Two-week Update

Haven't really been keeping up with the posts over the past few weeks.

For the most part, I spent a large portion of time working on a laser locking circuit which would combine peak-locking and DAVLL schemes.

In short, peak-locking is a very accurate (can lock laser to a specific hyperfine transition), but has a very narrow capture range and can be easily disturbed so that it unlocks. DAVLL is more robust, having a larger capture range and is able to maintain the laser's frequency with a range. The drawback is that the DAVLL scheme is not the most accurate, having a tendency to drift over time.

The purpose of attempting to combine these techniques is to have the accuracy of the peak-locking scheme and the robustness of the DAVLL scheme. So far, our attempt at combining these two methods is through direct addition (with adjustable gain) by feeding the signals through a summing amplifier. I completed the circuit earlier this week but only had limited testing opportunities since John wanted me to work on the PCB layout for AOM controllers.

Front view of the locking circuit, showing the various inputs.

 Top view of the locking circuit, showing the rats-nest that is the wiring.

The results of our limited testing showed that the circuit is somewhat quirky, not quite locking the laser as well as we had hoped. Possible issues may be the selection of the components used in the circuit, especially with integrator values. As an example, if the integrator provided too much feedback to the peak-locking scheme, it could possibly produce increasingly strong oscillations due to it over-correcting and losing lock that way. Another issue may be different locking points for the peak-locking and DAVLL schemes. One signal would tell the laser to correct to one particular frequency while the other signal may push the laser towards another frequency.


See Diode Laser Frequency Stabalization for a better explanation.

We also presented our mid-term talks this past Friday. Didn't quite go as well (for me) as I had expected. The "prompt" was to convey what we've been doing (not necessarily the overall objective), so I skimmed over a lot of details to focus on what I was doing with the laser locking circuit. Perhaps, I skimmed over too much when I compare mine to what others have done. I posted my slides below.

As a side note, I am using KiCad to layout the PCBs. It's a pretty neat open-source software package which allows you to design and layout circuits. I've messed a bit with the TinyCad-FreePCB combination, but I decided to work with KiCad instead because I feel that it more smoothly incorporates schematic designs (well, at least makes it easier) with one complete package compared to crossing over different programs.

KiCad Wiki
KiCad Step-by-Step Tutorial
KiCad Libraries

The finished saturated absorption laser spectroscopy setup.

Red = pump beam.
Orange/Yellow = Probe beams.

Second Week

Nothing much changed throughout the second week, did more of the same. On Tuesday, I called Isomet and Crystal Technology to obtain more information and price quotes for several models of AOMs for John Lorenz.

Most of my Wednesday was spent in laser safety training to become an unrestricted user for class 3B and 4 lasers (Laser Safety). Most of the lasers I'd be using in the lab are up to a few hundred milliwatts of power, but it was probably a good idea to obtain the training anyways. Most of the information was commonsense, but there was some interesting information, such as being able to injure the eye through expanding gases.

The rest of my week was spent on learning more about saturated absorption. John spend a fair amount of time explaining what we would be expecting to see in the saturated absorption, such as the orientation of the peaks for both hyperfine transitions and crossover resonances. In short, the laser frequency the atoms see are due to Doppler shifts caused by the atoms moving either towards or away from the laser, leading to Doppler broadening. If I remember correctly, if the Doppler broadened transitions are wider than the hyperfine transitions, we should expect to see crossover resonances because the atoms in the velocity of the atoms causes them to see different light frequencies and to become excited to a state not necessarily tuned to by the frequency of the laser. This causes the atoms to be absorb (reduced transmitted laser intensity) light of certain frequencies and not others depending on the state population of the atoms. Pumping atoms out of one state causes an increased transmission of light corresponding to that transition. If the width of the Doppler broadened transition is smaller than the frequency between hyperfine transitions, no crossover resonances should occur. This is because the atoms are not moving fast enough to produce enough of a Doppler shift to become excited into a different state than the one the laser is tuned to.

On Friday, I began setting up a layout for a saturated absorption laser spectroscopy. It should be a relatively simple setup on the optics table, but the space restrictions made it difficult to separate the probe beam from the pump beam. After a bit of fiddling, the setup is mostly complete with only a few adjustments to be made on Monday.