RPC weekly meeting

Monday 15 Jul 2024, 09:00 → 10:55 Europe/Zurich

5/R-026 (CERN)

Davide Boscherini (Universita e INFN, Bologna (IT)), Giulio Aielli (INFN e Universita Roma Tor Vergata (IT)), Mohamed Zaazoua (University of Science and Technology of China (CN))

Weekly meeting about the RPC system activities

09:00 → 09:15 Communications to/from the gas group

Speakers: Beatrice Mandelli (CERN), Enrico Pastori (INFN e Universita Roma Tor Vergata (IT)), Gianluca Rigoletti (CERN), Giulio Aielli (INFN e Universita Roma Tor Vergata (IT)), Kacper Kapusniak (CERN), Roberto Guida (CERN)

1.png

2.png

fig1.jpg

fig2.png

fig3.png

fig4.png

RPC_sensor_a.jpeg

Giianluca's GASelog:

In the morning the piquet received the warnings L/ATLRPC_Hu_MoistRegAA>2024-07-12 07:50:35.850. I went on place in ATLAS to check what was going on.

I found that since the 1st of July, every time purifier 2 changes column, the humidity in the humidifer drops (see fig1.jpg)

That means that purifier 1 is not able to absorb all the H2O of the system, so I checked the trend of the valves of purifier 1 and I found out that the time each column stays in run has changed (see fig2.png)

This behaviour could only be justified by the Khrone, and actually the flow read by the Khrone has decreased significantly (see fig3.png).

The suspect is that the Khrone was not reading the correct flow. If I compare Khrones of purifier 1 with purifier 2 during the columns run I see the following

	Purifier 1 (Column A/B avg)	Purifier 2 (Column A/B avg)
Flow before July 2024	8-10 m3/h	3.3-3.6
Flow after July 2024	4.5-5.1 m3/h	Same

 

To note that this trend is not really correlated with the issue of the two columns having a different pressure (see 4.png) and I also checked if it could be related to the operation of purifier 3 but it doesn't seem there is any correlation

 

-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

Col A is also affecting mixer consumption. When it enter in the loop (see 1.png) there is a spike in the mixer injection because the pressure of the loop (PT5011 and PT5005) slightly drop (see 2.png)

09:15 → 09:25 RPC LVL1 status

Speakers: Claudio Luci (Sapienza Universita e INFN, Roma I (IT)), Gian Luigi Alberghi (Universita e INFN, Bologna (IT)), Marco Sessa (INFN e Universita Roma Tor Vergata (IT)), Massimo Della Pietra (University Federico II and INFN, Naples (IT)), Sabrina Perrella (Sapienza Universita e INFN, Roma I (IT))

09:25 → 09:40 RPC Dectector Status

Speakers: Daniela Katherinne Paredes Hernandez (University of Hong Kong (HK)), Eric Ballabene (University and INFN, Bologna (IT))

• Summary of RPC cavern access: 8th of July [D. Boschereni, C. Luci, A. Polini, M. Zaazoua, L. Zhang, e-log]
  

  • BOS5A02.CO was showing a trigger hole in the corresponding tower due to the Vee not driving current. The connector was found unplugged--> reconnected. +8 gaps recovered.

  • A few HV channels that were operating at lover voltage were brought back to nominal voltage.

  • Disconnection moved from rack side to chamber side for the following HV channels:

    • BOL.A2.07.CO.Ly0: BOL3A07.CO.DP1.Ly0.DZ1 disconnected, driving high Igap current (68uA).  +3 gaps recovered.
    • BML.B.07.PI.Ly0:  BML1A07.PI.DP1.Ly0.DZ1  disconnected driving high Igap current (~60uA).  +1 gap recovered.

  • Noise checks:  BML.A.15.CO.Ly0

    • After checks on rack side we see that the problem is at the level of the patch panel.
    • There is a swap at the level of the splitter, between BML3/4 and BML5/6.
    • For now, we don't see noise with BML5/6 receiving HV. We keep the channel ON. +4gaps
    • Next action item: understand at which level of the patch panel the noise appears.
  • ADC board in Y_3605_X8_C4 apparently not working, has been reset remotely. Now it seems working properly.
    
    
• Summary of RPC cavern access: 10th of July [C. Luci, M. Zaazoua, e-log]

  • Noise checks : BML.A.15.CO.Ly0 

    • At the level of the patch panel, the BML3A15 was found responsible for the noise propagated in sectors 14, 15 and 16.
    • BML3A15 kept disconnected at the rack side:  BML3A15.CO.DP2.Ly0.DZ2 BML3A15.CO.DP2.Ly0.DZ1 disconnected and BML3A15.CO.DP1.Ly0.DZ2 BML3A15.CO.DP1.Ly0.DZ1 are connected to the recovery line.
    • Here we recovered two gaps: BML4A15.CO.DP1.Ly0.DZ1 and BML4A15.CO.DP2.Ly0.DZ1.
  • Trigger hole  in sector 5: BOL5A05.CO.Ly1/Ly0
    • Swap between Ly0 and Ly1 for BOL5A05, primary cable of BOL5A05.Ly0 disconnected at the level of the splitter. 
    • BOL5A05.Ly1 tripping at 3300V after connecting the primary cable to the splitter.
    • BOL5A05.Ly0 connected to the right splitter, looks fine at 9350V.
    • Here we recovered 4 gaps, and get rid of 1 trigger hole.
      
      
• BML.C.05.Pl.Ly0 HV channel ramped down after reaching the maximum value of the current [e-log].
  • I_calib and i0 were set to 260. The channel was ramped up to maximum voltage. However, the voltage was lowered by the safety script due to a very high current in BML6C05.PI.DP1.Ly0.DZ2 (47.59 uA at 9000V, fullflow=0.688).  The voltage was lowered to 7000 V and the channel was masked.
    
    
• RPC giving NOT_READY due to a LV channel in overcurrent [e-log].
  • Vth for BML5C11.CO.Ly0.ETA has been lowered from 1 to 0.8. No overcurrent anymore on BML5C11.
    
    
• LV tuning for BML2A11 [e-log].
  •  The Vpd of BML2A11 gives overcurrent error with 1.885V and 0.27A. The overcurrent error is seen even when all Vth are set to 0. A Vth tuning is performed for BML2A11.CO which increases the Vpd from 1.885V to 2.075V (shall be 2.1V). The Vpd current is still 0.27A. The problem originates from BML2A11.CO Vee which drives 3.8A at 6V (instead of 4.2A at 6V like the Vee of BML2A11.PI). I have checked the DDC rates and haven't found low-pT missing hits for s51t1 and s52t1. To be checked in the cavern.
    
    
• BML.C.05.PI.Ly0 HV channel ramped down due to BML6C05.PI.DP1.Ly0.DZ2 driving high current [e-log]:
  • I set v1 to 9250V the first time it ramped down but after a while it still ramped down again due to high current. At the end, I set v1 to 8800V and masked the HV channel on the DCS. BML6C05.PI.DP1.Ly0.DZ2 gas gap to be disconnected on the cavern.
    
    
• BML.B.13.CO.Ly0 HV channel tripped due to high current [e-log].
  
  • Increased i0 to 200A. HV channel back to 9350V.

09:40 → 09:55 Noise issue

Speakers: Claudio Luci (Sapienza Universita e INFN, Roma I (IT)), Mr Mohamed Zaazoua (University of Science and Technology of China (CN))

Noisy channel in sector 9 still off.

Need the Y cables for further checks for both channels  in sector 15 and 9.

09:55 → 10:05 Summary of Wednesday access

Speaker: Mr Mohamed Zaazoua (University of Science and Technology of China (CN))

e-log

Noise checks : BML.A.15.CO.Ly0 

• At the level of the patch panel, the BML3A15 was found responsible for the noise propagated in sectors 14, 15 and 16.
• It is kept disconnected at the rack side:  BML3A15.CO.DP2.Ly0.DZ2 BML3A15.CO.DP2.Ly0.DZ1 
• BML3A15.CO.DP1.Ly0.DZ2 BML3A15.CO.DP1.Ly0.DZ1 are connected to the recovery line.
• Here we recovered two gaps: BML4A15.CO.DP1.Ly0.DZ1 and BML4A15.CO.DP2.Ly0.DZ1.

Trigger hole  in sector 5: BOL5A05.CO.Ly1/Ly0

• 1st we found a swap between Ly0 and Ly1 for BOL5A05 caused by the primary cable connected to the wrong splitter. 
• Primary cabe of BOL5A05.Ly0 found also disconnected at the level of the splitter. 
• BOL5A05.Ly1  tripping at 3300V after connecting the primary cable to the splitter.
• BOL5A05.Ly0 connected to the right splitter, looks fine at 9350V.
• Here we recovered 4 gaps, and get rid of 1 trigger hole.

10:05 → 10:20 HV scan results

HV_scan_results.pdf

10:20 → 10:30 Resin update

Speakers: Ahmet Renklioglu (Istinye University (TR)), Dr Aydin Ozbey (Istanbul University-Cerrahpasa (TR))