Today we started vacuum chamber assembly work

• 10:16 am: Particle count measurement of clean room
• 10:34 am: Particle counting finished: meets ISO 5 standard
  1. Zone 3 :
     • 0.3 u: 3159
     • 0.5 u: 789
     • 1.0 u: 83
  2. Zone 4 :
     • 0.3 u: 498
     • 0.5 u: 41
     • 1.0 u: 0
• 10:45 am: Assemble vacuum feedthrough
• 11:08 am: Finish assembling feedthrough
• 11:14 am: Assemble inverted magnetron pirani gauge
• 11:29 am: Finish assembling magnetron gauge
• 11:31 am: Assemble calibrated Ar Leak
• 11:47 am: Finish assembling Ar leak
• 11:49 am: Assemble up to up-to-air valve
• 11:59 am: Finish assembling up-to-air valve
• 12:00 pm: Break for Lunch
• 1:00 pm: Came back from lunch
• 1:20 pm: Assemble 45 degree elbow [RGA Line]
• 1:35 pm: Finish assembling 45 degree elbow
• 1:35 pm: Assemble Reducing nipple [Pump Line]
• 1:49 pm: Finish assembling Reducing nipple

• 1:53 pm: Attempt to install gate valve [Pump line] but bolts could not fit to the gap between 2.5" reducing nipple (0.95" length, shortest 5/16 bolt is 1.5" in total length).
  Thickness of of CF flange on reducing nipple is 0.75" inch. Ideally, we want to have 0.2" engagement to the gate valve, thus0.95" + 0.25" head thickness = 1.2" > 0.95" gap.
  We thus most likely will need a replacement for the reducing nipple. We decided we would hold up on installing the rest of the pump line until we have got components to fix this problem

• 2:05 pm: Assemble gate valve [RGA line]
• 2:15 pm: Finish assembling gate valve [RGA line]
• 2:18 pm: Assemble 4-way cross [RGA line]
• 2:30 pm: Finish assembling 4- way cross [RGA line]
• 2:37 pm: Assemble manual bellow sealed angle valve [RGA line]
• 2:43 pm: Finish assembling sealed angle valve [RGA line]
• 2:45 pm: Assemble inverted magnetron pirani gauge [RGA line]
• 2:55 pm: Finish assembling magnetron pirani gauge [RGA line]
• 2:56 pm: Assemble vacuum hose, thin wall [RGA line]
• 3:04 pm: Finish assembling vacuum hose, thin wall [RGA line]
• 3:45 pm: Packing up for the the day
• 4:00 pm: End-of-date particle count measurement
  1. Zone 3
     • 0.3 u: 1080
     • 0.5 u: 124
     • 1.0 u: 83
  2. Zone 4
     • 0.3 u: 540
     • 0.5 u: 83
     • 1.0 u: 83

[Cao]

Continuity Test

• Chamber wall to optical table: continuity confirmed, resistance: 0 Ohm
• Chamber wall to ground point connection on chamber: continuity confirmed, resistance: 0 Ohm
• Turbo pump to ground point connection on chamber: continuity confirmed, resistance: 0 Ohm
• Turbo pump to optical table: continuity confirmed, resistance: 0 Ohm
• Optical table to chassis frame outside cleanroom: continuity confirmed, resistance: 0 Ohm
• Front of chassis frame to earth point: continuity confirmed, resistance: 0 Ohm

After Jon's comment yesterday that some of the connection did not seem to have good metal-metal contact, in particular the gate valve connection, I went through the ConFlat connections today and retighten them. I found a lot of the CF connections are not particularly tightened and there were a lot of range left that can be tightened with the wrench. After re-tightening, the copper gaskets are not visible anymore. For example, see the attached images for the difference before and after tightening.

Note for future installation of CF

• After tightening the bolts/ screws in jumping order (to provide uniform torque) and there is resistance appearing in further tightening, start going through each screw/ bolts in a direction, each time applying a small torque until it resists to further tightened
• After each time going all the bolts/ screw and returning to the starting point, one will find they can further tighten the screws/ bolts. Repeat the process until no further tightening can be achieved
• The copper gasket should not be clearly visible at the connection

Following the problem with contamination particles observed in the the Leybold thermovac TTR 91, we have taken the following step to clean the gauge:

1. Wipe with IPA-wetted Vectra Alpha: Cover the tip of a ziptide with Vectra Alpha wipe corner that has been wetted with IPA, push the wipe around and remove visible particulates
2. Fill the gauge flange with IPA: (following Jon's recommendation to use Lesker's procedure for cleaning their Pirani gauge), flip the gauge up such that the CF flange points upward, fill the flange with IPA all the way up. Use the tip of a SSTL tweezer to agitate the IPA. Let it sits for 20 minutes, agitate every 5 minutes. After 20 minutes, pour the IPA out then spray with dry pure nitrogen
3. Passive drying : Let the gauge sit inside the flowbench for 3 hours (3:30 pm to 6:30 pm)to ensure all IPA has evaporated
4. Baking : Leave the gauge in oven at 50 degree C for 48 hours (maximum allowed temperature is 65 deg C, we use 50 deg C to ensure we are well below this limit). Baking started at 7pm, should finish on Sunday 7pm and ready for assembling to vacuum chamber on Monday morning

The display resolution of the logrus has been 1064x768 and has been the only option, which is not great. While remote access to logrus using rdesktop allows rendering a virtual display of user's chosen resolution, it is not fast when using graphic-intensive program. NoMachine allows user to take over and control the machine remotely and thus appears as the same machine. However, NoMachine cannot render a virtual display like rdesktop. This has been limiting the resolution of using logrus via NoMachine.

On Friday, I found that even though we had NVIDIA Quadro P600 graphic card installed, we were not actually using it. The monitor has been connecting to the the integrated VGA display connector. This uses Microsoft Basic Display Adapter which limits the the resolution. Today I got a HDMI to Mini-display cable to connector the monitor to the Nvidia graphic card. Then in Display settings, go to Graphic settings, then enable Hardware-accelerated GPU scheduling. After restarting the machine, the machine recognised the Display 2using the Nvidia Quadro P600. In multiple displays field, selected "Show only on 2" and remove the VGA connection to the monitor. Logrus is now set at Nvidia Qadro P600 native display resolution of 1920x1080. This is now also the display resolution in NoMachine.

TL;DR:

Full version:

20 June 23:

• Turn off heater (short term testing) Close valves to RGA line and main volume. Pressure (Torr) readout:
  • Gauge 1 (main volume): 3.43E-6
  • Gauge 2 (RGA line): 2.61E-6
  • Gauge 2 (pump line): 3.8E-4
• Turn off turbo pump, let vacuum line vent
• Connect high temperature control unit to main power, power up, set setpoints:
  • Temperature setpoint: 150 deg C
  • High temperature setpoint (to switch off) : 175 deg C
  • Hysteresis: 10 dec C
• change setpoint of PID control unit to 150 dec C, with high alarm set at 165 deg C
• Connect power output of high temperature control units to PID temperature control units, ensuring both are off
• Turn off scroll pump
• Replace Pirani gauge in front of turbo pump with blanks (see image)
• Pump the vacuum line back down
• Open up valves and wait until pressure of full system drop back down to 1E-6
• Install the high temperature controller thermocouples onto the chamber (see images):
  • One thermocouple is installed on the other side of the cross before the turbo pump
  • One thermocouple is installed on the main volume next to the PID controller RTD
• Attempt to remove electronic box and magnetic shield of full range gauges:
  • Electronic box is easily removed with an Allen key
  • Magnetic shield cannot be removed since the gap between the vacuum flange and the gauge bolt is too short for a spanner to go in. This is mainly due to the bolt washers are in the way (see images)
  • To remove the magnetic shield, the main volume has to be vented and the gauges need to be readjusted.
• Attempt to power heaters (but only heat to 80 deg C):
• Leaving the gauges intact, the high temperature controller + PID controller are turned on to test heating the chamber
• At this point, both of the high temperature controllers are connected to the same power strip (Vacuum chamber end of the table)
• After heating up to 65-70 deg C range, the powers turned off on both controllers + the lights on one end of the table
• Upon inspection, the power strip has turned itself off, I don't know what the model of the power strip but we should check its current limit
• For a quick test, I connect one of the two heater controlling system to the power board on the work desk while leaving the other running off the same power strip: The heaters appeared to be function ok and able to reach 80 deg C, at which point I turned it off and left for the day

21 June 23:

The goals of today was to remove the electronic boxes and magnetic shields on full range gauges to enable high temperature bakeout

• Upon returning to the lab, the gauges readout of vacuum pressures were:
  • Gauge 1 (main volume): 3.31E-7
  • Gauge 2 (RGA line): 3.40E-7
• Venting the both main volume and RGA line
• Remove the two full range gauges from their flanges
• Electronic boxes and magnetic shields were removed according to the gauge manuals (see images)
• Mount the gauges (with only its probing volume) back onto their respective flanges, ensuring that the flange washers do not block access to the magnetic shield screws
• Only remount magnetic shield + electronic box for one gauge (gauge 2: RGA line) for monitoring chamber pressure
• Turn on scroll pump and let the chamber pump down
• Rearrange heater controller connections while waiting for pumping down:
  1. Controllers for bottom + pump & RGA line heaters remain the same: connected to power strip on vacuum end of the table
  2. Controllers for lid + upper chamber main volume: connected to power strip on clean&bake end of the table
• Once the pressure readout by gauge 2 reached 3 Torr, the turbo pump was turned on and left to run for 45 minutes
• Upon returning to the lab, pressure on gauge 2 read 5E-6 Torr and was still decreasing
• Disconnect the ethernet cable to gauge 2, remove its electronic box and magnetic shield
• Turn on heater controller units and let them drive the heaters to get up 150 deg C
• At 95 degree range, I could see smoke at some locations, mainly the RGA and pump lines where the insulation do not fit properly and the velcros are likely to be overheated
• Upon reaching 90 degree range (measured by sensors), the powers turned off on both sides of table (including heaters, lights and scroll pump)
  • Turn off both heaters controllers and unplug them from power strip. Attempt to switch the power strip back on but nothing happened
  • It is now not the power strips problem and must be further down the line from outside the tent, most likely the power supply that these strips connected to: What is the power supply model + current limit?
• The heater controllers left unplugged, I temporarily plugged scroll pump to the desk power strip so it can still run

After leaving the the chamber to bake at 120 deg C from Thursday 3:41 pm. Today I started to cool the chamber down to room temperature at 11:30 am (total bake duration: 91 hours 49 minutes). Ideally, this process should be ramped down slowly, approx. 6 degree per hour. However we had no ramping function with out controller. The only method to tune this cool/ heat rate is to tune the PID parameters, which are:

1. Pb : Proportional band, this quantity is inversely proportional to the P-gain
2. ti : Integral time. Increasing integral time makes the output response slower to error, thus the opposite effect of increasing integration gain
3. td : Differential time

1. Pb : 50
2. ti : 100 s
3. td : 25 s

1. Pb : 200
2. ti : 800 s

To access and control the Red Pitaya using Python locally on a machine within the local network, one should follow these steps:

1. Start the SCPI server. This is achieved by first log onto the Red Piatay page

   rp-xxxxxx.local/ 

2. Go to Development >> SCPI server and turn the server on. (Note : The server is currently running)
3. Communication with Red Piataya is done through PyVista, install PyVista with:

    sudo pip3 install pyvisa pyvisa-py 

   This has been installed on Chimay. Ensure that you have pip3 install, if not, you can install it using:

    sudo apt-install python3 pip 

4. To start talking to the RedPitaya, ensure you have the scipt

    redpitaya_scpi.py 

   in your local folder. This is the standard class that you will import to your code to establish connection with the Red Pitaya. This code can be found in directory

    ~/RedPiatya 
    or this link
   

[Aiden, Cao]

1. Helium leak test

• 2:00 pm: Uninstall regulator from nitrogen gas tank and move to helium gas tank. Place helium onto cart and move to the rear door of the clean tent. With the tank remaining outside, feed hose through the flexible wall to use in clean tent - See attached image
• 2:25 pm: Start Helium leak test: At each CF connection, He gas is sprayed while its level is monitored with the RGA in leak detection mode. A constant flow of helium is maintained until the level of helium detected by the RGA plateaus out.
  • All connections show low helium leak detected (< 3e-11 Amp)
  • The connections with highest leak detected are from reducing cross to flex hose and cross to RGA. They are on 2 - 3e-11 Amps. All other connections are less than 8e-12 Amps. We verified tightness of these two connection but they cannot be tighten any further

2. Argon calibrated leak test

• We opened the Ar calibrated leak. Upon opening the valve to allow Ar to flow through, we can see a surge in pressure measured by the gauges from 6.7e-7 mbar to 1.2e-6 mbar. This is then drop down and stabilised around 8.9e-7 after 5 minutes.
• Upon the the equilibrium is reached, we ran RGA scan twice in Analog mode with the following settings:
  • Points Per AMU: 10
  • Start Mass : 1
  • End Mass : 100
  • Focus Voltage: 90 V
  • Channel Electron Multiplier (CEM) : On
  • CEM Voltage: 1060 V
  • Unit: Amps (ion current)
  The settings is saved as rga100amu_scan.rga RGA application file in C:/Users/controls/Documents/Vacuum/VacuumChamber/RGA. The data file is saved in Data folder as ArLeak_230512.txt. This file contains the current values for corresponding AMUs. Using this dataset, we can obtain our chamber outgassing rate.
• Using RGA_scan_process.py , which is adapted from Mike Zucker's Matlab code E2000071, and the Ar calibrated leak posted in elog 95 we can compute the outgassing rate to be 150.5E-10 Torr l/s . Our outgassing rate is 37.5 times higher than the required (4.00E-10 Torr L/s per E080177)
• See image attached for the spectrum and the hydrocarbon (HC) tracer masses used to evaluate outgassing
• All codes and data are stored in Vacuum git repo . I'll qrite up an instruction to use the code and post it on our wiki page.

After finishing with Ar cal leak test, we close Ar valve, disconnect and turn off RGA. The He tank is moved back to the wall mount. The chamber is ready for baking installation

Cao and Aiden put the RGA back on to the chamber and measured the outgassing rate after the chamber has cooled down from its first bake. The figure attached shows the RGA measurements with the Argon leak open.

Also put the Full Range Gauge #2 [RGA line] back onto the system to measure the pressure and got a reading of 2.78 E-7 Torr. We believe it is actually lower as the chamber temp is still at 28 degC rather than the 24 it was previously measured at. The gauge also may have needed more time to equilibrate.

Pressure Measured Before Removing Gauge 2.31 E-7 Torr.

Went in and turned on the heating tape in increments of 30 degC until 120 degC was reached. I will let it equilibrate over night and then asses whether or not the chamber can be raised to 150 degC without having the flange to the turbo pump reach over 120 degC as it is not recommended for that flange to be any hotter.

If it can safely be raised to 150 degC, then only 2 days of baking is necessary. If 120 degC is the bake temperature, a week will be needed.

The temperature of the flange connecting to the turbo pump after reaching equilibrium is 74 degC. This means that it is safe to proceed with raising the temperature from 120 degC to 150 degC.

Raised the set temperature of the chamber from 120 degC to 150 degC in 15 deg steps.

Checked on the system today.

The flange on the turbo pump has reached an equilibrium temperature of 87 degC when set at 150 degC. While the main body of the chamber has reached an equilibrium temperature of 142 degC.

Checked on the vacuum system today to change the PID controls back to their default values in order to try and get the temperature of the cross in front of the turbo pump higher. Currently it is sitting at 90 degC before changing the controls. While the main body is at 146 degC.

Will update again tomorrow on new equilibrium with the default PID settings.

Checked on the vacuum system and it looks like it will either need to be increased in temperature or baked for a longer time. The main body is at 146C while the cross is at 91C.

Started to cool down the chamber today. Final temperature readings of the chamber were 146C and 91C for the body and cross respectively.

The insulation on the lid of the vacuum chamber has some discoloration and it may be wise to place a sensor closer to this spot in future bakes.

Took another set of RGA data after the second bake has cooled down to 27 degC.

Just by visual comparison to the other graphs, the HC levels of the chamber has gone down further. I will soon overlay the data for an easier comparison.

I also reinstalled the full range gauge above the RGA line. It is still adjusting itself downwards so I will have to check again to get an accurate measurement of the pressure.

Took some more data to look at the HC count. Initial data was from leaving the Argon leak closed and then left it open for 15 minutes and then took data for the open scenario. Attached are the graphs for these where attachment 1 is the closed scenario and attachment 2 is the open scenario.

From these graphs it looks like the signal-noise ratio of our calibration line is poor with the ratio between these two being 1.3. This means that the calibrated leak is barely being detected over the background. This should improve with more baking and pumping. I have also now attached overlayed graphs comparing the chamber when the leak is closed versus when it is open. As well as one comparing the chamber right after the bake versus the most recent data both with the leak open.

Dr. Richardson and I took off the insulation on the lid again to add more insulation in-between the heater tape folds in addition to what was added on top of the tape after discovering discoloration. Also cleaned the lid with IPA wipes.

Also took more RGA data, one with the leak closed and one with the leak open. The graphs below this show that the SNR is improving with the ratio now looking to be around 3. The chamber is still too dirty to be commissioned and we will need to be baked again soon.

Plugged the temperature sensor back in after getting another extension cord. Removed the electronics box for the RGA as well as the full range gauge above the RGA line.

Changed the pre filter inside of the 3 stage HEPA system next to the soldering station. It was rather dirty and I have attached images with a clean filter on the left and the used one on the right. I reset the pre filter age on the system. I tried to see if I could tell if the HEPA filter was dirty but I could not see it. I did not reset the age of the HEPA or UV stations. They are currently 378 days old.

Moved the temperature sensor on the PID controller to one of the switch backs on the lid.

Also started the next bake with both PID controllers set to 150 degC. When leaving the temperature sensors above the chamber read 119 C and 83 C. These sensors are on the chamber wall and flange in front of the turbo pump respectively.

Did some further testing to determine if it was within reason to increase the temperature of the chamber. I used a thermometer and pulled back some insulation to test many points on the chamber. Mainly focused on the barrel and the bottom lid of the chamber to see if it was possible to increase the temperature of just the PID controller that is powering the bottom half of the chamber.

Found that the barrel of the chamber seems to be hotter than what the sensor is saying, 125 C. Found many points on the barrel that reached 140 C. I believe this is because the sensor is taped on top of the heater tape rather than in between the aluminum and the heater tape.

Also measured many points on the bottom lid as it has the same configuration the top did before it damaged the insulation. It seemed to be consistently at 150 C but I was able to find a couple points that were hotter around 160 C. These hotter spots were most likely the heater folds. There is not much we can do about this as removing the insulation for fixing would required a very invasive procedure to do it safely.

Then moved to the stainless steel tube that connects the RGA and Turbo pump lines. This tube was around 150 C but many points on it were hotter around 160 C. Before another bake I think some of the tape around the tube may need to be pulled back and relocated to avoid damaging insulation while also providing more heat to the main body of the chamber.

Measured a couple more points on the cross that connects the RGA and the full range gauge and found it to be around 130-140 C which means this section could be hotter as the electronics of the full range gauge can reach 150 C.

Over all I believe that the chamber temperature should not be increased in its current state as the bottom lid is a concern with what happened with the top in the second bake. Some optimizations could be made for a fourth bake however. Moving the temperature sensors to be between the heater tape and chamber. Pulling some of the heater tape from the stainless steel tube and relocating it to the main body of the chamber. Moving the PID temperature sensor on the tube to be on the bottom lid or another hotter spot. These are some changes I would make for a fourth bake but maybe this bake could be the last with how the second bake went.

Vacuum Chamber equilibrium temperature for bake 3 are 126 C and 88 C for the chamber wall and turbo pump flange respectively. There were reached with the PID controllers both being set to 150 C.

Also noticed some discoloration around the port insulation. It could be burned from being exposed to the heater tape directly without insulation or it could be that the insulation, which looks much different, is depositing something around the ports.

The turbo pump overheated and the heating tape had to be turned off in order to get the pump back online. It was unusually hot at 52 C. Hottest it got in bake 2 was 38 C.

[Dr. Richardson, Aiden]

DO NOT TOUCH ANY VALVES ON THE CHAMBER. THERE IS A PRESSURE DIFFERENTIAL BETWEEN THE RGA LINE, TURBO PUMP LINE AND THE MAIN VOLUME. THE TURBO PUMP WILL BE DESTROYED IF OPENED.

Closed the valves to the turbo pump line and RGA line. Turned the turbo pump back on and saw that the pressure in the RGA Line and Turbo pump line was going down so there was no large leak within that line or with the valves as the main body pressure stayed constant. Then raised the pressure of the main body of the chamber as we suspect the viton o-ring has failed and needs to be changed. The bolts to the lid of the chamber were also removed but the lid was left on.

Then performed a helium leak test on the section of the system that was under vacuum. We found two flanges that had small but significant leaks compared to the others in the lines. The flange connecting the full range gauge to the cross and the flange connecting the stainless steel tube to the valve under the cross. These showed large spikes when sprayed with helium and were then tightened to hopefully fix it. After tightening, the flange under the full range gauge was fixed as the bolts were pretty loose and tightening reduced any spike seen on leak test. The other flange on the SS tube could not be fixed with tightening and it still shows signs of helium leaking through. This flange will need to have its gasket changed to fix this leak. Then left the system where it is in the picture for further changes tomorrow August 11. Again, DO NOT TOUCH THE VALVES ON THE RGA LINE AND TURBO PUMP LINE AS THERE IS A PRESSURE DIFFERENTIAL AND IT WILL DESTROY THE TURBO PUMP IF OPENED.

Note: We also found that the pump under the table had flipped the breaker and needed to be turned back on. This may be something to look into.