Vacuum is 2.5e-7 Pa (yesterday was 2.4e-7).
TORA signal is greater than around ### mV at injection.
FAB signal is large and requires 1 V/div at injection.
We will start with coasting beam lifetime.
Coasting beam lifetime will be acquired by recapturing and measuring intensity. The primary signal-of-interest is bunched beam before and after coasting. The FAB has strong harmonic-7 signal, so the TORA has been connected to Channel 2 of the scope. Because we have three active channels, the sample rate has reduced to 250 MSa/s. The FAB has an 80 MHz low-pass filter, so the sample rate is technically OK, although we expect increased noise. The TORA does not have a physical low-pass filter, so the scope channel was set to 200 MHz bandwidth and some aliasing is expected on C2. Since the amplifier bandwidth is 80 MHz the bunched-beam signal is still expected to be well resolved, the aliasing is primarily expected to add some noise; this is the best we can do as we don't have another suitable low-pass filter.
No Schottky data acquired because the coast is so short.
Delay 1 is 25 ms
Delay 2 is 50 ms
Delay 1 is 25 ms
Delay 2 is 72 ms
Delay 3 is 107 ms
Delay 1 is 25 ms
Delay 2 is 72 ms (47 + 25)
Delay 3 is 120 ms (47 + 100 - 25 - 2)
Delay 4 is 157 (47 + 100 + [25 -15])
Delay 1 is 25 ms
Delay 2 is 72 ms (47 + 25)
Delay 3 is 220 ms (47 + 200 - 25 - 2)
Delay 4 is 257 ms (47 + 200 + [25 - 15])
Delay 1 is 25 ms
Delay 2 is 72 ms (47 + 25)
Delay 3 is 320 ms (47 + 300 - 25 - 2)
Delay 4 is 357 ms (47 + 300 + [25 - 15])
We have increased the scope window to 100 ms (and necessarily reduced sample rate to 125 MSa/s) to get the bunched beam signal in fewer acquisitions.
Delay 1 is 25 ms
Delay 2 is 72 ms (47 + 25)
Delay 3 is 320 ms (47 + 300 - 25 - 2)
Delay 4 is 357 ms (47 + 300 + [25 - 15])
Scope moved back to 500 MSa/s, FAB and RF only, 50 ms window.
In all the two-beam AWGScripts we rapidly increase the RF voltage from 0 to 4 kV. This is shown in the adjacent example AWGScript, from the change in voltage around 33 ms.
We noticed that the noise signal on the TORA around this time the noise significantly overshoots and oscillates at a frequency other than the RF frequency. This then quickly dies down before giving a steady noise signal at the RF frequency. We wondered whether this large "spike" might be a transient signal on the cavity, and if so, whether it might be disturbing Beam-1.
To remove it, we made a new AWGScript with a slow ramp on the RF before Beam-2 arrives, such that the voltage when Beam-2 is injected is the same amplitude as it would have been for the sudden switch-on. The AWGScript had a 6.96 ms acceleration time to make sure that the two beams are separated.
It's only one-beam, so we can't make a comparison, but the initial result looks closer to the usual beam-1 than the usual beam-2. A two-beam case will be made to check. We will continue with the cases in the meantime.
Given the very limited time remaining to acquire data, we have decided to forego taking case-3.
Using the Welch method to compare data from the 22nd and 24th shows that dP/P on the two days was very similar. The slow RF switch-on does not appear to have improved the condition of beam-1.