=== Minutes of 2000/04/13 SVD meeting === Attendance: Alimonti, Hazumi, Ishino, Kaneko, Kawasaki, Lange, Marlow, Rozen, Stanic, Tajima, M.Tanaka, Tomura, Trabelsi, Tsuboyama, Varner, Yamamoto, Yokoyama, Zontar. 1. SVD damage summary (M.Tanaka) On April/1st between 18:50 and 22:35, noise and gain of VA suddenly decreased. (I wish this was April fool). 20% of layer 1 phi strips shows no gain. They are located at top and bottom side. No large dose was observed during this time although KEKB experienced injection problems. Tanaka and company tuned VA bias parameters and salvaged 25% of them. Vfp was moved to -0.2v from -0.3V (lower resistance), which is opposite of radiation damage effect. Since lower gain channel shows linear response to the test pulse, low gain is not caused by saturation of amplifiers. Kaneko measured leak current across the AC coupling capacitor and found less than 50 fA/strip after 10 MRad. So leak current across the AC coupling capacitor is ruled out. Tsuboyama found that we had leak current imbalance around 20:00. According to Tanaka, the DSSD with large imbalance is correlated with malfunctioning VAs. Tanaka continues investigation. 2. Radiation and BG status (Zontar) Accumulated dose is 53 kRad. Daily dose is about 500 Rad/day. We had two incidents with 6 kRad and 3 kRad dose. The latter was caused by earthquake under Tsukuba area. We might have received much more dose if we did not have hardware abort! We should praise the efforts by Zontar and Tsuboyama. Daily aborts are more or less stable (one/day for both hardware and software aborts) except for some bad days (6-8 aborts/day). --------------------------------------------------------------------------- Reports on the April 5th event: Dear BELLE colleague: This morning (11:30 5-Apr. JST) our SVD had another very big dose (~5 kRad according to Radfet reading). Very fortunately it is found that there are no new damaged channels in the SVD system by the investigation made immedeately after the accident. Anyway we add more 5 kRad (equivalent to 10 normal day running) to 45 kRad dose accumulated by this morning at once. I have to explain more abut the situation since I was at the KEKB control room and manipulated the LER Mask. In these days, Akasaka (KEKB) found very clear evidence that the IR vertical masks cause a significant vertical oscillation in a bunch train. We clearly observed that the oscillation frequency is closely related to the position of the IR mask. The amplitude of this oscillation is evaluated as an order of several micron at IP and the luminosity could be severely degraded by this oscillation. (Remember that our vertical beam size is as small as 2 - 3 micron.) With the preliminary study, we found that these masks have no effect to reduce the BELLE background now (previously when vacuum level was not good, they were very effective to reduce the background). This morning we tried to check whether these masks are effective for an injection background. If they are not effective again even in the Injection (this is true for HER and we have removed them for a better and cleaner injection). The check was done during the injection of Physics run (we intended to make a simple test in parallel). Mask were gradually opened from the nominal position 1mm by 1 mm with observeing background indications. Suddenly beam was lost without any indication of background monitors. SVD pindiode issued beam abort correcrly. QCS system quenched at the same time probably due to beam lost in its super conductor (we have a similar experience once when the vacuum leak accident in the mask happend in the last december.) The Radfet showed that the dose deposited in that moment is as much as 5 kRad. As mentioned above, it is our great fortune that our SVD shows no serious damage in the calibration run made just after it. The possible explanation of the present beam loss in the above incident is that the mask encouters some resonating points of the beam oscillation during its movement. It is not clear why it happens during injection and not in the test in the storage period. -- ------------------------- Best Regards; Junji Haba Dear colleagues, Prof. Tsuchiya (KEK, Cryogenics) asked me to check the F/B and PHI dependence of radiation level for the accident. I summarize the measured radiation by PIN and RADFET on both April 1 and 5. We have 4 PINs and RADFETs on forward and backward side. Unfortunately they are located behind Aluminum block and I found they show much smaller dose than those inside SVD region. Anyway, the figures are stored to /b01/svddaq/april1-2000-phi.ps and april5-2000-phi.ps. On April 1, we do not understand SVD was damaged, I follow Haba's assumption and extracted values around 21:30. Here, 'FWD' means downstream of HER or R side. PIN measurement (mrad/sec) April 1 21:30 April5 11:30 Position FWD BWD FWD BWD 000 160 220 60 50 090 130 80 35 80 180 70 --- 13 -- 270 70 20 28 14 (Degrees) RADFET measurement (krad) April 1 April 5 FWD BWD FWD BWD R / K R / K R / K R / K 045 0.05/0.20 0.02/0.02 0.80/2.5 0.65/1.5 135 0.02/0.02 0.01/0.01 0.12/0.3 0.40/0.2 225 0.02/---- 0.01/0.01 0.18/--- 0.18/0.3 315 0.02/0.02 0.02/0.02 0.14/2.8 0.35/0.8 The PINs and RADFETs inside SVD shows, PIN-F-180 175 900(maybe saturated) PIN-F-270 170 900(maybe saturated) PIN-B-090 250 470 PIN B-180 80 180 PIN VA1 35 130 (mrad/sec) RADFET-K 0.10 5.5 (krad) Summary It seems that the radiation around 0-90 degrees are high. On April 1, when HER was injected, FWD was higher and on April 5, when LER were studied, BWD suffered higher dose. It could be explained that radiation entered to monitors from the central region (Beryllium window), not from the behind which is protected tungsten or lead masks. Best regards, Toru Tsuboyama (KEK)