(on the fly)
some pieces of information on the initial difficulties to steer the beam through the pt 8 injection yesterday …
The vertical bending group upstream of the TED (5*B280, converter RBIV87715) had jumped in current by 4 % (!), from the correct value of 719 to 748 A, apparently when the energy of TI 8 was changed by ~ 1 per mil to match the SPS (~ 17:50, see item 12 in http://elogbook.cern.ch/eLogbook/eLogbook.jsp?shiftId=1014305). To note that the other magnets in the line changed their currents by the correct amount. Joerg seems to understand this; he had tested a new calibration curve for this magnet.
The polarity of the 3 vertical correctors MVIAV88115/6/7 (used as bends, i.e. with fixed, non-zero value) was wrong, while it was reported correct when it was physically checked (fortunately these magnets are supplied by the normal bipolar corrector converters, so no need to re-cable them). Rossano will re-check if the polarity was reversed at that moment, or whether there were some changes in the control system since (different treatment between bends and correctors ??). In the course of the afternoon we seem to have spotted a wrong polarity in the MAD strength file – to be investigated further.
The vertical read-out range of the BTVST (screen in front of TDI) is set up too low. While the screen goes to +50 mm we could only see from -30 to +39 mm, which means we just missed the beam off the top (nominal 37 mm). BI can re-adjust it with an access; it should be done in pt 2 as well. We should have spotted that before.
One pickup at the end of TI 8 showed dodgy readings, unfortunately just where and when we needed some steering. This was likely due of the beam losses before; the effect seems to have disappeared since.
Malika and Ilya have done sterling work fitting back errors on all quads, from R8 back into TI8 (and shall I just say it's lucky Ilya didn't have a PS Booster sequence to hand.....).
Result: the TI8 MQIF876 looks like a good candidate - a 9% strength increase fits the observed DX very well, and does not screw the response up (unlike some of the others) - Rogelio had a look at the data from last night. Other candidates like Q4,Q5, triplets didn't look anything like as good....even rotated and scaled by large factors. No beam yet to test the fix - we plan to first measure dispersion using the Malika version without doing anything on 876 (this matches for the new initial conditions), and then we will trim 876 down by 9% and remeasure dispersion.
When you see this you will probably know that we found a colossal beating of the hor. dispersion... probably originating in the last part of TI8.
But this is about something else...
- In case we go for 23, call me before to reload the settings (in case I'm not around). I could not because the converters are in the wrong state.
- In case we have to access the tunnel (or TI8), have a look at the logbook (LHC or SPS) or on
under RBI.816 (left side menu) for instructions on how to stop the switching of the RBI.816 converter to TI8.
Hi folks, PO has measured the voltage on the DCCTs of the MQIF87200 and MQIF87600 with a precision multimeter, in "DC-like" conditions (in fact a cycle with long flat top). For 300 A requested in the trim editor they measure exactly 5 V (600 A DCCTs). They propose to re-measure tomorrow the real currents on the flat top of the real LHC cycle, to exclude dynamic effects. Normally one would expect this to be tendentially on the lower side, meaning that the supply cannot reach to demanded level (fast enough); more difficult to understand that the current should be too high ...
In the meantime the magnet team is measuring all MQIF in the matching region (R, L, ...), including disconnection/reconnection. Coordination via Rossano. Cheers, Volker.
Nothing “squiffy” in terms of tilt has been found at today’s inspection of the end of TI 8.
All quads between MQIF87600 and MQID88100 are tilted to the right seen in direction of the beam, with values in agreement with the survey data base, within the precision of the simple measurement method applied (about +/- 3 mrad). The value at the end is as expected around 65 mrad, i.e. 52 mrad wrt to the LHC plane (see LHC Project Report 719, http://cdsweb.cern.ch/record/733786/files/lhc-project-report-719.pdf). The magnets MQIF87400/04 and MQID87500 were not measured but just looked at and were also found tilted in the prescribed sense.
A few simple things were superficially “checked” on the fly, for instance that the doublet MQIF87400/04 has the same polarity in both magnets (in fact, the same tape colour on the same sided connector).
A PO team will measure this afternoon with greater precision the current flowing towards the magnets MQIF87600 and MQIF87200, in DC mode, for a given setting, in complement to the simpler measurement made yesterday.
Thanks to the RP and SU teams and HWC to have supported this check at short notice.