Dedicated MD 2025-05-21

BCDs requested:

• Standard NA supercycle (2x or 3x SFTPRO1 + MD1)
• Standard NA + LHC1 supercycle (SFTPRO1 - ZERO - LHC1 - MD1)
• Standard NA + LHC1 (spare) supercycle as an explicit BCD (SFTPRO1 - ZERO - LHCINDIV Q20 1 inj (-ZEROxN) - MD1)
• Standard NA + LHCPILOT (SFTPRO1 - ZERO - LHCPILOT - MD1)
• Standard NA+AWAKE supercycle
• Standard NA+HiRadMat1 supercycle
• Standard NA+HiRadMat2 supercycle

MD plans

• Hysteresis compensation at flat bottom without MD1
• Eddy current compensation together with hysteresis compensation
• Q26 vs Q20 LHC cycle
• Eddy current compensation on LHC cycle flat bottom
• Tune offset studies at FT

Dedicated MD 2025-04-02

Cycles to play

5x SFT 4x AWAKE 5x SFT 4x HIRADMT1 6x HIRADMT2 3x SFT 5x LHC1 5x LHC1 spare 4x SFT

Then go for LHCPILOT LHCINDIV (Q26)

MD summary

We keep 3x SFT

Perturbation caused by missing MD1

Predictions show “ramping fields”

We apply 2e-5 T correction, but lose the beam after a few hundred turns. Turns out we hit the 0.66 tune.

We find to

MD analysis

We find that analyzing the following cycles:

Cycle 08 2025-05-21 17:45:04.935000: SPS.USER.MD1 MD_26_L60_Q20_2022_V1
Cycle 09 2025-05-21 17:45:08.535000: SPS.USER.SFTPRO1 SFT_PRO_MTE_East_Extraction_L4780_2025_V1
Cycle 10 2025-05-21 17:45:19.335000: SPS.USER.MD1 MD_26_L60_Q20_2022_V1
Cycle 11 2025-05-21 17:45:22.935000: SPS.USER.SFTPRO1 SFT_PRO_MTE_East_Extraction_L4780_2025_V1
Cycle 12 2025-05-21 17:45:33.735000: SPS.USER.MD4 MD_26_L60_Q20_2022_V1_Clone_MD_to_flat
Cycle 13 2025-05-21 17:45:37.335000: SPS.USER.SFTPRO1 SFT_PRO_MTE_East_Extraction_L4780_2025_V1

And designating the first SFT cycle as SFT1, and the third (after MD4) as SFT2

The measured current shows no difference between the two cycles down to noise - both cycles were uncompensated.

Plotting the difference in magnetic fields (raw B-Train, field predictions, eddy current from b-train or beam, and raw-btrain with eddy currents subtracted), we see:

We see the $\Delta B$ caused by eddy currents (as fitted on the beam) is 0.6 G at injection (C200) (dark red line). The $\Delta B$ as predicted is 0.5 G at the injection plateau (orange line). However while the $\Delta B$ as seen on the B-Train is 0.3 G at injection (navy line), it decreases to 0 - the b-train seems to only catch the eddy current decays, and shows no hysteresis (as seen on the red line).

However on the radial data we see a maximum $\Delta r$ at 9.7 mm (7.2mm + 1.5 mm), equivalent to a field difference of 3.07 G.

Whereas considering only the first trajectory (and taking the average of the non-masked BPMs), the SFT1 reads a ${\bar{r}}_1=-0.81\text{mm}$ and SFT2 reads ${\bar{r}}_2=4.46\text{mm}$, which gives a total $\Delta r=5.27\text{mm}$. With ${\gamma }_T=23.225$, $\rho =1100\text{m}$ and $B_{\text{ref}}=0.063\text{T}$ we get an equivalent $\Delta B=1.63\text{G}$, which is $0.4\text{G}$ more than as predicted avobe (purple), still an equivalent 1.3mm $\Delta r$.

If we consider the $\Delta B$ in $I$/$\Delta B$ phase-space:

We see that initially at low current, the $\Delta B$ between the two measurements are close to zero, and as the current ramps, we see $\Delta B\approx -7.5e-5$ T change up to 1200 A, peaking at $\Delta B\approx -1.7e-4$ T, and then stabilizing at $\Delta B\approx -1e-4$ T.

Comparison to BC MD 2025-03-18

In the MD during beam commissioning, we measured SFTPRO after an LHC cycle, and had 0 to 2 BPs between the LHC and SFTPRO cycle. Based on the first turn data in only the first sextant, the radial position goes -0.29 mm (0 BP) → 2.17 mm (1 BP) → 2.45 mm (2 BP), so largest $\Delta r$ was 2.74 mm, incompatible with what is seen above, as we asymptotically would realistically be around 3 mm.