end because it turned off 3 times in the last scan, but the data looked good.

Our estimate Magic is 497.4317(16)nm, but we will need more data.

Just learned that 497.4317(16)nm means (497.4317 +/- 0.0016) nm

698 1/2 waveplate: 311 for Horizontal, 21 for Vertical

497 Power to Set point raito: 165.5mW/sp

Tried tracking the Drift rate of the clock Holzworth with Grady, WE determined we need more scans over a larger period of time to get a better drift rate. We either are closer to Grady's 2.61Hz/min(today's date) of or Paul's of 4.48 Hz/min (08-05-19)

Grady added in new commands to the holzworth, Ill add them to my previous holzworth commands Here

Holzworth: 331,086 → 331,090 kHz

So, we can load back into the ODT from the lattice, because the ODT was moved. We determined this because, when you try to reporduce Grady's Data from Mar 13th, you will notice that DFG stays there after transferring to the lattice, its a little offset due to alignment, but its there. Four ours currently, the DFG just drops when transferring to the lattice. Almost like the lattice its near close enough. We believe the the ODT moved because, we noticed a increase in atom count yesterday all of a sudden, and whats worry some about that is the Magnetic Field was calibrated to be near 0 at the original ODT location.

To repeat Grady's Data from Mar 13th do the following: When imaging ≥

Parameters for Potential Rabi and <100Hz FWHM clock transition:

Lattice: 497.4205nm, setpoint= 0.7 , power=122mW

Repumper: 679.2925nm + scanning tp 0.5 marker,

Image Delay: 5.65s, TOF=0ms

Clock: power= 175nW(with room light off) , Pulse Length= 500ms, PID offset=0.1(b/c no photodiode)

Top image OD~ 0.9

Bottom image OD ~ 0.19

ODT and Lattice crossover time: 5s Lattice hold time= 100ms

holzworth 698: (583039.902936 +0.085)kHz center

Watched Grady run the Experiment, here is what he did:

• Had 2 images going, one for the atoms not excited for the clock state and the other that were repumped
• the goal is to get a high signal for the repumped atoms
• the Image Delay is 5s, and TOF= 0ms, Repumper is at 679.292450
• he took 3 averages at different Clock Pulse length, around 0.2 →1.1ms in steps of 0.05ms. He is running Rabi frequencys calculations
• he is using Rabi_with_679.ipynb, data is called “2024-05-28-04-rabi”
• d
• He decrease the clock power to setpoint of 0.5
• on the PID, when the Photo diode see the light, it sometimes doesnt stabilized it fast enough and you see a Damped sin wave when its increases quickly. Grady change the “i” parameter in the PID to minimizes this sin wave to be more square.
• Grady decreased the setpoint to 0.4 and 0.3, while adjusting the “i” parameter to get a more”step function“ in the clock PID
• Started collecting data again, filename: “2024-05-28-05-rabi”
• Didnt see repumped atoms at 0.1ms CLockPulse, changed it to 0.5ms and started taking data, with steps of 0.05ms.
• range from 0.5 → 0.9ms
• CHange the ID to 14.75s, TOF=0ms, Filename:“2024-05-28-06-noreload-reload-4avgs”

Grady's goal was to reproduce the Rabi freq graph has pulse length goes to very large times, the data didnt reflect this goal. He went through Boyd, and there is a strong chance the the angle between the Lattice and 698 beam effecting our rabi measurements. In Boyd's rabi measurements, he show that the larger the angle bwtn them the worst the signal, and our Angle is ~100mrad( which is bad). We plan to put a dichoric mirror to combine the 698 and lattice. We will realign the lattice and try to keep the alignment of the 698

we place the Dichoric mirror in the place and coarsely re algined the retro lattice and 698 through the windows of the chamber/ new mirror. We used the back retro reflect mirror of the lattice to aligne the DFG. Once that was algined, we went to the 698. Because of the realignment, we might need to add the 698 photodiode up the optics path, so we dont have to worry about combining it with the lattice. This will misalgin the 698, but we only have to do it once.

• new lattice OD at I.D.=14.2s and TOF =0ms is about 0.4
• Aligned the 698, with the a normal mirror, not the pico mirror. after finding it, we set the I.D. to 5s and TOF=0ms. and dual image with the repumping
• the OD in the lattice wasnt good enough, needs to be optimize
• Set it back to I.D. =14.5s and TOF = 5ms, and made atoms in the lattice look like the want the normal DFG is supposed to look like at that timing.
• New OD at this time is ~0.45
• this whole time, the lattice was at 497.42nm
• changed back to I.D.=5s and TOF=0ms, OD=~0.35
• went back and forth btwn, I.D. = 5s and 14.75s. and change ClockPulse length to optimize, but no luck. There is chance that the movement of the lattice change the trap freq, that would detune the clock transition.
• Grady checked that we cant load back into the ODT, which leads to the lattice is moved.

Had a brief meeting with Julio and Grady. The summary, to my understanding, was that we expected more atoms back via Repumping the clock state. When comparing to our groups and Pauls, we get a fraction of the atoms repumped, the major difference is that we would 497 lattice light and they use 813. 497 is also a repumping transition to the 3D2 state. We Belief we could be transferring the atoms to this state. To test this, we plan to excited the atoms to the clock state in the 497 like normal, but then Turn On the ODT and drop the Lattice, in order to transfers them, and then repump with 481 and 679 like normal and get them back to the 1S0 state. If we get more atoms, then we can be more confident that the 497 is exciting the atoms to the 3D2 state.

Grady looked at our Atoms at, what we believe is the 0 Magnetic field. Its very close to Paul's, but slightly off.

We moved onto setting the lattice back up.

• Set up the Cylindrical lens, focusing lens, and photo diode back up.
• block the retro lens.
• Set lattice to max power.
• set DFG to 14.005s I.D. and 5ms=TOF
• used Grady's chart to change the imaging
• go to Zoomed In
• Align the lattice until you see a disformation on the DFG
• change the imaging to Zoomed Out
• Change the DFG to 13.99s I.D. and 10ms TOF
• with out the lattice turned off, place the cursor onto the center of the DFG to know its location during this next steps
• When further aligning the lattice, you will encounter 4 different scenarios.
  • 1. When the lattice is very far away from the DFG. You will see no change to the DFG
  • 2. When the lattice is closer to the DFG, the Lattice will pull all the atoms out of the DFG, but wont trap them. This will result is the atoms being gone from the image
  • 3. When you get even closer, DFG will come back but misshaped. Continue to move closer until its more symmetric in that direction of travel
  • 4. Switch to the other direction of travel and Position the imaged DFG/lattice slightly above where you placed your cursor.
• Uncover the Retromirror and back reflect it to the entire path. This should get you very closer to the atoms for a lattice
• change the I.D. to 14.55 and TOF to 2ms
• A good lattice can be held for 20-50 ms, so hold them for 20ms and align the retromirror better.

Red imaging is when you use 689 light instead of 461 for imaging. First, in Main.py, you change self.RedImaging = True, take a image.

the room lights must be off, b/c we will increase the camera gain, and it could damage the cmos camera.

In the absorption Imaging matlab file. Change the gain from 600 →3000 in steps of 500. The image is weak enough to where you need ~3000 gain.

Grady creating a away to switch the polarity of the Shim coils. We took the DIO2.DIO6 and DIO2.DIO8 slots on the AdWin for BNCs labelled “Shim-Normal” and “Shim-Reverse” for DIO6 and DIO.8

• Set the red DDS to start at 190.8445 at steps of 0.005 for 3 averages. For 135 total runs. the X-shim start at 0mA. (-20 →kHz range)

Grady showed me how to do Y-shim calibrations.

you can change the DDS steps in flist.cvs in cd: strontium/

run on DDS red beagle bone

• Set the red DDS to start at 190.8495 at steps of 0.01 for 5 averages. For 205 total runs. the Y-shim start at 0mA. (0→400kHz range)
• Set the red DDS to start at 190.8095 at steps of 0.005 for 5 averages. For 205 total runs. the Y-shim start at 0mA. (-100→ +100kHz range)
• Set the red DDS to start at 190.8445 at steps of 0.005 for 5 averages. For 205 total runs. the Y-shim start at 0mA. (-20→ +kHz range)

When you get closer to the atom's B-field=0 location, the resolution of each splitting will be harder to to see. Grady had to change parameters when changing y-shim from 0mA to 500mA. He noticed lower camera gain(~3000) and high light power was better at 500mA. the opposite was true for 0mA.

Grady got the MOT back last night,

Today, the MOT moves evbery time Nanotraps run at the same time. There could be a grounds issue of a laser(since OVEn short and almost wrecked the 497 laser).

It probably a MAg field issue, grady is checking the MAG PID scope

It was the MAG field from nanotraps

in Sr_compter_ in terminal:

LaunchPS

Set up the ODT Photo detector and pump block.

the Red Patiya for the ODT1 needed to be rese

Starting Science

Grady started where he left off( doing the Mag field calibration)

in qcontrol: BBtrigger; triggers the Beagle bone. to get a DDS to change freq for many runs(ie, changing freq every 3 runs for 200+ total runs) edit file name:

IF you shine light ALONG & in the same direction as the quantization axis: you CANT drive Pi-transition, All types of linear polarized light will drive a combination of Sigma(+) and Sigma(-), Right hand circular light will drive Sigmas(+), and Left hand Circular light will drive Sigma(-).

IF you shine light ALONG & in the opposite direction as the quantization axis: you CANT drive Pi-transition, All types of linear polarized light will drive a combination of Sigma(+) and Sigma(-), Right hand circular light will drive Sigmas(-), and Left hand Circular light will drive Sigma(+).

IF you shine light Orthogonal to the quantization axis: Vertically Linear light drives Pi- transitions, Horizontally Linear light drives a combination of Sigma(+) and Sigma(-) light, and Both Right Hand or Left Hand Circular light drives a combination of all 3 transitions. Incoming light, that is Orthogonal to the quantization axis and Circular polarized, leads to a combine of Pi, sigma(+), and Sigma(-)

My Current understanding of how we do Spin-polarization measurement:

The OSG, Downward stirring beam, and re imaging are used.

In Stellmer, he uses 2 OSG beams(with B-field) to apply a force onto the cloud of atoms, this force changes strength depending on the mf-state of the atom. Give enough TOF, he can resolve all mf-states in a single image.

:stronitum/code/old/mfscanning.py

Steps to pumping down molecules:

• Heat at hotter temps
• Heat move areas
• cycle ION pump
• decrease volume
• cycle NEG
• Heat the oven at Hotter temp(350-400C)

AOSENse cant go above 120C. and always being 30C below max

windows and valve can go to 200C but temp rate needs to be < 1C/min

1. on the zoom computer, open Remote Desktop Connection
2. connect to address: …zzz.74.52
3. open REOLINK
4. choose camera

It has been a while since i added to the notebook. This is because, we were baking and doing an oven exchange and i put everything i did there.

I left off at setting up spin polarization measurements (S-P) measurements for the experiment. In Paul's thesis, th he did S-P measurements with the OSG beams. There are 2 paths, vertical (vOSG) which combines in the vertical rMOT path right before the large shutter. the Horizontal (hOSG)….

in his thesis, Paul states that the vOSG can only do Sigma(+) transitions, but that would mean its permanently in Right Hand Circular Polarized, this is because the vert rMOT path is takes has a PBS in it.

the vOSG Needed 100mW and the hOSG needed 30mW. Paul opted to stick with the hOSG because it could do Sigma(-) transitions

When looking at the experiment, we are missing everything at the 170MHz AOM and before. I need to find a 689 light source for the OSG.

The Wollaston is a PBS that separates the to light paths be degrees between <1 and 45.

Potential light sources could be the 698 guide guide light. but i should only take this once i know the clock light is hitting the atoms. which could be next week. Also, I currently down know the power out of the fiber for this beam. I believe i need 30mW at the atoms, so the probably means 100mW to this path imaged above or i can direct add the guide beam fiber to the hOSG path on the Sr table??? maybe!

In short answer NO, we dont have 30mW to spare on the OSG, but i read the SPin-polarization wrong. Paul didnt use the OSG to spin polarize the atoms. He used the Downward Stirring beam At Sigma(-) transitions. This Right Hand Circular. We use the Downward rMOT light for the 698 guide beam, but we can still take it bake. so my original idea works, but i was going to use it for the wrong path. From Experimental Logs, Paul had 0.45mW at the atoms

I had forgotten about changing the 3 valves in front of the sieve for 1 UHV rated valve. I went through steps approved by grady to make sure it was done correctly. below is a the instructions i followed and new layout of the turbo setup. The turbo was able to reach 7W steadily.

Last night i had trouble opening the bakeout screen from my personal laptop, and it wouldnt load from another lab computer. I was able to open it from nanotraps tho. Today, I learned, i had killed the main bakeout screen and made a new one on my laptop. so when i tired to run python bakeout.py not on the nantraps computor, it would try to open on my laptop. To solve this, i killed all screens named bakeout, and created a new one on the physical nanotraps computer and it now works as normal.

Side note: make should to use screen -rd to enter a screen. and use ctrl-a d , to exit a screen

I got these errors when tryng to access the bakeot software on any computer not being the nanotrap's. More importantly i needed to run it on my laptop so i can adjust the heaters when necessary. Grady reminded me of when we connected my laptop to the network using XQuartz and adding in ssh -X or ssh _Y to the networking. I tried using those commands in difference ways and lookied up instrustions but no luck.

I decided to change course cuase it was getting late, but i tested the PWMs and they work, i see the Relays blinks 50% percent of the time, and i made sure the kill switch turns on and off. and i finished wrapping, expected for around the aosense valve.

I am not connecting the heaters because i havent figured out how to get the bakeout software from home, and this is the way to change the relay's duty cycle. I will try to figure out my networking issues.

I change the Bakeout screen UI file, It was originally bakeoutUI_sieve_bake_030824

to bakeoutUI_20240408_tweezer_oven_exchange

I Started a new Tweezer Bakeout page

http://tmpups.barreiro.ucsd.edu/bakeout.png

in terminal run: ssh nanotrap@[IP address] IP address is in vlan

password is physically on computer….

note not to forget, the panic script is running in the panic screen, to close a screen without stopping the script, press Cntrl-a d

If you are in the bakeout screen, and run python bakeout.py , it will open onthe physical nanotraps computer

to see the bakeout.py , you need to not be in any screen and then follow the steps from 08/2023

Create a screen name panic

run: conda activate bakeout

go to: /src/bakeout/src/

vim into variables.py to change your settings, than save/exit

run: python temperature_panic_check.py

you should always test that is works by setting the max turbo to a super low value and see if you get alerts. and then set it back

For the turbo power, we have set it to 10W

to let the alert python run in the back ground

this should help you go from screen to screen more easily, I also found this website with screen commands

Creating screen named “Guess” : screen -S Guess

Seeing all attached/Dettached screens: screen -ls

Attaching to a screen named Guess: screen -r Guess

Dettaching to a screen named Guess: screen -d Guess

This is for the TMP Power and TMP Tempurature Graphs in BakeOut Screen to display data.

Locate the 9-pin connector under the turbo cart. The 9-pin cable we used for this could be connected to the back of the nanotrap's turbo controller located in the image below. Make sure you can unplug it, and then do so. Add a extension 9-pin cable to it and connect it to your turbo cart 9-pin slot.

The next steps will be in the terminal and SSH into nanotraps.

Create the a screen called turbo by typing in: screen -S Turbo , Turbo is changable, its only a title

Creating this screen will take you out of any screen your currently in. You will need to do the ” Conda Activate Bakeout“ again

go to the directory: :~/src/lab_sensors/turbo/

run: python Record.py , this start the recording of data from the turbo.

Go to the bakeout screen and you should be seeing the TMP Power and TMP Temperature Datas.

Since the Turbo has not been turned out for couple months, it needs to be Soft Started first, Soft started means that the turbo will increase its power by small steps and stabilize at that power before it goes to the net increase. The manual for this turbo is this and a complete video of how to get to Soft Start Mode is here.

Just in case the video isnt available here are the exact steps on How To Do Soft Start:

1.Once you have the Roughing pump connect to the turbo, and all power connects are made, turn on the turbo cart by flipping the power switch. The Fan should turn on because the fan is set to always on.

2. Make sure the valve between the roughing and turbo is open, and Flip the roughing pump On via the switch on the turbo cart

3. Once the roughing has officially pumped all the air that it can pump out, we will now use the control panel on the cart.

4. At the same time, press and hold the (Increase/Counters) button and the (Measures/Next) button for 1-2 seconds or until the screen displays “Mode Front”. This “Mode Front” means you are in Programming Mode.

5. Once in Programming mode, The (Measures/Next) button acts like a Next Button and the (Increase/Counters) button acts like a (Select/Change) Button.

6. Hit NEXT until “Pump Settings” is display on the screen and SELECT it.

7. Hit NEXT enough times until you see SOFT START ON/OFF on the screen. Hit SELECT to turn it either ON or OFF. This will save as the default until you change it later.

8. Once you are satisfied with the setting changes, Press and Hold the (Increase/Counters) button and the (Measures/Next) button for 1-2 seconds or until the screen change back to the original screen.

9. Press the (Start/Stop) button, to start the turbo. The screen will alternate between “Starting Remote” and “Soft Start Mode”, for about 15 mins until the Soft Start has been completed.

I have 10/13 thermocouples connected to an ampilifer. I need to find 3 more amps and label them. 7 out of the 10 thermocouples have documented locatations, the other 3 slipped out of their tape and i need to retape them. I notice fore a lot of the Amps, that the thermocouples would slip out of the clamps. I trying soldering a littler piece of solder to the tip to make it larger but no luck, the wire would get hot enough to attach. Iended up using a rubber band to the label zipties to hold them togther, it not super tight, but it should stay if left undisturbed.

I was able to find 6 bolts with enough silver plating to attach the large bellows to the turbo. It clamped down in its orginal spot and the small bellows is able to reach the AOSense valve with a proper bend. Couldnt find screws for that connection point tho. There is a pot holding up part of the large bellows, so it doesnt rest on the mirror for the lattice, this post has a right angle clamp holding it in place, so its hard to move, but it will move.

The orange ducttape on the old roughing pump box wasnt sticking well, so i replaced it with packing tape that i got from Brad ( i couldnt find our). The Return Label and Return Form are taped tot he box, plus i print extra just incase it needs to go in a different box. I need to get the fedex label i believe, and talk to someone named Susan to process transfer to fedex.

This was my first week being an LTAC, and I found myself having to do a lot of presswork for the TAs and next week's lab.

I spent the week preparing and setting up the turbo/rough pump. I set the orders for more Valves, screws, and gaskets. I relabeled the thermocouples.

Khang showed be how to change the Bakeout page. The .UI files are in the Nanotrap's computers home/src/bakeout/src/ui_view

editing UI file in QT 5 Designer

after editing the ui file, I put : ./convert_all cd in ui_views

vim in bakeout.py

and change the .py file name

Holzworth 330,835 → 330,839 MHz

in the 698 path there are two waveplates, one after and before the telescope lens. You change the one before it to go btwn vert and hort. the number are 283 vertical, 240 horizontal

Grady is adding in the variable ” self.ODTLatticecrossover“ in the sequence, This is to specify when the ODT and Lattice will overlap in the sequence. This variable is in the “EvaporationTime.self” with a offset. The purpose of th is addition is to be keep the lattice and ODT overlap the same while changing different image times. Before, when we image earlier or later, the ODT Latte crossover would not be the same, but now, Grady made it where the crossover length and location moves with the ImageDelay.

Started looking at theRepumped clock atoms. Grady changed the pulse length. But he also change “BRD: False” which is in self.ODTParams. He is adding in the Z-shim to the clock sequence. variable called “self.Bz_mf_split”

Seems like it didnt effect the atom count as much so he switch to self.BitterCoilRampMax instead

Our lattice and bittercoil quanization axis is vertical while our probe beam is hortizatial. This drive the Sigma+ and Sigma- , From Grady's graph on 01/10/2024, he found the clebsch gordan coefficients for the 689 transition. They should be similar to the 698 transistion

Went through Sara cambpell thesis, and checked out 3.3.1 Magnetic field cancellation and calibration.

Grady went over subtle difference in the 689 locking. That could cause your Rmot to drop dramatically. Its in laserlock, but pay attention to the size of the noise blue line when you increase the gain, it should be as small as possible and perfectly horizontal.

Grady scanned the lattice freq btwn 994.86nm - 994.87nm, but then went to 994.84nm. We set it to 994.84nm because we expect magic to be at 497.42nm, based on Grady estimated graph.

Grady added to AbsorptionImaging.m, he is adding for another image to taken, where the first is the ground state,second is the post clock repumped , third is the reference, and fourth is the dark image.

Grady set up Dual absorption imaging. We get an OD imagw after the 698 pulse, and a second image for after their repumped with 679, in the same sequence. He also made rabi_with_679_rempump.ipynb.

Repumper set to 679.292730nm

Currently doing a freq scan of the clock transition with this imaging. (0.0005MHz steps at 10-averages)( i noticed I.D.=14.55s , but TOF=0ms, probably is the lattice), 800us 698-pulse

He Plot the 2 sets of atoms count on the same graph. I notice the non-repumped and repumped count trend opposite to each other which makes sense.

Changed the 679 to 679.292490nm

Grady checked the DFG and Rmot. They both look good. Lattice at TOF=2ms looks good.

He went through different TOF for the lattice. 2ms is the best OD

In the imaging computer, the file “Imaging_config.json” has a new input called “Clock_Mode: #” this # probably can be either 0 or 1. And it changes the single OD imaging to dual clock set up.

Grady is looking at the Dual Clock images changing either the setpoint for Lattice(in main.py) and the for the TOF.My best guess is, he is maximmizing the repumped atoms

Trying to find how other groups get high extocation ratios, Ours is around 0.2. I cam across Three-dimensional optical lattice clock with bosonic 88 Sr atom which compares their experiment to Sr87 in 1D lattice. I need to go over madjarov thesis, he talked about destructive and non-destructive imaging, this could relate to our trap depth being too high. Lastly, I believe we turn off our magnetic field, but i see other people having it on,( even Paul i believe). I know this connects to the splitting of the mf states, but i need to know why other applied a B-field.

Holzworth 330,813 MHz → 330,823 MHz

Grady and I reset the 679 to the 481 path. I got 7.6mW out the the fiber, but due to the last PBS in the RMOT path, we only got 2.2mW at he atoms. Now we get ~6.8mW at the atoms in the 481 Path.

I checked that the Both repumpers were hitting the back of the chamber( near the oven). Because we dont initially see anything when checking for the 679, I wanted to make sure we still had atoms. We did, but the lattice OD was 0.6 instead of 0.95. First checked the lattice, still hitting set point, then checked ODT, and that was worst( so NOT lattice issue). THen we checked the RMOT and it was OD 1.2 instead of 1.8. The bMOT looked good. Grady relocked the 689. But then checked the 481 optics, and saw the shutter was only half covering the light. Recentered that, adn the OD went back to 1.8 , we checked if it was moving out of center due to the running, but it wasnt. Conclusion, When i checked the overlapped of 679 and 481 through the whole chamber, i most likely didnt put back the shutter to the center position.

While trying to find the 679 transition, I did a rough scan from 679.3 - 679.228, at step sizes ranging from 0.001 to 0.01. I should have done more consistent steps, but i was also scanning at every point. I didnt see anything. I need to read the laser manual more to see how much we are scanning.

One of the quastion that we dont know how to answer is, How do we know we are drive the clock transition? Especially if we dont see any 679 repumping. We Grady decided to leave the 679 on for the entire clock-sequence and then start changing the frequency. He started at 679. 31 but around 679.2925, we noticed our Latt vanished completely. THis is super weird becuase it was low due to do “resonant” 698 light being shined on it, but we would expect it to have more atoms, Not less.

Currently Grady think we could be anti-trapped in 3P2( i need to determined what exactly anti-trapped means in this context). We will try having 481 and 679 on for the entire clock sequence Grady tried changing the clock pulse to longer times( ~250ms to ~450ms). Then changed to orginal clock repump sequence, and shorter clock pulse ( 50ms). We saw repumped atoms, finally but there is a catch. If the clock pulse is shorter, then we less clocked atoms but more repumped atoms. If the clock pulse is longer, we see more clocked atoms, but less repumpd atoms. (Repumper :679.292500nm

Started changing the 679 freq, to increase the OD. Set scanning ON and going lower to 679.2922nm but got worst. Then went up to 679.2937nm(kept it here for the day) and still nothing. The clock pulse was set to 5ms.

we observe longer 679 pulses, lead to smaller OD. We stopped seeing anything, the 698 was unlocked, but FOr how long?? We increase the clock to 36mW(max) to power broaden, in hope to really drive the clock transition.

found some iteresting papers involving Sr clock.Cooling and trapping of atomic strontium by jun ye, Clock transition for a future optical frequency standard with trapped atoms by lemonde

Grady started change tot he clock pulse length to sub-ms ( 0.02ms). 0.01ms was too fast for the AOM, Then went to 0.05ms- 1.05ms

got slightly better OD at TOF=0ms. Grady started changing the Clock pulse 100us steps to see a rough Rabi oscillation at 0ms=TOF. Grady changed the Setpoint of the clock laser, in steps of 0.05, to see OD oscillations with intensity. didnt see oscillations

He started rabi measurements, keeping the setpoint at 0.02, and clock pulses starting 1ms, 1.25ms, 1.5ms, 2ms. (5-averages)

I thought of imaging with 679 to see if the atoms are in the 3P0 state. But this cant work well enough, because when we image with 461, the atoms go to the 1P0 state and back down to the 1S0 state, and as long has we have the light shine for longer then tau=5.22ns, we will see a larger shadow on the atoms. But with my idea, the atoms would go from 3P0 to 3S1, but then decay to any of the 3Pj states and not recirculate to the 3P0 state. Making it very inefficent.

I looked over what Grady did to the 698 Holzworth beagle bone. Some commands to remember are:

hz.get_freq() :Tells you the current holzworth

hz.set_freq(#) :Resets the holzworth to a specified #

hz.track_drift('on') :Turns the drift of the holzworth ON

hz.track_drift('off') :Turns the drift of the holzworth OFF

hz.set_relfreq(#) :Changes the Holzworth #-amount, relative to the to its current freq WHile Drifting hz.reset_count() :this Zeros out your count. your new loaction will be your hz.get_count()=0

hz.get_count() :tells you how many MHz you are from your zero(starting point)

hz.offset : tells you a more exact hz.get_count(). Use eqtn hz.offset=hz.offset+# to change it mid way

hz.return_to_zero() : bring your holzworth back to starting freq(while still drifting)

We use hz.set_relfreq(#) from now on to scan.

as an example. We find the clock transition at a hz.set_freq(#). Then to do a scan, like we normally do, hz.set_relfreq(0.0004) to move it 400Hz from the transition. And doing hz.set_relfreq(0.0004) again will move it another 400Hz. And doing hz.set_relfreq(-0.0008) will move it back to the clock transition.

Some time ago, I added the 679 clock repumper to the sequence, but was able to add the 481 light so both repumpers are on. The shutter would turn on, but not move. When looking at the shutters the majority of them have the channel names formatted by “ExampleSutter”, and the moving/changing direction of the shutters is written as “Example_Slp” and “Example_Dir”. But with the 481 case, the channel name is “Repumpr481”, and to move it you use “Repumper481Shutter_Slp”. This is the opposite of all the other shutters in the sequence. But with this find, both the 481 and rMOTXY shutters open and close at the same time for the Clock Repumping sequence.

I scanned for the clock transition, to measure how much its drifting

from 1:25pm to 1:51pm. I got this

From 2:15pm to 2:47pm. I got this

• when i got into lab, Grady had the lattice running. He was changing the I.D. from 14.02s to 14.08s. then to values of 14,14.5,15.55(2msTOF),16.55(2msTOF),17.55(2msTOF),18.55(2msTOF),19.55(2msTOF), 20.55(2msTOF).
• set and 14.55s,(2ms tof). turned on the Clock light and change the pulse time.~400ms
• change the clock freq. steps of 0.001MHz, from 582.750→582.752
• change back to 14.05s(2msTOF). change the clock freq. steps of 0.001MHz, from 582.750→582.762 (3 averages)
• saved data in “497 magic” folder lattice=993000. code used in 497magic_rough.ipynb
• change back to 14.05s(2msTOF), lattice to 0.4. change the clock freq. steps of 0.0005MHz, from 582.750→582.762 (5 averages)
• change clock pulse to 450ms, ID 14.55s, clock freq of 582.755MHz
• Grady checked the pid for the Clock beam. but then the computer froze
• Computer back on, changing 698 Holz around 582.7550MHz
• grady change the latt Freq to 993.99500 and refounding the clock transition(3avrgs over 582.739→582.749MHz, 0.001MHz steps)

I wanted to learn more about finding the radial trap freq, to get a accurate waist measurement. i found this paper that provides and equation with radial and axial freqs, as they change with r,z.

I found this paper at calculating polarizablilty for off-resonant light with arbitary polarization.Here. Should help in find it for our ODT

This could help to finding AC stark shift for ODT

SBench: Vpp should be 160mV. and Trg1 should be turned on

Tried to get the Bmot spect up. This can be seen on the bmot PID. the red line needs to at 0.2568, or the sp-value in

• Grady says its impossible to see Kapitza Dirac diffraction with our current setup, and i should check it myself.
• He re-found the new clock holzworth for the DFG at 582.724MHz.
• He was originally going to do sideband spectroscopy, but change to finding the magic frequency.
• He is doing the same procedure as in his 473 magic wavelength paper. You first measurement the resonant probe frequency without the trapping light. Then you turn the trapping light on, and find the change in resonant frequency of the probe light, which should have changed due to the AC stark shift. DO this for several powers and get this slope of $\frac{Resonant Detuning}{Trapping Power} = Differential AC Stark Coefficient(DSSC)$ This slope becomes 0 at the magic wavelength. So you change the trap freq, and repeat the measurements above until you have a graph of Lattice Freq vs DSSC, and of it crossing 0 DSSC.
• This procedure was done by 2/3 groups that measured 813nm. They measured it up to 6 Sig Figs ~813.428nm
• Grady wants me to theortical find the AC stark shift of the ODT beams. I foudn a good reference in Steck's book 7.7.3 Total Shift. Which talks about calculating the AC Stark shift for the scalar, vector, and tensor parts.

I read over Boyds Resolved Sideband Spectroscopy section. It discusses how to properly estimate your trap's waist. First he uses an aligned and high intensity clock light to see the axial sidebands. This separation bwtn the clock transition and a single sideband in the Axial frequency and the trap. Then he lowers the power and misaligns the clock beam so that the much smaller radial sidebands can be observed. The distance is still the radial frequency. Note that, he says 'Resolved sideband spect' becuase you cant have your clock transition be too broaden and overlap the sidebands in either the axial or radial case. After experimentally measuring these values, you use the equation $\frac{\eta_{z}}{\eta_{r}} =\frac{\sqrt{2} \pi w_o}{\lambda_{lattice}}$ to find experimentally find you trap waist.

I read over this kind of cooling. Its a way to get our atoms to the 'motional' ground state, which refers to the ground state of the harmonic trap, and not the internal energy states of the atoms. What happens is we can be int he internal ground of the atoms (1S0), but in an excited state for the harmonic trap(n=3). Using sideband cooling, we can excited the atoms int the internal excited state(3p1), but it will be in a lower motional state(n=2). Once it spontaneously decays into the internal ground state, the motional state will stay the same(1S0, n=2). WE can continuously do this until the atoms are in both the internal ground state and motional ground state(1S0, n=0). Cooling down the atoms further. WHile looking this up, i found a paper on the side band cooling of Strontium in single atoms tweezers. This could be helpful in the future.

I tried to confirm why the Kapitza Dirac paper uses 2-photon recoil energy. Boyd only uses single photon recoil energy. this is because short pulse times, the second photon being retro reflected back need to taken into account in order to accurate identify the kapitza-dirac diffraction, but if you have the lattice light for too long the 2-photon momentum kicks turn into the stand wave lattice trap, and the 2 -photon recoil energy can just be used as the single photon recoil energy. So, at short lattice times, we should use double photon recoil energy, and longer times we should use single photon recoil energy.

WHile trying to find this Photon Recoil Momentum in Dispersive Media, i can across this paper of a proper fit to the photon momentum transfer to cold atoms. It states, the momentum equation of $p = \hbar k$ should change to $p_{n}=n \hbar k$, for n is the index of refraction of the cloud of atoms.

Clarifications from the meeting.

I had a question regarding the connection bwtn Kapitza-dirac diffraction, ramsey spect, and the tweezer waist measurement done in a Schrek member's thesis. Kapitiza-dirac using the lattice to give a momentum kick, that affect the velocity of the atoms if they are in the Raman-Nath regime(short pulse of the lattice then TOF). The tweezer waist measurement turn lattice on for varying time, then off ,and back on, this lets us see how the atoms are oscillating in the trap and depending the initial varying time, we can measure the waist from those oscillations. The Ramsey Spect use resonant light(clock), in which the energy being added to the atoms is is not necessarily transferred to the velocity, but ot the internal energy states ( 1S0 →3P0).

When looking at the number of photon recoil for the Trap depth measurement, i need to confirm with boyd if he uses n=2 for his lattice. If he does, i would be more confident in saying we should use n=2 for the lattice due to the retro beam, and n=1 for the tweezer because there is no retro beam

Here is what i got for tranmission of the 497 lattice to estimte the power of the forward beam and the retro. I got the chamber windows to be ~97.0%. From measureing the power before the first window here are all the optics and there respective Transmission rate.

1st Measure Beam→lens(99.75%)→window(97.0%)→ atoms→ window(97.0%) → mirror(99.1%)→ 2nd measure beam→ lens(99.75%)→ Retro Mirror(99.1%)- lens(99.75%)→ mirror(99.1%)→ window(97.0%)→atoms.

At a setpoint of 0.5, i got a first measured beam of ~87mW, and a 2nd measured beam of~81mW. This gives be a forward beam of 84.2mW and retro beam of 76.8mW at the atoms.

With our current setup and an average lattice waist of 27.5um, I calculate our parameters below. The main thing to note is that we are not in the Raman-Nath regime. Our parameter needs to be much less then 1

I calculated the same parameters, but only changing the Trap Depth. with the goal of getting to the raman-nath regime.

I read a bit on Raman-Nath Regime, Kapitza-Dirac, and Ramsey Spectroscopy.

Ramsey Spectroscopy: Used in Boyd's thesis as a more precise way to get the clock transition as compared to Rabi spectroscopy. Rabi spect use a single Pi pulse to get atoms to the excited state. This Pi pulse adds in inhomogeneities, which can cause broadening. Ramsey spect, using two Pi/2 pulses separated by longer separation time. This separation time lowers the inhomogeneities and lowers the broadening. This next part is an educated guess.Because we use Pi/2 pulses, those times could be small enough to cause diffraction when hitting the atoms, and have fringes due to the velocity (KICK) added to the atoms. There is a connection to bloch sphere that i currently dont understand so i need to read some more. How the line width is smaller because we measure fringes in the transition, and the center transition is thinner than with Rabi spect.

Kapitza-Dirac Diffraction: This diffraction is very similar to how aom works. As a reminder, RF is sent into a medium which causes phonons. when light passes through that medium, it diffracts into different modes. In KD diffraction, the atoms act like the medium with the atomic frequency as the RF signal, and the lattice light is the Laser. You only see diffraction at 1us scale because the trapping light is more of a momentum kick for the atoms and you see diffraction after a time of flight(15ms). This, The count of atoms at different Pulse times, fits a Bessel function, and from here, you can estimate the Trap depth of the lattice.

Raman-Nath Regime: is when your lattice pulse time is much shorter than your atomic frequency period. this is how you get KD diffraction.

Questions I need to answer and Difference bwtn Raman-Nath diffraction and KD diffraction? Understand the fringes in ramsey spectroscopy.

I recorrected my previous Trap freqs calculated on. Here are our update parameters for our lattice.

$\alpha = 1377.5 a.u.$
$average waist=27.5um$
$power=65mW$
$U_{trap} = \frac{4P\alpha}{\pi c \epsilon_{0} w^{2}}=1.34MHz=66\mu K $, in Lamb-Dicke_calc_rj.ipynb
$ \nu_{r} = \frac{1}{2\pi w_{0}^{2}}\sqrt{\frac{16\alpha P}{c \epsilon_{0} M \pi}}=915Hz$
$ \nu_{z} = \frac{1}{2\pi w_{0} \lambda}\sqrt{\frac{32 \pi \alpha P}{c \epsilon_{0} M }}=225kHz$

From Tylers notebook: Boyd's questions for how to design trap (pgs. 87-90):

1. Is λm is a practical wavelength, and is the required frequency stability reasonable?
   1. ($U_{trap}=149 E_{r}$)
2. Can we perform spectroscopy in the Lamb-Dicke limit (νT rap > νRecoil) and in the resolved sideband limit (νTrap>γClock)?
   1. ($\nu_{recoil}=9.2KHz$), ($\gamma_{clock}=1mHz$)
3. Can we efficiently load atoms into the lattice (U0 > kB T )?
   1. ($U_{trap}=66\mu k$)
4. is the absorption rate for lattice photons at λm and U0 an issue?
5. is the light shift polarization dependence at λm and U0 significant?
6. Is the fourth order Stark shift at λm and U0 negligible?

Analysis of Kapitza-Dirac diffraction

Watched Grady characterized the lattice some more:

• Checked the location Florescence of the DFG on the CMOS camera in the side Side imaging path.
• DFG 8s I.D., 5ms tof then 10ms .Started at ta low gain. 4k gain for DFG with 5ms bimage
• check OD of lattice, started OD is ~0.98. ( you can see the images on this date)
• went back to florescence with the side path

I went through 679 polariztions in Boyds thesis and with Grady.

The currently holzworth for the clock is 582,708kHz, and using the 582,612.5Khz on 02/20/2024. we get an drift rate~ 4.15kHz/day

to get a better undertanding of the degeneracy measurement, look at DeMarco's Thesis or Pauls thesis

when reading the 1S1 state or the 3P1 state, the S, P are the L for oribtial angular momentum and the subscript is the J combine with spin. With fermions, you need to know what manifolds are possible, and that its determine by the F state, which is a range from {9/2 - J, 9/2 +J}. Example, the 689 transition, from 1S0 to 3P1, we go from a 1 manifold to 3 manifold, or F=9/2 to F= 7/2, 9/2, or 11/2. Depending on the detuning of the laser. This is shown well in Stellmer's Fig 2.3

This showns how the manifolds are reached by a specified detuning of the laser. The 461 transition is different from the rest because the linewidth of too broaden compared to the detuning. They overlap too much and it becomes too challenging to resolved the manifolds of the 1P1 state. But since the others, example: like the 3P1 state are ~ GHz detuned, the kHz linewidths do not overlap and the {7/2, 11/2} can be determined.

I found the Boyds repumps the Atoms from the 3P0 → 3S1, specifically into the 9/2 manifold. This determined in Table 2.3, where the 3S1 detuning is -542MHz, which is the 9/2 manifold.

This next part can be used to determine the polarization of any of the lights use. Here I will talk more about Sigma and Pi transition, the when the mf state are degenerate or not, how it connects to quanization axis, and how to experimental achieve these.

I will use the General 679 setup, where we go from the 9/2 Manifold to 9/2 manifold.

NO H or E field:

The -9/2 → +9/2 mf states would be degenerate, meaning 3P0 mf=5/2 would require the same energy to excited to 3S1 mf= any state. The polarization wouldnt matter because there isnt a defined quantization axis, and all mf state are equally likely

H-Field(Our Least likely senerio):

For our experiment, a magnetic field is achieve through either anti-helmholtz or helmholtz configurations. An anti-helmholtz makes the the H-field in the center 0, while the Helmholtz makes it non-zero. Our Shim coils are in a Helmholtz configuration. When turning the coils on we can define the the quantization axis as point up/orthogonal to the Sr table, aligned with the H-field. Turning on the H-field splits the energy levels of the 3S1 mf states, which now gives us the ability to drive Pi or Sigmas transition. With our current setup, the 679 and orthogonal to that axis. To drive a Pi-transition, or polarization needs to be aligned with the quantization axis, so only a vertical polarized light would cause this. If we want to drive a superposition of Sigma+ and Sigma-, we need to be horizontally polarized. We cant drive isolate Sigma+ and Sigma- from one another in this configuration.

If we send the 679 vertically UPWARD (along the quantization axis). We can drive Sigma+ with a Right Hand Circular light, and Sigma- with Left Hand Circular light. We can not drive the PI-transition because the oscialltion of the polarization needs to be aligned with the quantization axis and the E-Field of light changes in the transverse direction and not along the direction of light.

E-Field(Our Likely senerio):

the Starke shift is done by an electric field, which can be cause by an Optical trap. Using the 497 lattice as an example, we set our quantization axis along the Lattice. Everything in the H-Field section is the same here but i will dicuss what we will likely do for the experiment.

Currently,we dont know if the light of the lattice in splitting the mf states. And because we are not currently needing to be in a specific mf-state, we can scan the 679 to reach all freqs, in order to reach all states. From this scanning, the polarization of 679 in anything direction does not need to be specified, and can be left unpolarizated or set by the optics of the path its adding onto.

Holzworth= 330,752 → 330,755 MHz

current lattice trap depths is 121.6Er for waist of 34um, and 318.7Er for waist of 21um, both with a power up 81mW at 0.5 setpoint

change the exposure time to 10ms because the smallest time

I tried setting up the florescence for the other transverse imaging path, I had change the exposure time (10ms)and adjust the imaging delay. but the One thing i forgot was to flip up the retro mirror for the main bImage path.

Holzworth= 330,749 → 330,752 MHz

Since Khang has been using the 1064, I need to be able to chang ethe current back to our normal settings. Our setup has the ACC at 3.82Amps on the IR box

I set up the 679 in the sequnce. I added in the REpumper to change the timing of the clock beam. and open the transvse red shutter at the right time. I might need to change the 679 to be added to the 481 path instead because if i align the 679 through the vertical fiber, it will be off center in the transerve windows. I believetha tth Rmot transverse beam arent perfectly centered on the atoms, and aligning the 679 perfectly through the transverse window will lose power through the vertical fiber

this is the steady state of the side imaging

Before i chose to use the OSG aom, I tried changing the DDSblue DDS5 DAC3 from 170MHz to 290MHz, This DDS5 should be able to go up to 500MHz, but i saw more noise the closer i got to 250Mhz, and it wouldn't reach 290MHz. This is when I decided to use the original OSG AOM and keep the RF at 170MHz. TO turn on this AOM, you Turn on SwOSG to send 170MHz input to the RF Amp and than turn on LiftTrig to the VVA.

Input Rf power: -11dBm,

Output Rf pwr(SwOSG ON): -16dBm

Output Rf pwr(SwOSG and LiftTrig ON): 29.6dBm

When setting up the 679 through the optics, The input power was 16mW, I got 14.2mW of power into the -1 mode, than 4mW of power out of the collimator. As long as i can get the beam size to be close to 1cm diameter ( like Boyd), 4mW should be sufficient.

In order to get the 1cm diameter at the atoms, I needed to add lenses in front of my collimator to get it to the correct size. It was decided to go through the 689 path with an HWP and the magnification is 3.2x, So in order to get a 1cm diameter at the atoms, i need an input beam of 3.13mm Diameter. The 2 collimators have a diameter of 1mm and ~2.3mm. Using the available 60mm and 75mm lenses, I can increase the 2.3mm collimator to ~2.88mm, and then it will be about 9.2mm at the atoms. The 679 has not been aligned through the lenses yet.

Holzworth= 330,711 → 330,749 MHz

I spent the day trying to better understand who to set of the 679 aom and be able to control it. I went through the DDSs and some of the corresponding RF amps. The original 679 RFamp and DIO slot was taken by the 11/2 aom for the 689 quad pass

For the DDSred:

1.SWAP MOT

2.Stir

3.689/OFL Shift

4.689 Cavity(holzworth)

5.11/2

6.Single Freq MOT (87 or 88)

WHile looking into these, I remembered the OSG is not being use, and instead of setting up a new 679 setup, i moved the OSG aom over to the 679 setup.

while going to the DGFG at 14s I.D. and 5ms TOF, i notice the OD was a lot lower then normal.I tried to change the timing of the I.D. and tof but iit stayed lower, even though its has a normal OD at 1ms TOF and 1s I.D. But what seems likes like a fluke, it came back to the Expect OD. I didnt change anything in the sequence, alignemnt, or powers. This happens once before. On Feb 28, prior to changing the tau to 1.4. It was low at 1.7, than i changed it to 1.4, and when i changed it back to 1.7, the OD was back to expected. I dont know why this happens. Changing the image from [400,600] to [400,400] didnt affect the OD, i just changed it back.

to update where the clock is for the DFG, its between 582.667MHz - 582.677MHz when using 4mW of power

jun ye for boyd thesis

i got 679 to the wavemeter, i only used ~49uW of power. i spent part of the day drilling larger holes on the IntraAction AOM mounts. they didnt fit the 6-32 size screws. it doesnt take long to fix, but i t took time to find the best placement.

Holzworth= 330,703 → 330,711 MHz

I spent time trying to see the DFG and mot through our other transverse imaging path, but didnt see it. I must be missing something small, and i cant think of it currently, probably due to be slightly under the weather. I switched gears to cooling the atoms down, and what parameters i could change to do so.

I started with Changing the self.Lattice497HoldTimeHigh from 1s to 0.5s

then changed the Intensity from 0.4 and 0.5

But then i notice grady had the Tau=1.7, and Only today i changed it to 1.4 because i thought i made a typo when changing the parameters. so i testing it.

this is at tau=1.4

the is is tau=1.7

Tau is 1.7, becuase it reaches 50nk

I had my exam today, Grady helped me properly set up the 679 laser. I finished the optics order. and here is my set up for those optics.

I took images of the of the beam, it doesnt look good coming out of the laser, but we only need ~3mW at the atoms. i should be able to get 3mW out of 16mW (Currently)

I comparing to Boyds 1cm diameter with 2.5mW beam

679 at 1 and 5/8 inches

679 at 5inches

679 at 17inches

679 at 33inches

I started making the list of optics we will need for the 679.

Also, when reading Madjarov paper, i learn their reasons between repumping the clock transition and not. The call repumping with 679nm 'Destructive', this help them get s 99.8% clock population count. IF they are just imaging the leftover atoms in the ground state, thy call this Non-Destructive. This comes withs 2 infidelities: Ramam scattering and Leaking into the 3P2 state.

One of my goals for this week is to image the lattice and dfg from the side to make sure its centered becasue we could be causing the heat up. and cause a high broadening.

Holzworth= 330,703 → 330,703 MHz

The Rmot was >2.1 and ODT at 1ms TOF 1s I.D. was > 1.8. Plus, now change needed to the Holzworth

I found out how much the atoms were being heated by the Lattice. Inside the DFG only, the temperature is ~52nK, and when i turn on the lattice while the ODTs are on and turn off the lattice before the ODTS beams are off, the new temp is ~5200nk, 100x hotter. I need to adjust the hold times and powers to decrease the how much its heating up.

Here is what i changed in the sequence to get these parameters

Original→ heat checking:

TrapsOff: with self.Lattice497HoldTimeHigh → without self.Lattice497HoldTimeHigh

self.ODTHoldTime: 20ms → 2s

self.EvaporationTime: 13.5s → 11.5s

self.Lattice497ImagingTime: self.ImagePulseTime → self.ImagePulseTime - 3s

self.ImagingDelay: 14s → 15s

Original

Heating Check

I also traced the ODT sequence to better understand what i was changing, here is what is found. In blue are the Self.labels of the values we are changing and in red at the groups they are in when you looking at the EvaporationWaveform2_MOD inside EvaporationWaveforms.py(in virtualbox)

Holzworth= 330,700 → 330,703 MHz

started the day with a RMOT OD =1.7. I adjusted last vertical rmot mirror and the both Red transvese retro mirrors. got the OD to ~1.9. I could get much higher then that. I checked the ODT at 1ms TOF and 1s ID and it was 1.8, which is where it has been. ODT at 14s ID and 5ms TOF. and its >0.78, which is where it should be. The 14s ID and 0msTOF for the lattice is showing > 0.5

Holzworth= 330,695 → 330,700 MHz

I was too narrow. I needed to take a wider scan

Some attempts for finding the Rabi freq

Holzworth= 330,691 → 330,695 MHz

spent the day looking for the clock narrow transition, and I also finished the Rabi Freq graphs from yesterday

This is the wide scan at 37mW and 50uW, we notice the 2 side humps seem to go away. These humps could be due to Rabi detuning, where New_Rabi=sqrt(rabi^2 + Detuning^2) Atom count= 0 at 582.6125 MHz, in DFG it came back at 400uW (DFG)

Used 50uW in the DFG and the Lattice. the DFG is more clear to see the transition. I saw the transition in the lattice but it was at 4mW. I didn't collect that data, but the drop was near 582.545MHz

These are my attempts to find the Rabi Freq. I subtracted the 6us or 12us steps from 500ms. Even though it's very damped at 500ms, there should still be an oscillation.

Holzworth 330,687 → 330,691 MHz

I check my Trap depths with tyler's equations, who checked it with Boyd's calcultaions. The key difference with my equations was they used a single transition instead of using the polarizability calculated.

$\alpha = 1347.5 a.u.$
$waist=300nm$
$power=1mW$
$U_{trap} = \frac{4P\alpha}{\pi c \epsilon_{0} w^{2}}=179MHz=8.6mK $, Im currently unsure how to check this further
$ \nu_{r} = \frac{2}{2\pi w_{0}^{2}}\sqrt{\frac{16\alpha P}{c \epsilon_{0} M \pi}}=18MHz$
$ \nu_{z} = \frac{2}{2\pi w_{0} \lambda}\sqrt{\frac{32 \pi \alpha P}{c \epsilon_{0} M }}=48MHz$

From Tylers notebook: Boyd's questions for how to design trap (pgs. 87-90):

1. Is λm is a practical wavelength, and is the required frequency stability reasonable?
   1. ($U_{trap}=19276 E_{r}$)
2. Can we perform spectroscopy in the Lamb-Dicke limit (νT rap > νRecoil) and in the resolved sideband limit (νTrap>γClock)?
   1. ($\nu_{recoil}=9.3KHz$), ($\gamma_{clock}=1mHz$)
3. Can we efficiently load atoms into the lattice (U0 > kB T )?
   1. ($U_{trap}=8.6mk$)
4. is the absorption rate for lattice photons at λm and U0 an issue?
5. is the light shift polarization dependence at λm and U0 significant?
6. Is the fourth order Stark shift at λm and U0 negligible?

Wasnt able to see the clock. after Grady troubleshot the issue, he determined that sweeping was making it hard to see the transition if we were close, so he stop sweeping and scnaned for it. we saw it at 582.545MHz I took scans at 37mW and 3mW in order to find the expected power broadening. And i scan the pulse time at 37mW to find the rabi frequency.

Holzworth 330,682 → 330,687 MHz

Each table correspond to a folder with the data file, located in :media/strontium/data/clock/*date*_*attempt#*

Wavelength used: 497(993.770950nm,301.671580THz), 698(429.228086(5)THz{into UV wavemeter}, 429.228083(2)THz{into IR wavemeter})

My goal for today was to find a more precise frequency of the clock transition. I took a good amount of data, but I'm having trouble with the code.

Note: attempt 1 was taken after midnight of (02/15/2024) Attempt 1:1mW,Pulse=70ms 550um waist, 1MHz Sweep,DurationUP=10us, 3 averages

Attempt 2:4mW,Pulse=35ms 550um waist, 500kHz Sweep,DurationUP=10us, 3 averages

Attempt 3:4mW,Pulse=35ms 550um waist, 500kHz Sweep,DurationUP=10us, 3 averages

Attempt 4:800uW,Pulse=35ms 550um waist, 1MHz Sweep,DurationUP=10us, 3 averages

Attempt 5:800uW,Pulse=35ms 550um waist, 1MHz Sweep,DurationUP=10us, 3 averages

Attempt 6:800uW,Pulse=35ms 550um waist, 1MHz Sweep,DurationUP=10us, 3 averages

Holzworth 330,679 → 330,682 MHz

Each table correspond to a folder with the data file, located in :media/strontium/data/clock/*date*_*attempt#*

I searched for it too, and here is what i found. Wavelength used: 497(993.770950nm,301.671580THz), 698(429.228086(5)THz{into UV wavemeter}, 429.228083(2)THz{into IR wavemeter})

Attempt 1:39mW,Pulse=150ms 550um waist, 1MHz Sweep,DurationUP=10us, 3 averages

Attempt 2:4mW,Pulse=70ms 550um waist, 1MHz Sweep,DurationUP=10us, 4 averages

Attempt 3:4mW,Pulse=70ms 550um waist, 1MHz Sweep,DurationUP=10us, 4 averages

Attempt 4:4mW,Pulse=70ms 550um waist, 1MHz Sweep,DurationUP=10us, 4 averages

Attempt 5:400uW,Pulse=70ms 550um waist, 1MHz Sweep,DurationUP=10us, 3 averages

I went over how paul and Tyler got the clock transition in the 813 trap. I believe the 50mW of power used was not for the 698 incident power but for the trap power. Because the Rabi Freq and Power broadening do make sense. I might be missed something, but i cant reproduce the the same values. Tylers took note from this paper of 3D lattices, where the the polarization of each dimension is different which results in a different magic wavelength seen by the observer. I.E., if the quantization(magnetic field) axis is in the Z directions. the Z-trap needs to be Vertically linear polarized, and for X and Y axes, it needs to be Horizontally linear polarized. Being linearly polarized minimizes the the Tensor part of the AC STarke shift, being Horizontal or Vertical, minimizes the Vector part of the AC starke shift. and being either horizontal or vertical depends on if you want to drive a Pi, sigma- , or sigma + transition.

Grady found the clock transition, check his notebook on (02/14/2024)

Holzworth 330,664 → 330,669 MHz

Holzworth 330,649 → 330,664 MHz

went over the Power_Broadening code with grady. It was good, but can be improved. THe graph below show how the Power Broadening and rabi frequency change, depending where the beam hits the atoms

698 Clock, 39mW 550um waist

Best 698 Clock, 39mW, 170um waist

698 Clock, 39mW 600um waist

also went over picomtotor alignment. we determined that we are within 200um from the waist, we found this out by moving the picotmotor in 1400 steps range, when using eqtn (tan(7e-7 *1400)/2rads)*0.4) this gives was 196um range

when it came to the guide beam sequence, we were imaging the mot during the guide beam, we switch it to after

Holzworth 330,644 → 330,649 MHz

I went over power broadening again, becuase i was missing something crusciual. I looked in Tylers note book and i had missed the line where he said the FWHM in Foot's book is the Power Broadening. I had already used that equation in mt Power Broadening code, but i hadnt made that connection. Once i realized that, i was able to quickly find what our power broadening will be for a give power/waist/wavelength.

Currently, at 39mW at 550um, our power broadening is 30kHz. And 39mW at 170um, it goes to 90kHz.

using that and some math i went over with grady, wecan find our pulse length the atoms need to see.

Time of resonant light is on the atoms= T_res= (Power Broadening)* DurationUP/Sweep

Number of Sweeps =N_swps= Rabi Time/ T_res

Total Pulse Time= T_pulse= N_swps * DurationUP.

ie) PwrBrdn=30kHz, DurationUP=10ns, Sweep=1MHz, Rabi Time= 43.5us

T_res= 0.3ns, N_swps=145,000 ,T_pulse=1.45ms

The purpose of these 3 attempts is to get a wide 20MHz scan around where we expect the clock transition to be, with a tighter scan near the expected spot.

Attempt1:39mW,Pulse=1.45ms 550um waist, 1MHz Sweep,DurationUP=10ns

Attempt2:39mW,Pulse=1.45ms 550um waist, 1MHz Sweep,DurationUP=10ns

Attempt3:39mW,Pulse=1.45ms 550um waist, 1MHz Sweep,DurationUP=10ns

I finalized the Power Broadening code, not only does it give the power broadening but it also tells you how it change depending on where the beam is hitting the atoms. Power_Broadening.ipynb

Holzworth 330,642 → 330,644 MHz

I kept thinking about how to make sure the guide beam is aligned to the atoms, how can i be 100% sure. I recalled the Guidebeam in the sequence opens the StirSHutter while keeping the MOTZShutter(under Sr table) closed, but the StirShutter also opens during the normal RMOT. SO i check the RMOT and its getting blown away, from OD of 2 to 0.9. It gets fully blown away at 0ms TOF and 70ms Image delay, at every point beyond that, it never appears. This tells me, when i thought i was aligning to the (5ms TOF, 14s I.D.)atoms I was wrong, and i was actually aligning to the rmot.

I used the OFL shutter directly in front of the input collimator of the guide beam. I added it to the same sequence as the Stir shutter for the guide beam code. Note for future shutter work, look at the TX1 in the sequence. and make sure the lines arent too close to others. with the shutter in place, i could see if the beam was aligned to the atoms. at the 0.01 intensity, the 14s I.D. & 5msTOF came back. I aligned to the located of this mot, and reach a florr intensity of 0.00001 where the mot would entirely go away no matter how good the alignment was, this leads me to believe that i am closest i have ever been to the atoms.

Holzworth 330,637 → 330,642 MHz

I wanted to make sure we were on the atoms, and i remembered that i could decrease the power of the guide beam. This would cause less atoms to be blown away, and i could better align it the atoms. I used the half wave plate to lower the power of the guide beam, and the atoms came back. But after 3-4 runs the atoms went away again. I measure the power of the beam, and it was fluctuating, between 20uW-60uW( this was not guaranteed to be the power at the atoms). i did an ER measurement and increase the dB form 25 to over 40. this stabilized the beam at 30uW.

while fixing this, i had bumped something and i spent the rest of the day attempting to bring back the RMOT, as detailed below.

at 4s load time got it to low 1.5s at 7s load time got it to mid 1.93 +/- 0.3

only touch the trans @ red and its retro, and the HWP to divide the powers, i was as to increase the coupling into both vertical red paths but now the stir and rmot vertical arent the same as on the wiki

Holzworth 330,634 → 330,637 MHz

SInce move the 698 telescope lens, i dont measure the polarization. It was not horizonal, like Paul and Grady sad it should be. I will retake the data from yesterday but also, move(remove) the focusing lens to increase the beam size at the atoms. to get the same 10uW 125um waist that paul had. with our ~25mW, we need a waist of ~6mm, and the colimated 698 is at 500um waist. Plus i can use the beam profiler to near perfectly overlap the guide beam and 698.

Pauls Intensity was 203.7 mW/cm^2 I set the powers to max, its was at 25mW, but it was clipping on the stir shutter ethernet cable. the at 39mW and 550um waist he pulse pulse range in 33us- 50 us

Attempt1:23mW, 33us & 50us Pulses 550um waist, 2MHz Sweep

Attempt2:23mW, 33us & 50us Pulses 550um waist, 2MHz Sweep

Attempt3:39mW, 33us & 50us Pulses 550um waist, 0 Sweep

I didn't find the clock transition, but I saw general dip bwtn 583.83-583,86 and I need to check (583.92-583.93), (583.98, 584.10) at 500Hz steps

I got the OD back to 0.85 at 14s Image Delay and 5ms TOF, throughsome adjustment of the last ODT2 mirror

I tried scanning for the clock transition but I didn't find it. Here is how I scanned for it:

Attempt 1: No Sweeping, 15uW(later to be incorrect, actually at 8uW), 250ms Pulse, from 583.42 to 585.42 at 0.01MHz steps

Attempt 2: No Sweeping, 15uW (corrected), 500ms then 250ms pulse, from 583.42 to 585.42 at 0.01MHz steps

Attempt 3: 1 MHz sweeping, 30mW, 12.5us then 10us pulses, (these were from the range of 40 - 50 kHz rabi frequencies.

grady found the problem. It was alignment. I had bumped the last transverse2 red mirror. Check retroreflections of the transvers beams and see if there are pointing some whatcolinear. also check Zshim and move atoms, this can tell you if the mot has been moved (with the mot beams)

I spent saturday and sunday trying to get the MOT back. I made a check list over all the is things i could check, with alignment being the last, because its not perfectly repeatable.

• Shutters(Are they working, Are they resetting to the same position. Is light getting through)
• Are the lasers locking to the right frequency
• on bmot PID, make sure red line is reaching set point. if not, better couple 461spec on laser table
• in 922 Solstis: Output PD should be around 780mV, Do this by locking below wavelength and ramping the RefTuner up
• imaging DDS is at 162MHz
• Is the chiller ON
• Are unnecessary beams turn ON
• Check laser stablility, aka extinction ratios
• Is the timing for camera correct, go over the sequence
• in “imaging_config.json” Filter is 0 for ODTs(1s,1ms). and Filter:1 for everything else
• can you get into steady state, how does the BMOT florescence look
• Check all the powers, is everything reaching setpoint.
• check the shape/ size and location of the mot
• Check the Z Shim, (is the mot gets better, a optics was mostly bumped
• check the Mot beams retroreflective alignment.(in the retroreflector beam don't coalign w/ the beams, something was bump
• Check oven absorption
• Last, Alignment
• Ask the atoms what's wrong. -Delanie

i scanned for th eclock transition, I started with Holzworth from 584 to 594 MHz, at 0.1MHz steps, sweeping 1MHz over 11ms. with a waist of 160um and ~30mW power,

I then tired 583.42 to 585.42 at 0.01MHz steps, at 15um. I notice no dips in OD, but I saw the OD slowly drop over time form about 0.8 to 0.6.

I tried increasing the waist to make sure we are hitting the atoms. I thought of moving the second lens in the telescope, but I bumped the cage mount for the first telescope lens. I took some times but i got the telescope back with alignment and both on a cage setup. The side-to-side movement could be repeatable with some effort, but the forwards and back along the z-axis are a bit harder. I measued the 2W of the beam when collimated is at ~1000um. The focusing lens focuses the beam for about 4cm away from the atoms. but the waist at he atoms is still 160um.

After fixing the bumped telescope, i saw a low atom count, but since it was late, I checked for repeatable solutions such as sequence timing, and beams pathes for potential blockage.

Here is it the location of the guide beam when enters and exu=its the chamber, its sitting on a 1inch post

here is the 698 path on the Sr table

I tried to run but the 689 wasnt locking, , i was reading sls manual to lock it when grady showed up. He messed with it a bit and determined it was multi mode and you can adjust the Master set currenet knob( top right) 1 or 2 turns to get it away from multimode.

I notice the 698 was eliipical after the aom in -1 order. I tire adjusting the knobs of the aom and the last mirror before the aom. ther ewas change but nothing seemed to make it gaussian. grady help it become circular, i didnt physically see the adjustment, but he change the last mirror going into the aom. Im unsure what he did differentlly.

i shadowed grady today:

• he added in the redguidebean in the main.pys. i triggers the shutter after the TOF time.
• he adjusted the last small mirror of the guidebbeam/clock path until our 1ms:1s ODT went away.
• we lowered the intensity but even at the lowest setting it still goes away, i move the tip and tilt to the center of where the beam when away and we were good
• make sure to block the guide beam, because min amount of light is enough to blast the mot away
• he added in the clock to the sequence. we currently are keeping the the spectrum card on and only switching on the OFL1 aom.
• Next, we are doing at temperautre meassure of the atoms
• I beleive the next steps are to adjust the holzworth until we see a dip in the OD, becuase due to the guide beam, we assume we are hitting the atoms.
• you change the Holzworth via the beaglebone green, and the intrustions are Here

I work on the clock laser alignment and watched grady lock the laser. but after TAing, Grady determine we needed the -1 order of the aom, so i realign to the that order and copropagated the guide beam. I will look at the waists tomorrow

Holzworth 330,610 → 330,610

From CSV :698Holzworth IP:172.16.74.146

• Updated the label on the Physical 689Holzworth from 172.16.74.62 to 172.16.74.173(from CSV)
• Paul connected the Beagelbone(IP:146) in the 698Holzworth path. It is currently turned off.
• the beagle bone is it the CVS file in column 146, with password claroqueclock (ssh debian@IP)
• make sure the phase noise on SLS has a signal, increase the power to the cavity
• if able to get in run : ipython -i test.py
• at this point we are unable to access the beagle bone green, we looked up Booting BB from the SD card
• we tried to force the BBG to boot form the SD by, Holding the Boot button while turning it ON
• Current theory, the script inside the SD card was not backed up and was go during the lock shut down
• connected the SD card to the Sr computer.
• tried searching for the Static IP for the BBG
• in terminal, cd into media/stronitum/engineering/ software/clock_bbg
• we physically connect the BBG to the computer
• grady tried a wide range of things to connect to the BBG via ethernet, but most things didnt work, he noticed it would connect while being physically connected to the computer.
• we switch the power supply cable and It Connected.(something was wrong with the old one)
• Here is how to control inside the BBG from tyler Notebook
• Grady tried the freq i calculated and eom didnt change the beam. As well as the last found freq didnt work.
• Its not a EOM, its a AOM.
• In strontium/experiment/sls, Use the FNCS-1000-1_manual.pdf to turn on the AOM. In section 3.5 AOMScreen
• got the aom to turn on and we see it on the dlc
• the way to lock this laser is the same as the 689 but we need to know our starting point
• Using Tyler notebook entries of 5/3/22 and 5/12/22, we know what the wavemeter should read to the hundreds of kHz.
• Note: The SLS cavity freq is constant, and it equals the sum of the wavemeter freq + the holzworth freq, sowhen you lower the holzworth you are increasing the wavemeter frequency
• We decided to hold the holzworth at 350MHz( at the middle of its range 50 -750 MHz) and scan the DLC, while looking for dips in the blue line(Fast In 4)
• Max Current for the Master Current on the DLC is 70mA
• 689 wavemeter freq = 429.227970THz , then subtract 350Mhz from holzworth, to see 429.277616THz
• Currently as we change the holzworth by 5Mhz, the wavemeter shows 15MHz change. This should be 1:1 not 1:3
• the 698 was in multimode, and that why we see a 1:3 shift. at the TEM00 mode, the freqs changes are 1:1

Holzworth 330,592 → 330,608

Intial output Stirring: 0.314mW/1.7mW =18.47% current output stirring: 0.47mW/1.7mW = 27.65% (can couple it higher, but it was already reach set point)

initial output Rmot: 10.6mW/ 15.5mW =68.39% current output Rmot:

In Foot: a pi pulse is the when the rabi freq * duration of pulse = pi

In Ivaylo Madjarov thesis they had a rabi freq of 2pi* 2.8kHz,there peak pi-pulse should be at t= pi/(2pi*2.8kHz) = 1.7ms which it shows on figure 3.2

I tried to determine the reason for the non perfect Gaussian shape of the 698 after the chamber. After back reflecting it through the last lense, it was still there. I went to align and overlap the guide beam and notice it was terrible. I determined it was due to the collimator, and i replaced it. I adjusted the waist to be with the Zr of the 698 beam and aligned it as well.

I looked at the waist of the 698 beam on the atoms and determined it was at 180um waist, located maybe 16mm away from the atoms with a Zr of 13cm. But i notice the beam leaving chamber was not gaussian. I reflected the beam from beofre the chamber window back to the computors and saw the disfigurment was due to either the picomotor mirror or the 400mm lens. I used the last adjustable mirror(not picomotor) in the path to Back relfect the beam onto the the 400mm lens and it might have fixed it

The values of Z correspond to the pos=[] list in the pwr_broaden_example.ipynb

z1_.5

z1

z2

z2+.5

z3_1

z3

z4

z5

z6

z7

Holzworth 330,587 → 330,592

grady went over with me the different s ways to bring rf to a amo.( bimaging power), you can use a VVA for better control, you can use the DDS for speed.

we determine the size of the 698 through the aom was too large and was clipping. after he iterated between the colliamtor and tip/tilt of the aom he got a better looking beam enough through it wasnt collimated. the “telescope 30 to 35mm lenses were there to re-collimated the beam into the chamber and the last 400mm lenses focuses it down. I notice the beam coming out of the change wasnt nicely guassian,but since grady needed to run. I will coming in tomorrow.

Holzworth 330,580 → 330,583

I went to set up the flowmeter/pressure gauge. and found that we got shipped a valve instead of a flowmeter. It comes the next day.

I determined where the 698 holzworth is suppose to be by comparing it to the 689 holzworth and the drift per day. All my calculations can be found in pwr_broaden_example.ipynb. At a 6.5kHz/day drift, the 698 Holzworth should be near 743,516kHz

grady found holzworth

when tyler found clock transistion

Holzworth 330,575kHz → 330,580kHz

I took more waist measurements, I seems like the waist is (154 +/- 4)um but its positions about 8 cm closer to the laser, so i need to make sure that either the laser is collimated or change the lens positions

Here is the paper on the 813 measurement

Spent the day designing a flowmeter/ Pressure gauge set up for the chillers, because if we are adjusted them we should be sure what they are at.

afterward, I took some waist measurements of 698 through the chamber.

I tried to get the EFf up on the AOM, I noticed if i changed the RF power, it would increase and decrease the Eff. This would make sense, consider what a VAriable Volatage Attenuator does. I change the Voltage to the spectum card and it change the rf power to the aom. Even though the spec sheet says 24.8 dBm, the Eff for that rf power isnt that good, so I will keep it around 85% or 170mV, its not super stable by that probably because of the PID being consistantly on.

I set up the 689 guide beam path, it shouldnt be in the way of another laser. Plus is pretty aligned to the current 689. I need to characterize the waist at the atoms for both 698 and 689. As well as implement it into the sequence with a potential shutter.

120mV 65.2% 26.2dBM

130 mV 71.6% 26.9

140mV 76.9% 27.5

150mV 81.4% 28.1

160mV 84.7% 28.7

170mV 86.6% 29.2

180mV 87.6% 29.6

190mV 88.2% 30

200mV 87.4% 30.4

Holzworth 330,570 → 330,573 kHz

the distance to the 698 collimator to the window is ~84.5inches

When turning on the SBench from the computer restarting. In either 698 or RJ starter project. Assign the 80MHz_200ms.sb6dat to AO-Ch1, and change the Amp to Vpp=120mV. This should get you 24.9dBm at the 698 aom.

Holzworth 330,565 → 330,570 kHz

I started to set up the 1064 for the guidebeam needed for thr 698 path. After talking to Julio and Grady, we switch to using the down wards 689 beam,and put it along the path i was just creating for the 1064. While checking the waists of 698 to calculate the damage threshold for each optic. I learned the 698 waist was changing way too much and decided fix it. I reflected it a total of ~143.5inches or 363cm at 2.7mm diameter at the lowest waist. I had to TA, and didnt measure the distance to the atoms

Holzworth 330,560 → 330,565 kHz ludlow thesis maybe for clock stuff

I called Laird to ask about the wetted materials for both there chillers but no answer. I hoping they will message me back soon. I also called digikey, they were unable to answer the question, but will contact Larid aswell.

New Eff for the 698 path:

Pbs1:56 to 51.5= 92%

aom:51 to 48=94%

pbs2:48 to 46= 96%

Currently their is ~42mW at the atoms of 698. the output of the fiber 60mW which gives a 70% eff.

I think i got a working power broadening graphs, but my Full Width Half Max equation seems off. I using Foot's eqtn(7.88).

Holzworth 330,550 → 330,553 kHz

Holzworth 330,547 → 330,550 kHz

I have tried different optical confiurgations to get the badly shaped 698 diode laser to produce a better Eff out of the first fiber. So far the best coupling i can get is 58mW out of the 110mW(no iris), which is 52.7% Eff. This Eff is done by using 2 mirrors, 0 reshaping optics, and schafter+Kirchhoff 60fc-4-M20-10 collimator

Holzworth 330,542 → 330,547 kHz

I estimated the waist of 698 to be about 260um. To get the power broadening near Pauls, we will need a power of 40-45mW at the atoms. While using the power broad eqtn from steck, i was confused about the relationship btwn Omega/Gamma to Int/Isat. The example graph in the book lead me to believe since Omega was close to Gamma, so Int should be close to Isat, but this isnt correct for our case. Isat for 698 is extremely small, on the order of pW/cm^2, mean while we want Int to be a factor of 14 larger. So, refering back to Steck's graph, using Omega close to Gamma is a rescaling. To understand this more Omega is Rabi Frequency and i need to read about this.

Int/Isat scaling for different pwr(260um waist):

pwr at the atoms(mW)[25,30,35,40,45,50]

Factor of Int/Isat[5.5e13, 6.6e13, 7.7e13,8.8e13, 9.9e13, 1.1e14]

698 Waist:

Wz[419,320,287,262,308,325]um

pos[-273,-163,-55,25,206,269]mm

419um at 133mm from window: -303mm

320um at 23mm from window: -193mm

from center to mirror is 93mm

287um at 130mm at mirror : -85mm

262um at 210mm after mirror: -5mm

308um at 36mm after window: 206mm

325um at 99mm after window: 269mm

windows are 340mm apart so center is 170mm

Holzworth 330,537 → 330,542 kHz

Called Andor again, they said their engineers are in, and they will get back to me.

I worked on learning the Power Broadening, and i was able to replicate the graphs for Dan Steck, but the I/I_sat is huge. I need to read more about it

tracing the 698 power in order

the largest drop in Eff is the aom at 58%. The manual says the Eff should be 85%, which would change the total Eff to 48%, which isnt good enough. the 2nd largest drop is the first PBS at 77%. It could get to 90%.

If the 2 largest drops of Eff were at their Spec range, our total eff would be 56.4%, and the power needed to get 50mW at the atoms would be 89mW out of the fiber. Also some of the mirrors have lower than normal eff, so they should be cleaned up to hopefully increase it.

Currently we are missing a Dichrotic mirror, so our 689 power are lower. For now we will need to Load for 9s.

the 922 solstis ECD output should be about 1.1V

Everything else should be normal

I called andor about the Chiller they recommend not meeting their specs. They said they should email mean with the infomation.

Found our gloves, and they are order.

I got a incomplete equation for power broadening working in jupyter. I need to complete it

I move to a thorlabs f=11mm collimator becuase the the shap of the beam is similar size.

I was able to immediately got to 1.8mW but nothing. But as writing this, i realized if the actual power of the laser is ~71mW without the garabe part of the beam, and i had ~50mW out of the fiber, i could increase the Laser Curent higher to get 50 mW at the atoms????

I meant to post this Picture yesterday, its the 3x mag + 250mm Cyl lens

Just the 250mm CYl

Ideal beam with f=5mm collimator

Ideal beam with f=11mm collimator

when using the F=5mm collimator,I got to 15mW pretty quickly, so i will keep trying to get to a higher Eff

I work on getting the 698 power higher. I needs to be at 50mW at the atoms. Before i got 50mW after of the fiber, but high 20s at the atoms. i took images on the of the beam to try and change it. I currently using a 25mm to 75mm telescope for 3X mag, and a 250mm Cylidrical for reshaping.

This is what the beam should like(Taken by shining light through opposite end of the fiber(schafter+Kirchhoff 60fc-4-M20-10))

What the beam looks like out of the laser

Immediately after the telescope without cyl lens

Far after the telescope without cyl lens

I forgot to take images with the cyl lens, but the most spherical beam shape was at 282mm after the first mirror. While the output of the laser is 106mW, i put a iris to block the bad part of the beam and got 70mW. I will try to get a high eff through the fiber and then increase the current of the laser, because it should be lower then it needs to be.

In any terminal type:

cd ~/.scripts/

python laser_log.py sroven 689 on

Spent part of the day reading over the 2 papers about magic wavelength measurements.

1. Sr87 813 magic-clock transistion by YAMAGUCHI Atsushi

2. Sr88 473 Magic Measurement by Barreiro Group

My todo list for the quarter is:

1. get the 698 power to be as closer to 50mW as possible at the atoms.(for power Broadening)

2. Get 698 aom to turn on during with in the sequence

3. understand the stability of the 698 laser, and know how to adjust it.

4. Remeasure the 497 Spec I did in the previous quarter, becuase the entire breadboard with changed, and i need to reference the 2. paper above for insight on the reference scan.

5. Determine the excitation ratio for varying clock frequencies, as shown in the 1. paper above.

6. More to come

0. Look at Boyds thesis and his 813 measurement