Lamp intensity with 654nm MSE filter in LED lamp control via Lamp 1 is outboard most - MSE(fibres) ch2 fib5 - MSE11(ADC) Lamp 2 - MSE ch6, fib 3/4 - MSE6(ADC) Lamp 3 - MSE ch7, fib 2 - MSE17(ADC) Plasma angles relative to lamps (roughly) Lamp2 (central), they read -13 ±2 Lamp1: -11 Lamp3: -24 Plasma: -2 to +8 (-5-+10) Varying Oven angle: Lamp1 AUG/42 (10 - 15) Lamp2 AUG/41 (9 - 17.5) Lamp3 AUG/41 (15 - 27.5) AUG/44 Lamp3: 12 - 23 Lamp2: 9 - 29 Lamp1: 15 - 27 Hmm, one end of range is broken, where 1 component is min, and brackets are moving image 30 is good near th most anti-clockwise Lamp3: 24 Lamp2: 34 Lamp1: 31 so moved it down to nearly most clockwise pos so that they are all < 20°, the direction that puts the plasma at ~30° ref pulse 28071 - mse reference, Rout ~ 2.1 low density h-mode, beams 3 & 8 +beam 3 on during current ramp down + period of 1, 2 3 4 3,1 3,2 3,4 2,1 2,4 4,1 1,2,3 1,2,4 1,3,4 2,3,4 ardSeq can store around 5976 bytes of data, which is 498 half frames HGW 0383488xxxx Robert HGW 1205 Ralph Dux 1256 Wolfgang Jacob 2618 Guenter Haas 1459 gmds/AUG 1 - 33 Thursday 17/01 Offline in-hall test - radiation etc 36 - 47 Wednesday 23/01 Setting up on torus, lamps, calibs etc 48 - 77 Thrusday 24/01 Plasma Day 1 78 - 98 Friday 25/01 Plasma Day 2 99 - 117 Monday 28/01 Rebuild and set-up 118 - 143 Tuesday 29/01 Plasma Day 3 144 - 153 Wednesday 30/01 Relocate camera with bolts, Remount Oven but trying to maintain alignment. 154 - 181 Thursday 31/01 Plasma Day 4 (inc. IMSE ref) 182 - 187 Friday 01/02 MSE polariser calib 3 sets of calib, name CalibSet: Plas1+2, 44 -- Start - Oven rotated at 45 Pol scan, Ne lamps, Defocused 49 rads --- Plasma 1 --- 50 - 56 Camera wobble blurring fringes? 52 interesting comparison of long/short exp - w.r.t ELM dents --> camera movement 57 strapped camera down, better, but occasional dents 61 polariser at 2500 = m(~31°), but is messy near edge 73 polariser at 3000 = m(~5.9°), not messy but too near 0 62 45kV beam, and next?? 62 - 72 field/power scan 74, 76 good current ramp -------------- 77 - 78 rads --- Plasma 2 --- 81 Polariser scan on zeeman BG, stops in middle 82 Polariser at 1500, m(~4°), no beams, not much use 85 Polarisaer full rotate, good µ calib for zeeman 86 - 89 Beam blips 91 Very good reflections data (pellets) 92 - 93 Polariser scans, def. good enough for µ variation, maybe also for µ abs 94 Good reflections and zeeman ellipticity data 95 A few really good reflections images 97 Beam blips of Q2,3,4 near end -------------- 98 rads 103 Focus changed 100 Pol scan, Ne Lamps, Focused 101 Ne Lamp, Focused, moving lamp to integrate 102 103 Defocus 110 -- End - Rebuilt CalibSet: Plas3 --- Plasma 3 --- 118 All ellipt 119 Good interlaced plasma data, ellipticity is a bit wtf 122 Ok plasma data for early 123 Good plasma data during ramp 126 Messy but viable plasma data with ECCD drive on q surfaces, some movement near end 127 Good reflections data with interlaced FLC 128 129-131 Attempts at polariser movement, failed, but has small steps 132 *µ CALIB* Polariser fixed at 2660, Good, home ret -0.14° 139 Polariser at 2660 again, home ret +2 142 Plasma edge in soft-disrupt ---- 143 rads... CalibSet: Plas4 144 152 Rebuild --- Plasma 4 --- 154 160 quite near end 162 Rotating polariser, stuck at ~21° but not very useful due to low contrast in places 163 Fixed polariser, unknown loc ~29°, low constrast in places but shows background vs beam µ varies measurement by < 1° 170-177 Bad plasma data but good Refelctions from PSL tiles in places 178 Comparison pulse 180 Polariser fixed at 3100 (~ 3.8° - not much use) ---------- 181 - rads 184 - defocused lamps 185 - 187 MSE ex-Vessel wheel Plas4 moved camera so changes µ calib a lot. offset should be similar MSE ex-Vessel calibs: ** All the MSEθ here are 1/2 what they should be, since they are actually per tooth. There are 180 teeth over 360° - but MSEθ=0 is still the same, and is the start of the blocked tooth ** # Desc λ θ(MSE0) 185 Fast 45.004 21.79 186 Slow Filt 45.004 21.82 187 Slow Ne Lamp 44.998 21.81 186: MSEθ=0 at θa=21.813°, setting µ=79% global θm(µ=100) here = 18.59 Nearest 22.5 is MSEθ=0.345 Nearest img to that is 121, at MSEθ=0.133, θa=22.08 µ image has weird bottom right corner like the plasma images 187: MSEθ=0 at θa(fit using µ=79)=21.84 θm(µ=100)=18.8 θa=22.5 at MSEθ=0.331 Nearest img = 26, MSEθ=0.272, θa=22.38 ... That was based on a definition of 0 as the rising edge of the blocked groove. A.Mlynek's paper says 0 is half way through gap, which is 1.6° (actual, not halved) further around (now corrected this in script, and redone in IMSEProc:)) So: 185: MSEθ=0 at θa = 23.36(µ=100), 23.4(µ=79) on upswing 186: MSEθ=0 at θa = 23.6(µ=100), 23.4(µ=79) 186: MSEθ=0 at θa = 23.375(µ=100), 23.41(µ=79) So... 23.4 132 + 139 from CalibSet'Plas3' gives ok µ for image center for beam 3 This /should/ be the same for every calib set - we don't expect µ to change significantly but the absolute angle will change (both of the calib wheel and the plasma) There is more likely to be a small change in the µ of the edges after the camera was moved between Plas1+2 and Plas3 and then especially Plas4. 93 from CalibSet'Plas1+2': Pol scan, part without beams, part with went to end, stopped and turned around imgIdx 210-410 - symlink to IMGSCAN to remove non-scan at start/end and turned around bit Radiation remove image, no smoothing Demod Save as DSH_CALIB Time seq of Average large box Manually fit triagnel λ=170, φ=0.05, A=22.5°, y0=22.5° gives: 45° at 211.15 22.5° at 253.65 (253=22.85°, 254=22.3°) 0° at 296.15 22.5 at 338.65 45 at 381.15 Varying global µ in DSHDemod module, to fit non-beam by eye = 79 Best µ for beam is at worst 77 For average of nearly whole image is µ~74 Using non-beam 254 at 22.3° as calib, saved trace for centre average to pres/firstResults/imageWork/firstCalib/p_93-plas1-muCalib-nonBeamImage-trace.bin beam calib is then a little off across image. nearest clean looking to 338.65 (22.5° during beam) is 335 which is supposed to be 20.57 Acutally, on image: L=20.5, R=21.25, T=20.75, B=20.75, BR(broken by?)=21.75 so using non-beam calib image gives beam error of +0.75 - 0 There's a slope in X of almost 1° across the image all the way up the beam data, so it's probably not quite right for some reason. Using 335 (beam) at 20.57°, as the calib: At m(13°) with beam, now around 0.3° of slope At m(36°) with beam, now around 0.5° slope At 27.... ~0.3° Seems to be better, although it's hard to tell what the bottom-right contamination is actually doing. 348 is possibly a better bwith beam calib, supposed to be at 27.45 92, also for Calib Plas1+2 Looks like stop and reverse motor between non-beam and beam data, so very hard to get a lock. Broken in IMGSCAN1_* with no beam, and IMGSCAN2_* with beam. Both are ranging ~20 - 45° and back to ~20 so locking is hard. Working on just plasma, and manually fiddling to λ=180 and φ of triangle fit to centre avg gives something that kind-of works. Looks about <µ>=78 for beam but that looks a little wrong for non-beam at end That gives that img 439 should be 23.1 (+/- 1°) Using that as cal, gives: At m(19), 0.5° variation over view, flat within noise Flat within noise at 27° (noise is ±0.5°) At m(36), 0.4° variation (n oise is 0.5) So, need to see how that compares to calib from 93 This is probably good enough for the general sense of µ being around 76-80 and also those two give a single calibration image for cross-image variation although 93's is containated on the RHS. Should compare these between 92 and 93. Will still need to check field variation and abs µ for other calib sets Save #93 calib to AUG/93/CALIBFIT_DSH_BEAM_3x3_0 and ..._NOBEAM_... Using these for 92 doesn't look so good - there is up to a 2° drop in the bottom right corner (beam edge) Using it for #61 of the previous day is good, and gives <0.3° over image variation. Using it for #74 - the nice plasma current ramp 'looks' good - it gives a pretty much entirely linear slope of 10° as in the sims (not actually compared yet) Using it for #178 (imse comparison) gives a bit dent in the lower right corner so the µ field variation is definitely different for Plas3 For Plas3: 132 is overexposed but had ok beam data for imgs 90-114 Says polariser is at 2600 (rahter than 2660, which was calced 22.5°) 2600 should be ~24.5 with no µ, m(cent) θ=21.2°, with µ=79, θ=24.8, which matches, for 22.5° actual, µ would need to be 93 which seems unlikely. 139: looks ok, (and has interlace) but bottom right corner is broken as before Should be at 2660, which is supposed to be 22.5deg This is either wrong, or µ is now ~92 which seems unlikely Alternativly, polariser is actually at 2614 (24.7°) which could easiy be 2600 set rather than 2660. Without any calib, variation is about 20 on left, 21.5 in middle and 19 at bottom RHC (which might be broken) Bottom corner is a problem, and might cause up to 1.3° error there, hmmm Saving as CALIBFIT_DSH_BEAM_3x3_0 Using that for 123 current ramp looks very linear 132: For Plas4, the µ image variation appears to be much less, esp on the RHS Probably due to better camera alignment There is no proper polariser scan for this set 162 and 163 provide reasonable references. Assuming that the central µ is still around 78% as in Plas1-Plas2 (hmmm) the actual from the measured are: 162 m=20.9 a= µ=81-->θ=23.9, µ=77-->θ=24.6, µ=79-->θ=24.2 163 m=29.1 a= µ=81-->θ=31.7 , µ=77-->θ=32.3 , µ=79-->θ=32.0 From 93, linearity is <0.5° down to 3°, and <0.2° down to 5, at 10° it's 0.1°. Image plane position Back in HGW lab graph paper on plate in lens holder short rods Moved plane to get focus best for same lens focus as at AUG plane from end of short rods = 20.0mm (really!) plane to back surface of lens holder = 29.6mm big positioning rods: dist from end to front mount = 152mm dist from first MSE surface (at end of big pos rods) to end of small rods = 24.5mm dist from fist MSE surf to image plane (direct meas) = 46mm dist from fist MSE surf to image plane (to + from small rod ends) = 44.5 so 45±1 mm