From: Serguei Zelepukin Sent: 13 June 2010 16:15 To: Ken Bell Cc: Serguei Zelepukin; Dave Cockerill; David Barney; Diogo Raphael Da Silva Di Calafiori; Wolfgang Funk; David Anthony Petyt; Nicolo Cartiglia; Matteo Marone Subject: RE: FW: stability plots Attachments: PTHM_EE_mappings.pdf Dear all, pls find EE probe mappings in the attachment a few comments: 1) backplate PTM probes (4x) are all mounted on the backplate but very close to cooling pipes (~1-2 sm) -- so they can be affected by cooling... (each probe monitors a different cooling circuit) However, from EE plots one can see that backplates have rather big thermal inertion... (while EB SM grids do not) That means PTM backplate probes show backplate temperatures... and not those of cooling pipes 2) As I remember Dee's backplate has no thermo-isolating layer from the side of electronics (EB SM has an armaflex layer) Also, cooling pipes on the backplate are much less "powerful" than SM cooling channels inside the grid -- the cooling channel density in the grid is much much higher than the one on the backplate 3) PTM probes on the env.screens show a noticeable change when LV is switched on/off.. which one would not expect to see I suspect that env.screen cooling pipes are somehow "connected" to output pipes of cooling blocks -- maybe they are routed very closely to each other somewhere outside Dee's... 4) EE PTM probe calibration: -- backplate probes have the same relative accuracy as EB probes (<0.01C) and tolerance of 0.03C in the range [15C - 20C]. -- Env.screen and CB probes are partly those which were sorted out in the EB probe "calibration" in the lab -- some of them may have relative accuracy ~0.02C and tolerance ~0.08C in the range [15C - 20C]. Some others may have relative accuracy ~0.05C and tolerance ~0.2C in the range [0C - 70C]. If you need more details I can look into the installation log files to find exact info about each probe. Terms explanation: Relative accuracy means that of measuring a T change. Tolerance (a term from manufacturer): gives a max difference between any two probes at a any absolute T point in the range [0C - 70C]. Also the factory calibration has some (unknown but hopefully very small) offset from the real absolute T in calibrations points... Thus, all probes have some individual constant offsets well inside the tolerance window... Manufacturer vs lab calibration: manufacturer individually calibrated all probes and provided calibration coefficients for each one. (Non-calibrated probes have guaranteed tolerance of 0.2C) Lab "calibration" at CERN was just a series of measurements and sorting probes by their actual tolerance (mostly <0.03C in the range [15C - 20C]). Probes with bigger tolerance than the above are installed in EE CBs and env.screen while backplate probes are of the same "quality" as EB ones. Hope this will be helpful... Cheers, Serguei On Sat, 12 Jun 2010, Ken Bell wrote: > Thanks, Serguei, > > Plots extremely interesting.... > ...extremely reassuring wrt EB... > ...& extremely depressing wrt EE... > > I've forwarded iyour mail to PFG (maybe one can't email it if not on the email list oneself?). > > I need to think a bit, & also investigate what happened ~3 weeks ago to EE > when LV+Cooling went off for ~6 hours... > > ...OK, couldn't resist looking, instead of watching the football... > OK, times in UTC, on May 13th. > 10:48 Cooling PLC shut off cooling. ==> HV/LV powered down. > (from memory, ES should have stayed Powered & Cooled) > D2#19 = "backplate temperature" rises steadily from 18.7 to 19.2 until 14:00. > At that point Diogo got Xavier to restart the cooling, to stabilise things, even tho' Xavier was still working. > ==> temp drops to 18.5 at 17:00, which is presumably when LV was powered back up, since afterwards > Temp slowly crawls back to 18.7. > PS (or EPS?) file attached... > ... > > OK, so when LV goes off in EE, backplate sensor: > * Cooling stays ON => sensor cools down by 0.7'C over 8 hours (14-Apr) > * Cooling goes OFF => sensor warms up by 0.5'C over 3.5 hours (13-May) > ==> > Conclusion: EE "Backplate" sensor more stable if cooling STAYS ON ! > > So much for my hypothesis/hope... > Well done Serguei ! You were right ! > > OK. > Thanks a bunch - bon weekend - Ken. > > PS Serguei, remember, UXC is supposed to be both temperature & humidity (dewpoint=< 10'C?) stabilised. > "Plug" can't be opened, for radiation safety reasons, when beam circulating... > Not sure whether it could be opened during 4-day Tech Stop...Perhaps, but unlikely to be neeed... > > PPS Serguei: Is there a good map somewhere, for EE, between numbering of sensors in WBM > & their physical locations ???? > (and eg is the "backplate sensor" on the BACKPLATE, or on the COOLING PIPE on the backplate ???) > > Ah, and what's the ABSOLUTE calibration accuracy for EE sensors - if I remember, > you didn't yourself calibrate the EE Pt100s ??? > > Dave_B, > Can (aka shouldn't!) the ES sensors likewise be accessible via WBM ??? > Aha ! Just discovered they fully ARE ! Time to shut up ! > K. > > ________________________________ > > From: Serguei Zelepukin > Sent: Sat 12-Jun-10 01:53 > To: Ken Bell > Subject: Re: FW: stability plots - mail delivery problem > > > > > Hi Ken, > > I've got a message on delivery is failed to: > > > Delivery has failed to these recipients or groups: > > cms-ecal-commissioning-prompt-analysis (Prompt analysis of CMS ECAL > commissioning Data) > (cms-ecal-commissioning-prompt-analysis@cern.ch) > Your message can't be delivered because delivery to this address is restricted. > > > It's up to you to decide whether it's really needed... > > Cheers, > bon weekend, Serguei > > > --