Jared lab notebook
--Jmiles2 (talk) 12:39, 13 May 2014 (CDT)
I need close to 17 mW of seed power for the 2nd FORT beam to work at 2 Watts. Although this is is in the range of acceptable powers for this TA I still want to contact M2K lasers and make sure the Intensity is not too high. I can only get about 40% through the AOM when powering it with 2 Watts. I get around 33% out of the fiber at higher power (~600 mW). I get over 200 mW from the output of the fiber.
Need two AWG outputs for evaporative cooling of both FORT lasers. For the FPGA, FORT on/of Channel 14, is a switch to the FORT 1 amplifier, and turns it on or off. The RF in to the amplifier is controlled by the AWG which is triggered from CHannel 10, FORT AWG. OFF means the AWG is triggered.
--Jmiles2 (talk) 14:25, 14 May 2014 (CDT)
For the AWG/hittite chip on FORT 1, 3db input on CW setting is equivalent to 33 to 33.2 dbm output power, 2 Watts, maybe a tad more. Since I'm not running this all the time, I think slightly more than 2 Watts should be fine. On pulse setting, 0 V input gives 11.5 dbm and -700 mV gives 32.8 dbm. A little over 20 dbm attenuation, that should work fine for what we're doing. The majority of attenuation happens from 0 to -200 mV.
I made an identical circuit for FORT 2. at 0 V, the output is 33 dbm for 4.5 dbm input power. The input power might change based on what RF generator you use, always check it first. At 700 mV the output is 4 dbm, 29 dbm range, I think the circuit for regulating the current to the chip is slightly off in the FORT 1 circuit. 4 dbm input give 33 dbm at CW setting.
--Jmiles2 (talk) 11:05, 15 May 2014 (CDT)
Need to get FORT 1 laser to one polarization. Alignment affects polarization, looks like most of the light is vertically polarized. 19 amps give me 20 Watts in vertical direction. ~20 Watts is max for the AOM. Put in a wave plate to change to horizontally polarized so it would go through the beam cube. Still get good efficiency through beam cube, 45% to 50% at high powers. I get a little over 7 Watts after the beam cube.
FORT 2 is now aligned to the MOT. I can get atoms trapped with either FORT now, they are aligned over top of one another and I can load FORT 2 from the atoms in FORT 1, with no MOT. Need to check lifetimes of both.
--Jmiles2 (talk) 16:14, 19 May 2014 (CDT)
Lifetime of FORT 1 is now much longer than it used to be. I'm not sure why, but it might be because the laser power is not jumping around as much. It was moving around by as 20% on the photo diode when evaporative cooling at the end of March. The only thing that is different is that I have the laser going through a beam cube, and have optimized the alingment so that the Horizontal linearly polarized light is maximized at ~18 Amps of laser current. 18 to 19 amps should be operating current, around 20 Watts on the AOM.
Taking different temperature measurements for different evaporation times, up to ~400ms.
--Jmiles2 (talk) 10:09, 21 May 2014 (CDT)
Lifetime data from Both FORTS.
Lifetime Data
Fort Lifetimes. Green is the high power FORT around March 2014 and Black is the same FORT in May 2014. Red is the TA FORT, also taken in May 2014.
--Jmiles2 (talk) 17:17, 24 May 2014 (CDT) Put a shutter after the TA for the MOT beams. This improved the lifetimes by around 150 ms. Can get temperatures equivalent to ~16 nm/us. I should do a temp scan with more data points, around 10 to get a more accurate number. All of the temp measurements with the 1064 FORT have 200us exposure time.
--Jmiles2 (talk) 18:13, 30 May 2014 (CDT) Coupling beam on resonance 204.5 Mhz. Probe resonance at 220 MHz. No EIT yet. I think it's a laser issue, this happened about a month ago when I tried to get EIT with the 1064 nm FORT. I'll do the threshold test on the laser and find the peaks again, maybe put it through the interferometer.
--Jmiles2 (talk) 18:18, 9 June 2014 (CDT) Need to find EIT again. Polarization might be an issue, need them to be opposite linearly polarized I think. Only using the Large FORT laser to trap atoms. Peaks look fine on the laser, and linewidth seems reasonable. I could try another diode, but I don't why I'm not seeing anything :(
--Jmiles2 (talk) 14:17, 12 June 2014 (CDT)
EIT works, although the transfer should go to zero at high control powers, it seems to go back up at much higher powers, ~50mW control with 15 mW probe. beam sizes around 1mm. I can check my background better by looking at the two spots where I add together the pixels with no FORT. Subtracting these two should give me zero but there will be some small error in this.
--Jmiles2 (talk) 18:01, 16 June 2014 (CDT)
I can get EIT on both beams, and they both do basically the same thing, the curves aren't exact, but they never have been. The second EIT pulse looks slightly different than the first but the most transfer still occurs around 3mW. I think the standing wave is working but it isn't consistent. There is a pattern there, but it's much noiser than ones that I have seen before.
--Jmiles2 (talk) 12:42, 20 June 2014 (CDT)
Can't get a good standing wave pattern, even though EIT curves look good. I think it might be something with the FORT. It's possibly too inconsistent to see the small change (~20%) I need to see in the standing wave patter. I looked at scans of the Large FORT yesterday and it looks like the signal was oscillating much more than it was when I did EIT before with the D1 line with the 850 nm FORT. I'm going to try to get the 1064 TA FORT working so I can see if the signal from those FORT atoms is more stable.
--Jmiles2 (talk) 10:19, 23 June 2014 (CDT)
Need to watch the power out of the TA, I think something might be wrong. Today I have 12.8mW input power and 1.41 Out at 3.106 Amps. I thought I was getting more a few days ago. I also think I need a longer focal asphere to better couple to the TA. I have a very short focal length asphere right now.
--Jmiles2 (talk) 18:35, 24 June 2014 (CDT)
I'm going to work with the TA_FORT for now. I can get some evidence of standing waves with either FORT laser, but the better data was taken with the TA FORT. I'm going to look more into the statistics of my measurement. This could be a problem that I haven't addressed up to this point. The standing wave hasn't always worked in the past, the one Think that I thought fixed it was a loose mirror, but I was never sure that was the case. The standing wave shouldn't be that hard to get though, once it works, it usually stays working fine for awhile, over a month.
I get a decent signal with the TA FORT with only ~220 mW of power. The beam size could be expanded bigger to make a deeper trap.
--Jmiles2 (talk) 10:37, 26 June 2014 (CDT) I think noise might be a problem. I started looking at the fluorescence from the MOT with a photodiode. The signal oscillates at around 120 Hz, the same frequency that the feedback to the laser piezo is oscillating, I think this is changing the frequency of the beam slightly and a different number of atoms is loaded into the MOT. It would also change the number of atoms I detect in the FORT because I use the same laser to look at the fluorescence of the FORT atoms with a photodiode.
--Jmiles2 (talk) 15:56, 27 June 2014 (CDT) Noise in the FORT atoms go down when I turn up the gain on the MOT seed laser. There is a standing wave, but I can't get a good plot that shows more than one peak. I think this is the same problem I've had before but never quite figured out what was wrong.
The TA before the second High frequency seems very sensitive to alignment. Over the course of the day it got unaligned, and I had some trouble getting it back, about 265 mW out of that to get ~170 through the high frequency AOM.
--Jmiles2 (talk) 23:08, 1 July 2014 (CDT)
Standing wave working much better now, up to 25%, not sure why. The last thing I did was realign the coupling beam from ChB on to ChC, it didn't seem to do much, the power I got through the fiber went from ~4.98mW to ~5.02mW, not much at all. I also did the threshold test on the seed laser again, it did improve it, but I didn't see a better standing wave immediately after. It also looks there is some asymmetry with the standing wave. I get different results whether ChC or ChB has the larger coupling beam power.
While things were working, I checked to see if the collection size (how many camera pixels I take as data) affected the percent difference between nodes and anti-nodes. I doubled the size and saw little difference.
I see a good standing wave with the Big FORT laser. Maybe some narrowing too. I might have even been using the wrong settings, I think I was using the saved state 9 on the delay generator instead of 8. They should both be very similar, so I don't think it's a huge problem, but I can't remember for sure which one I was using for the data I was taking, it should be 8 but I think it might have been 9 for some of it.
Either way, I still got really good contrast with the standing wave, and contrast with the larger FORT.
--Jmiles2 (talk) 15:55, 3 July 2014 (CDT) When the laser is pulsed, there is a different amount of power that gets through The AOM than when it is on CW. For Ch.C. the pulsed power is about 10% less that the CW power. For Ch.B. the pulsed power is about 5% higher than CW. This could cause some differences in what I set the power to. I want to check this every few days, because I think I'm getting slightly different numbers when I test it. I'm focusing the CW into a photodiode, and then switching to pulsed, seeing what the difference is. I got 15% instead of 10% for CH.B. yesterday...but I'll go with 10% for now.
It would be helpful if I could measure that pulse power by itself.
--Jmiles2 (talk) 14:02, 17 July 2014 (CDT) I still get a different signal depending on which coupling beam I have larger. I increased my signal to noise by using the binning feature on the camera. This combines pixels on the camera, making my signal about the same, but the noise that I subtract is now much less. The noise is more off an offset now, and doesn't affect the pattern that much. This also lets me go to lower power, which should reduce the Fort Temperature.
There are still other noise issues I'm not sure about. I might just have to accumulate more.
--Jmiles2 (talk) 16:45, 18 July 2014 (CDT)
My signal is still noisy compared to the feature I want to see.
[Noise Data]
Here is an example of the standing wave pattern producing some contrast in the total number of FORT atoms. This percentage change is higher than the Noise, but I would like the noise to be much less. [Standing Wave Example]
--Jmiles2 (talk) 16:59, 25 July 2014 (CDT)
I changed the labview program so that it will for a trigger before starting the sequence for the MOT and FORT. Now it triggers off of the line voltage, so far I haven't seen any big improvements in noise.
[LineVoltage to TTL pulse]
--Jmiles2 (talk) 14:41, 29 July 2014 (CDT)
I have the camera working so I can two pictures during one cycle of the FPGA. I don't see a direct correlation between taking an image after evaporative cooling and before, but this may help if the laser power drifts over time. I also don't see a huge effect with triggering off the line voltage. I'm looking into the evaporative cooling being a part of the noise. Also check noise with and without repumper.
--Jmiles2 (talk) 10:20, 7 August 2014 (CDT) Noise seems to be much better when I take data more randomly and not in a row. For example, don't take 100 data points for the first data set, then 100 for the next and so on....Instead, take 200 points and every other point counts for the first data set, and the other points count for the second. I saw a big improvement doing this with this data that was taken with all FORT conditions the same and no EIT beams. Here are the graphs using the two types of data taking.
[NoiseReductionData 6Aug2014] [NoiseReductionData 5Aug2014]
--Jmiles2 (talk) 14:28, 13 August 2014 (CDT)
Took a lot of good data yesterday. A few notes though. On the 1.4mW to 81mW, the powers were 26 and 15.5 mW to start, but ended at 27.5 and 13.8. For the 3 to 100, the powers started at 17 and 34.4 mW and ended at 31 and 20. The 3 to 30 mW case started at 3.5 and 13 mW and ended at 13.5 and 3.1. The probe beam stayed in between 15 and 16 mW throught out all the runs. The average was probably around 15.5, I don't think this small change would have a huge effect on the data for the Probe beam. I need to figure out what is going with the coupling beam though.
--Jmiles2 (talk) 12:45, 14 August 2014 (CDT)
Best Narrowing Ever. The probe power was 16 mW, and was the same after the scan. Both coupling beams were 2.5mW to begin with and at the end CHB was 2.25 and ChC was 2.7. Noise eaters would be great to get rid of this small drift. The scan took around 70 minutes. I fit the three peaks using 3 gaussians added together as well as small linear offset.
[Less than 100 nm narrowing] [Picture of narrowing]
--Jmiles2 (talk) 20:23, 14 August 2014 (CDT)
One scan was with both beams at 2.5mW; At the end ChB 2.7mW, ChC 2.6mW.
ChC 3.5mW ChB 2.5mW. Ended with ChC 3.65mW ChB 2.73mW
ChC 2.5mW ChB 3.5mW. Ended with ChC 3.4mW ChB 2.4mW
Probe was always around 16mW
All three scans looked similar, saw some neat features at the low points of each plot.
--Jmiles2 (talk) 10:17, 20 August 2014 (CDT)
I think I figured out the narrowing that I was seeing. We were transfering atoms in a very large area using EIT and the remaining atoms were in a very small spot, less than 100 nm. So the feature we were seeing was not from EIT in a small region.
I'm changing the probe power to 5mW from the 15-16 mW I've been using. I notice a difference in the two photon detuning needed for the coupling beam right away.
--Jmiles2 (talk) 13:55, 25 August 2014 (CDT)
I've decided to try the experiment with longer EIT pulses. Up until now, I've only tried EIT with a ~50 ns pulse probe beam and ~100 ns pulse coupling beam. I think there are weird effects happening in the fall time of the of both pulses. During this time, the standing wave has different min and max powers and can be transferring atoms in areas other than the nodes of the original standing wave. The pulse shapes were accounted for in the simulations that Deniz ran, but maybe I'll see a more narrowing. I'm trying with a 100 ns probe and 150 ns coupling.
I get good transfer with only 2 mW of probe, and 3 mW of coupling.
--Jmiles2 (talk) 11:38, 29 August 2014 (CDT) For the plots from yesterday, 3to50 had 19.7mW and 6.7mW after the scan. 3to40 was 16.2 and 4.9mW. 3to30 was a bad scan. 3 to 15mW was 7.9mW and 1.16mW.
--Jmiles2 (talk) 17:54, 2 September 2014 (CDT)
I still have an imbalance with the coupling beams. I took some data on Aug 29 2014 and it showed that when I switched the beams I got a different standing wave pattern. Also, the average number of atoms throughout the scan was different. It looked like there was an overall higher amount of atoms when I had the higher powered coupling beam on CHC B, the same side that the probe laser is going into the chamber. I repeated the experiment today after checking all the timings, and got the same result.
I thoroughly checked the timings today. I checked the shape of each pulse by putting the photodiode at a distance from the fiber launch that was similar to the distance the laser would travel before reaching the FORT atoms. I averaged over many pulses on the O-Scope, saved the waveform, and then moved the photo diode to the next laser. I would align the pulse of this laser to the saved waveform. I then repeated this for the final laser in the EIT sequence. I had to repeat the entire setup for the second EIT pulse. In the end, I found that the timings were up to 5 ns off, but it doesn't look like this caused the asymmetry in features from switching the two coupling beams.
--Jmiles2 (talk) 17:32, 11 September 2014 (CDT)
Fort Became unaligned after replacing the RF generators with the small NOVATECH instrument. A coupling beam was also misaligned, so I went through the alignment process for all the beams and also checked the spot sizes of the coupling beams. Beam sizes are about 1.24mm X 1.15mm.
The asymmetry problem still occurs, and is very apparent depending on which beam's frequency I scan over. For a standing wave from 3 to 40 mW, scanning over the higher power coupling beam, around 16mW, causes a large slope in the standing wave data. The slope is also there if I turn the Lower power beam off, and use the high power coupling beam for two consecutive EIT pulses. If I scan over the weaker beam, I don't see this slope in the data.
--Jmiles2 (talk) 18:42, 1 October 2014 (CDT)
I've taken some good data from Sep. 24 to Oct 1. The data on the 24th is with larger coupling beams, approximately 2.2mm by 2.4mm. This was measured with the newport camera. Probe power was 3mW. The plots show a sine wave at lower power, and then some localization at higher powers. On the 26th I took some more scans at higher powers, but this time I had the probe power at 4 mW, this was a mistake, I wanted to mimic some of the results from the other day, but forgot to adjust this power to 3 mW. After taking the longer scans, I decided to reduce the beam size to 2mm so that I could got higher in power for testing 1 EIT pulse. The data on Sep 29th shows that as I increase the coupling power, the features get narrower and then this stops at even higher powers, I am not sure why.
--Jmiles2 (talk) 16:22, 22 October 2014 (CDT)
Taking scans from 0 to 10, 0 to 15, and 0 to 20. The power has been fluctuating on one of the beams, for 0 to 15, the power went from 3.75 at the beginning of the run to 4.2 at the end. The other beam stayed the same. At these lower powers I think these fluctuations might make a difference, but we should be able to test that with the simulations.
--Jmiles2 (talk) 11:05, 28 October 2014 (CDT)
The data I took Oct 22 looked very interesting, and I'm now trying to simulate it. I need to figure Out what things we'll need for the grant proposal in December. I think simulating these plots will be one of them.
I thought more about the coupling beam standing wave acting as a FORT for the ground state and we thought it might be heating up the atoms. I'm trying to prove it so I had the coupling beams on but no probe beam, and tried to measure the temperature of the atoms after the coupling beams were turned off. I didn't see any change in atom number or temperature with the beams on or off.
--Jmiles2 (talk) 10:13, 31 October 2014 (CDT) Took Beam size measurement of Probe beam, it had a 1/e^2 of 2.6mm, although I wouldn't be surprised if it was between 2.55 and 2.65, maybe a litte more. The coupling beam were measured a few weeks ago and were each 1.5mm. I took these measurements with the Thorlabs beam profiler.
--Jmiles2 (talk) 16:38, 3 November 2014 (CST)
Interesting things about the localization. When the standing wave is traveling toward the probe beam, meaning the beam on channel ChC is larger than ChB, the data becomes broadened in that direction. I notice that for lower, power coupling beams, 0 to 20 mW. I took data to show this more closely on Oct 31.
[Standing Wave Asymmetry]
Weird things happen when I change the Shim current. The features look different but they're still there. Changing the shims affect the total atom number, but doesn't move the atoms in any way, so I don't think it's an alignment issue.
--Jmiles2 (talk) 14:32, 4 November 2014 (CST)
This data is from Oct 30, when I noticed that changing the shim current on #2 from 4 Amps made a difference in the features.
[Shim Current Changes]
--Jmiles2 (talk) 14:00, 17 November 2014 (CST)
Since changing the shim currents changes the features and contrast, I think the problem is with the m levels. Not sure how to see that though, so we've but a large magnetic field on the FORT to split the shim levels by about 10 MHz. This takes 15 Amps with the current set up. The coils were already on the MOT chamber but haven't been used for awhile.
The idea is to seperate the m levels and only use 1 level to do the experiment. Then I'll change the shim currents slightly and see if I see the same thing.
The problem now is that I would need about 30 amps to separate the levels far enough using 100 ns pules. So I'm switching to 200 ns pulses, this cuts the width in frequency space by half.
--Jmiles2 (talk) 15:21, 2 December 2014 (CST)
I'll try to stick with the 100 ns pulses, but I want to make sure the levels are split enough with the current we have. We might make some more coils with 14 AWG wire instead of the 12 we're using now. We would use less current with 14 but be able to make a lot more turns, so I think we could still get a bigger B Field.
--Jmiles2 (talk) 15:54, 3 December 2014 (CST) Checked the Beam size of the Coupling beams, I thought they hadn't changed since the last time, but ChB was ~2 mm diameter 1/e^2 and ChC was ~1.7 mm diameter 1/e^2. Probe was still about 2.6 mm. Bigger than the coupling beam but it makes alignment much easier, and there is plenty of power to use for the Probe.
--Jmiles2 (talk) 15:21, 9 January 2015 (CST)
We are trying the experiment with different m levels. Part of the hyperfine repumper is used as a pump to get most of the atoms into the m=0 F=1 level. This way we can more easily model what will happen with the EIT pulses.
I'm also trying the experiment with higher intensity lasers. My model shows less oscillations in the populations of the three levels over time, and therefore the system reaches a steady state faster
Some data taken today shows that there is a difference with EIT between the two coupling beams. I want took transfer data from 1 of the coupling beams. Then I put the beam into the other fiber launch, so it's propagating in the opposite direction. The two transfer curves did not look the same, but I didn't check the beam sizes. I want to try this again more thoroughly because this could be a big deal... I don't know why they would be different though.
I check to see if either beam was clipping the chamber, and it seems that both are missing the chamber and a large amount of the power is going out the otherside.
--Jmiles2 (talk) 15:49, 11 January 2015 (CST) I think something is wrong with the beam coming from ChB. I can move the ChC beam to the ChB fiber launch and get very similar data. I took some data on Jan 9 about this. I should look at the fiber output and make sure it's anice gaussian mode. I might switch fibers and redo timings too. The output looks fine on the camera, but it's still something to try. The ChB beam was not transferring as many atoms with EIT as the ChC beam was. the powers were the same and the fact the ChC beam did similar transfers and similar powers from each fiber launch means the beam sizes are correct.
I checked the timings of the ChB beam and it looked really good.
--Jmiles2 (talk) 15:44, 12 January 2015 (CST) Light from the ChB AOM set up transfers less atoms the the ChC set up. It's something with the AOM timings or something with the fiber.
I'm convinced that CHC transfers better, but I'm not sure why. I'm going to move the circuit that pumps everything to F=1 to the CHB AOM to see if that circuit is messing things up.
After I did this both seemed to working fine. Not sure what was happening but I did notice that the lock on the diode for the EIT lasers was not that great. It would move around a little when locked. To fix it I turned down the sensitivity of the lock in amplifier.
--Jmiles2 (talk) 21:23, 28 January 2015 (CST) Today I took some scans from 0 to 4 mW and 0 to 2 mW. A coulple of days ago I took very good data on the EIT transfer of both beams with each of there pulses. The figure has all 4 plots on it, and are very similar. I'm not getting good narrow data though. Not really sure why, I get coupling from ChB control beam into ChC fiber. I used thorlabs differential screws to better align the CHC coupling beam and the probe to the fort atoms.
--Jmiles2 (talk) 17:17, 5 February 2015 (CST)
I get a standing wave with more than 10% contrast when I optically pump to m=0. I also narrowing when I went from 2 mW to 50 mW, but not from 2 to 30. The simulations say both of these regions should get a narrow feature, but the simulation has not been matching the experiment very well. I was trying to get the optical pumping even better, but I think I will take a long scan of this narrow data, with the majority of the atoms starting in m=0, this would be a new data set.
Then I'm going to make the pulses rise times longer by changing their size through the AOM's. This should put us more in the dark state and eliminate the oscillations we see in the simulatinos for the transfer data.
--Jmiles2 (talk) 17:00, 24 February 2015 (CST)
Changed the pulse times by using a longer focal length, approximately twice the focal length of the original lenses. the rise times are closer to 50 ns, rather than 25 ns.
The FORT power decreased the other, tried to fix it but couldn't get the same amount of power out, maybe something is wrong with this Chinese laser. Got the FORT back though with similar atom number, now trying to get optical pumping to work better. I get EIT, but I'd like to do EIT closer on resonance.
Need to do standing wave closer to resonance to ensure that we're better in the dark state. This will match simulations better.
I checked the beam sizes, The probe is 1650X1500um +- 50 um Both coupling beams are about 1600X 1600 +- 50 um.
--Jmiles2 (talk) 16:10, 31 March 2015 (CDT)
I keep seeing some asymmetric stuff with the standing wave depending on which way I move the standing wave. We're moving the probe beam to be perpendicular to the coupling beams. We'll need to change the polarizations to so they are perpendicular to each other and the B field which defines our z axis.
We used to use vertically polarized coupling beams and horizontally polarized probe beam, now we'll be using vertically probe and horizonatll coupling.
--Jmiles2 (talk) 13:39, 24 April 2015 (CDT)
I've been trying to get narrower data, but no luck. I checked the beam sizes and they were all around 1700x1700 um. easily + or - 50 um though. I will got through the data I have and put some of it together to try and put together the paper.
I used 10 Amps on the coils for optical pumping, and I have proof that this does pump atoms into the m=0 level.
--Jmiles2 (talk) 14:49, 3 May 2015 (CDT)
I'm trying to make some figures in Matlab for the paper, if you go into more properties when having the figure open with "plot tools" you can manually set the size of the figure through PlotBoxAspectRatio. Useful for getting all graphs to be the same for the paper.