X23A2 Beamline Evaluation: Detector Noise

This is a pretty simple test -- just a series of time scans in which the detectors are polled repeated for 3 minutes while dwelling under fixed conditions. To make the latter parts of this test, I borrowed a 50 micron tungsten foil from the EXAFS equipment pool. At 10 keV, the energy of this test, this foil is about 9.4 absorption lengths. The ion chambers were filled with nitrogen and the harmonic rejection mirror was in place at an angle of 0.2 deg.

The W foil was affixed to the edge of the sample stage with masking tape in such an orientation that it would be simple to block the beam with either the horizontal or vertical edge simply by moving the sample stage. This means that these noise tests are made by comparing the I_t and I_r chambers rather than I₀ and I_t. All three chambers are of the same type and have been in service for the same period of time. Of course, the I₀ chamber is 10 cm shorter than the other two, but I assert this test is applicable to any measurement at our beamline. It was much less work to set up the test this way than to move ion chambers around.

Each of these scans shows the ratio of the signals on the I_t and I_r chambers as a function of dwell time.

First test, 10 keV, full beam

• Here we have an rms noise of about 0.0001 on top of a signal of 0.646. This is noise of about 1.5 parts in 10,000.
  

Second test, no beam

• Here we see electronic noise.
  

Third test, 10 keV, beam half blocked by a horizontal knife edge

• Here the rms noise is about 2.6 parts in 10,000.
  

Fourth test, 10 keV, beam half blocked by a vertical knife edge

• Here the rms noise is about 2.6 parts in 10,000.
  

The noise is slightly higher in the half-blocked scans. I suspect that the culprit is the silly job I did mounting the foil. The mechanical instability (floppiness) of the foil is probably the cause of the higher noise level in those two scans. In any case, the noise level in our ion chambers and signal chain is excellent.