EPR Polarization Measurement - AFP Sweep

This configuration uses the lock-in amplifier with a PI feedback box to lock into the EPR resonance frequency and then track its behavior. With AFP spin-flip, a shift in EPR resonance frequency will be measured, which is proportional to the pumping cell polarization.

Construct the circuit described in figure :

1. First construct FM Sweep circuit described above.

2. Connect Lock-in amplifier channel 1 output (from rear panel) to PI circuit 'in'.

3. Switch the 100:1 attenuator from ``FMS (no attenuation)'' to ``AFP (100:1)''.
4. Set Modulation source DS345 parameters:
   Frequency: 200 Hz (This frequency is arbitrary. 200Hz is good since it is far away from most noise sources.)
   Waveform: sine wave
5. Set Wavetek 80 parameters:
   Amplitude: 8.0 V, higher amplitude will be needed if the signal extracted by FM Sweep is less than 25 $\mu$V;
   

6. Bypass the RF capacitor by setting the switch to 'EPR' on RF capacitor box front panel.

Figure: Circuit for EPR measurement with AFP spin flip

Now that the circuit is built, you can begin the initialization of the setup.

1. Set the modulation source DS345 amplitude to approximately 0.6 Vpp. This number can be optimized for different conditions such as different target polarization and EPR D2 signal amplitude. At an EPR D2 signal amplitude of 25 $\mu$V the optimized DS345 amplitude is roughly 0.05 Vpp for P$< 20\%$, between 0.05 and 0.6 Vpp for $20\%<$P$<40\%$, and 0.6 Vpp or higher for P$> 40\%$.

2. Find the EPR resonance frequency:
   

   If you have done a FM Sweep measurement, then you already know the EPR resonance frequency, continue to the next step.
   

   If you don't know where the resonance frequency is, you may want to do a FM sweep measurement, or you can find the current resonance frequency by hand, To do this,

   • Temporarily remove the connection to Vin on the PI feedback box, turn off the integration function, set the 100:1 attenuator switch to ``FMS (short)''. Now you are back to the FM Sweep circuit setup. Manually adjust the Wavetek 80 frequency with the smallest step 0.01 MHz, while observing the lock-in. Look in the area given by the formula:

     EPR base resonance = (466000 * holding field in Gauss)

     This is typically about 11MHz.

   • You should be able to observe the lock-in signal changing when adjusting the frequency. It should show a maximum ($>$0, peak) followed by a minimum ($<$0, dip), or vice versa, and the resonance is between them when the lock-in signal is small. Keep in mind that the lock-in will read zero when you are far from resonance as well, so be sure to look for this characteristic resonance behavior.

   • Reconnect the Vin input, and set the 100:1 attenuator switch from ``FMS (short)'' back to ``AFP (100:1)''.

3. Set the Wavetek 80 frequency to be 0.1 MHz lower than the current resonance frequency if the light is right-handed polarized, or 0.15 MHz lower than the current resonance frequency if the light is left-handed polarized. The lock-in signal should be large since you are not at the resonance frequency.

4. Turn on the integration on the PI feedback box.

5. Observe the Vout of PI circuit on an oscilloscope. If it is ``jumping around'' erratically, or looks as if the Opamps have been saturated, you must adjust the gains. The best advice is to begin with the relative gain, and if this does not rectify the problem, move on to the absolute gain. Basically, adjust these parameters such that the feedback signal is not too strong, neither too weak.
   

6. Once this is done, the circuit should track the resonance frequency and the counter reading should change towards it. The lock-in signal should be stablized to minimum.

7. Wait until the counter reading is stablized, change Wavetek frequency manually by 0.01 MHz. You can observe a jump of 0.01 MHz in counter reading, but then it should change back to the resonance frequency, which means the circuit is following the resonance. If not, then the frequency it was stablized before is not the true resonance frequency, check the circuit and try again.

8. Once this is done, the circuit should take care of itself. Left alone, it will track the movement of the resonance peak.

To measure the shift in resonance frequency due to AFP spin-flip, open the LabView program AFPSweep.vi. Enable the program by clicking on the large white arrow in the upper left corner of the panel.

1. Set the sweeping time to be 6 seconds and the waiting time to be 20 seconds. Set the number of sweeps to be 2.
2. Run the program.
3. Download parameters by clicking the corresponding button, you should read ``downloading parameters'' on the program status dialog.
4. Wait until the program status shows ``waiting for data taking to start''.
5. Start data taking by clicking the ``pause :: run'' button. Now the program should read resonance frequency from counter, and the lock-in amplifier signal. The program status should show ``waiting for sweep trigger''.
6. Wait for roughly 20 seconds, start sweeping by clicking on the ``start sweeping'' button. The program status should show ``sweeping...''.
7. You should then be able to see the jump in resonance frequency. When the light is right-handed polarized, this should be a positive jump, while for left-handed light it is negative.
8. After the sweeping is completed, wait until the program status shows ``waiting for sweep trigger'', now you can stop the program.

9. Disconnect the output of Wavetek 80 to the cable leading to RF coil.

10. Never stop the program during sweeping, this will cause the $^3$He to stay at the wrong state and hence a big loss in target polarization.