Hysteresis Compensation

Dipole measurement campaign 1

6 min read

Measurement planning

Date: 2023-10-31 Place: CCC Time: 15:00 - 20:00 (best case) Depending on how long the magnet patrol is, and if RP patrol at 18:00 is, the time for our measurements might be significantly less, if at all. This did not happen.

Sensors

We will only have access to the B-Train:

  • Measured I
  • Measured B

Cycles to play

Supercycles

We have asked for 5 supercycles: All supercycles are STANDALONE, and prefixed by HYSTEST in the BCD editor

PRECYCLE supercycle: LHCPILOT - LHCPILOT - LHCPILOT - LHCPILOT - LHCPILOT

LHCFILL supercycle: LHCPILOT - SFTPRO - ZERO

  • We need the ZERO to get the down.

FT supercycle: SFTPRO - ZERO - ZERO - ZERO - SFTPRO - ZERO - ZERO - ZERO

  • We need the ZEROs to get the down.

LHCFILL + MD1 supercycle: LHCPILOT - MD1 - SFTPRO - ZERO

  • Closest to operational supercycle.

FT + MD1 supercycle: SFTPRO - MD1 - SFTPRO - MD1

  • Closest to operational supercycle.

FUNKY supercycle: LHCPILOT - SFTION - AWAKE - HIRADMAT - SFTPRO

FUNKY + MD1: LHCPILOT - MD1 - SFTION - AWAKE - HIRADMAT - MD1 - SFTPRO

Optional supercycles

These cycles are not approved by MSC or EPC to play. We keep them as standby.

STAIRCASE supercycle:

  • Staircase on repeat. Cycle name: XXX HYSTEST supercycle: 50GeV - 100GeV - 200GeV - 300 GeV - 400 GeV
  • Cycles 6000 ms - 8400 ms with flat flat top and flat bottom.
  • is satisfied.

Standard cycles

For LHC → SFT cycles we need to add a zero at the end of the supercycle to make I RMS low enough.

Staircase cycle

Stephane has programmed a cycle in LSA:

Not possible to make full staircase cycle with high granularity as the I RMS would be too high, and cycles would be limited to 90s. In this case the steps are 400 A each, however while IRMS is below 3250A, EPC and MPC are hesitant to play it. See below.

Meeting with EPC and MPC

Staircase cycle over I RMS not possible, thermal and electrical problems. We can investigate the possibility of running it, but it will take more than one week to investigate the hardware limitations.

We can play the staircase if we rescale the Y-axis so that the . This will not provoke pronounced hysteresis effects, but we can still use it to map some minor hysteresis loops.

Notes

  • When can we do magnet tests, exactly
    • Michael checks
  • Who is coordinating during the measurement week
    • Michael checks
  • Can we do cycles with flat flat bottom and top?
    • Ask magnet guys

Ask about pre-cycle supercycle feasibility

MD plan

Discuss with Vincenzo

Preparations

Ensure that the marker is correct on the following cycles:

  • LHCPILOT
  • SFTPRO
  • STAIRCASE
  • HYSTEST
  • MD1
  • ZERO
  • AWAKE
  • HIRADMAT
  • SFTION

Ensure that the proposed supercycles are ready

Cycles to play

Preliminary ideas:

  1. Pre-cycle
  2. FT (x5-10)
  3. LHCFILL (x5-10)
  4. FT (x5-10)
  5. Repeat
  6. Pre-cycle
  7. FT + MD1
  8. LHCFILL + MD1
  9. FT + MD1
  10. Repeat
  11. Mess up hysteresis loop with HYSTEST or RANDOM
  12. Repeat above
  13. Play STAIRCASE in between.

Execution

MD planned last minute for Thursday 2 November 2023 10:00 - 12:00 after EPC HV test.

Tests executed:

LHCFILL - PHYS

This cycle was a mistake 2x PRECYCLE 5x LHCFILL 5x PHYS

PHYS - LHCFILL - PHYS long

This test to see what happens if we switch from PHYS supercycle to LHCFILL and back, with relatively many repetitions of each supercycle (total 10 SFT before and after, respectively). This should allow time for hysteresis effects to settle (even after some 3 cycles it should be fine) 2x PRECYCLE 5x FT 5x LHCFILL 5x FT

PHYS - LHCFILL - PHYS short

This test to see what happens if we switch from PHYS supercycle to LHCFILL, and back, but with few cycles of each. This is to see how long it takes for the hysteresis to settle. 2x PRECYCLE 2x FT 2x LHCFILL 2x FT

PHYS - LHCFILL - PHYS very short

Same as earlier, but one of each cycle. Ideally we should have played more FT cycles after the fill, but this will have to do. 2x PRECYCLE 1x FT 1x LHCFILL 1x FT

PHYS - LHCFILL - PHYS w/ MD1 long

Same as before, but with MD1 before SFT, as during operations. For comparison against previous measurements, to see what happens to the dipoles when we remove MD1 from the supercycle. 2x PRECYCLE 5x FT + MD1 5x LHCFILL 5x FT + MD1

PHYS - LHCFILL - PHYS w/ MD1 short

Same as earlier, but with MD1 2x PRECYCLE 2x FT + MD1 2x LHCFILL 2x FT + MD1

PHYS - LHCFILL - PHYS w/ MD1 very short

Same as earlier, but with MD1. 2x PRECYCLE 1x FT + MD1 1x LHCFILL

FUNKY supercycles

Here we aim to mess up the hysteresis loops so we get a large variety of hysteresis loops to model.

We:

  • Played our FUNKY supercycle with and without MD1
  • Switched to PHYS and LHCFILL supercycles in and out
  • Went in and out of dynamic economy when pulsing FUNKY + MD1, for all physics cycles
  • On the fly changing the order of the cycles in FUNKY/+MD1
  • Adding more AWAKE (making 3x in sequence), or interleaving AWAKE with HIRADMT (3x), and switching one to DYNECO
  • A clone of FUNKY + MD1 was added to make it easier to switch out cycles or change the order.

PHYS - LHCFILL - PHYS long 2

Repeat experiments from previous measurements, playing supercycles (without MD1) 5x FT 5x LHCFILL 5x FT

to investigate the effects after messing up hystheresis, without precycling.

PHYS - LHCFILL - PHYS long 3

Repeat above, but with pre-cycling, to investigate pre-cycling wipes out memory to the first cycles played.

2x PRECYCLE 5x FT 5x LHCFILL 5x FT

This measurement did not make it past the first PRECYCLE.

MD finished early at 11:53 as an access at BA6 accidentally triggered the IMAINS to trip. Since we had access to 12:00 sharp, we cancelled final measurements at 11:59.

All data is confirmed to have been logged to NXCALS by the B-Train (yay!).

Graph View

Backlinks