# Background

The Hall A Compton polarimeter uses Compton scattering between the electron beam and polarized photons confined within a Fabry-Perot cavity to monitor the polarization of the electron beam. The Compton asymmetry between events where the electron and photon polarizations are parallel and antiparallel is proportional to the beam polarization $P_e$:

                  $A_{exp} = \frac{S^+ - S^-}{S^+ + S^-} = \langle A_l \rangle P_{\gamma} P_e$


The Compton cross-section is low enough that this polarization measurement can be taken simultaneously with a running experiment (e.g. d2n) downstream. Ideally, we can detect both scattered electrons and scattered photons in coincidence.

During d2n running, active Compton polarimetry was limited to the scattered-photon side (no electron polarimetry data are available). The laser feeding the Fabry-Perot cavity (and hence providing photons for Compton scattering) has a wavelength of 1064 nm and the power in the cavity was in the 400 W range throughout the experiment, requiring periodic re-tuning over time. The cavity runs in a cycle:

• Laser on, cavity locked, photons right-circularly polarized (~ 90 seconds)
• Laser off for background measurements and polarization switch (~ 30 seconds)
• Laser on, cavity locked, photons left-circularly polarized (~ 90 seconds)
• Laser off for background measurements and polarization switch (~30 seconds)

During d2n running, we had two separate DAQs running:

• Original (Saclay) DAQ, computing asymmetries based on counting rates
• New (CMU) FADC DAQ, computing asymmetries in energy-weighted, integrated signal

Since these two DAQs are measuring slightly different asymmetries, using two different methods with two different analyzing powers, we do not expect them to yield identical asymmetry results. However, both asymmetry results track the same beam polarization, so (for example) one asymmetry should increase at the same time the other does.

The photon detector was a GSO crystal, 6 cm in diameter by 15 cm in length, which was installed in the hall in December.

# Hardware Changes During d2n

• 02/08: CMU DAQ was running at 15 Hz. Now running at 30 Hz (triggering off MPS instead of pairsynch): HALOG entry
• 02/11: 0.3x attenuator added to photon signal path to Saclay DAQ (since it expects a much smaller signal than the CMU DAQ does) HALOG entry
• 02/15: New scalers added to CMU crate (MPS and trigger rate): HALOG entry
• 02/15: Trigger threshold on CMU crate raised from -30 mV to -50 mV: HALOG entry
• 02/23: New photon detector PMT/Base swapped in (still GSO crystal): HALOG entry
• 02/23: Shortly afterward, busy bit on CMU-DAQ FADC quits working. CRL modified: HALOG entry

# Calibration

## Detector Response Function

The relation between the Compton asymmetry $A_{exp}$ and the electron beam polarization $P_e$ depends in part on the analyzing power $A_l$. Understanding how the detector responds to light is necessary to understanding $A_l$, and hence to arriving at a polarization number from a Compton asymmetry.

To understand the GSO detector response function, we rely on GEANT simulations and results from tests undertaken at HIGS.

## Detector Linearity

Apart from the response of the GSO detector, the PMT/base/GSO combination may have additional nonlinearities. CMU grad student Megan Friend has worked on mapping these out.

## Identification of Cavity States

Correct identification of laser cavity states is crucial for computing Compton asymmetries, since Compton events only occur during cavity-on states and we use the cavity-off states to perform background corrections. A careful separation between events with left-circularly polarized photons and those with right-circularly polarized photons is also essential. Misidentifications will introduce systematic errors and dilute the asymmetries.

Cavity-state identification relies on signals from EPICS variables and real-time bits.

We recently discovered that the real-time bit meant to indicate whether the laser cavity is on or off is unreliable. This may be true only of the copy sent to the CMU DAQ, or it may also affect the original DAQ -- we don't know yet.

See main article for further details.

# Cross-Check with Moller Polarimeter

The Moller polarimeter serves as a (destructive) measurement of beam polarization independent from the Compton. By comparing the Moller and Compton results, we can get an empirical calibration constant for mapping from a Compton asymmetry to a beam polarization. This can eventually be compared with the theoretical proportionality constant (computed from $\langle A_l \rangle$, that is from QED and from the detector response function).

There are a few caveats to be considered when comparing Compton and Moller data. Moller measurements are taken at very low currents (1-2 uA or less). Since the Compton polarimeter does not work well at this current (signal is too low compared to background), we cannot take Compton and Moller measurements simultaneously. Thus:

• If the beam polarization is not sufficiently stable in time, the few hours' separation between the end of a Moller measurement and the first Compton result may make the calibration problematic.
• If the beam polarization changes with beam current, then comparing the low-current Moller measurement to the high-current Compton measurement is problematic. In this case, we must remember that the Compton measurement is taken with the production current.
Comparison of Compton asymmetries (measured by both Saclay and CMU DAQs) with beam polarization as measured by the Moller.

# References

• Talks
• HAPPEX meeting (18 April 2009; simulations, preliminary d2n results by D. Parno): [1]
• d2n phone meeting (26 May 2009): Media:Parno_26May09.pdf‎