On 19/11/2018:

During the target ladder calibration (see elog entry #41) we also replaced the collimator (thickness ~24mm, diameter 3mm) located immediately upstream from the target ladder - that is between the upstream and central sections of the chamber. We mounted an aluminium collimator (thickness ~1mm, diameter 5mm). See attached picture 1.

Using the same telescopic sight as that used for the target ladder calibration, we aligned the 5mm collimator with the beam axis.

This morning I unpacked the chamber parts.

I mounted the downstream and central lids with the iris arm hooks to double check we could still successfully mount them, see previous eLog entry (#25) for photos from this procedure performed in Edinburgh.

When attempting to move the aperture arm, I noticed the hook was catching on the Al tube used to support the aperture onto the side of the chamber collar. I cut the hooks used for both the downstream and central apertures, they still hooked onto the iris arm.

The actuator positions were found as follows:

Downstream:

Mounting (iris "closed" to ~18mm diameter) aperture setting = 82.5mm

During measurement (iris open) aperture setting = 1.0mm

Central:

Mounting (iris "closed" to ~18mm diameter) aperture setting = 87.0mm (Old = 85.5mm)

During measurement (iris open) aperture setting = 14.0mm

Note: these were measured by moving the actuator arm away from the chamber collar (105 -> 0 on actuator scale). This is to avoid backlash in the position, particularly because the iris arm hole allows moderate (~4mm) movement of the aperture arm hook.

I then cleaned the upstream, central, and downstream lids, the chamber collars o-rings and top and bottom surfaces, and the metal part of the diode mount all with alcohol.

Me and Raffaele mounted the chamber collar onto the chamber target section (already mounted on beamline from February 2018 shifts)

We then attempted to mount the downstream lid, with S2 detector mask, onto the collar, however we had difficulty connected the iris arm to the actuator hook. This was also the case for the central lid.

We decided to implement a new approach, this requires 2 people to ensure the aperture can be adequately supported during mounting:

1. Leaving the aperture off of the chamber collar, mount the chamber lid (downstream/central) with iris arm mounted

2. Using tweezers or another small tool, gently pull the iris arm out through the chamber collar side hole

3. Using the iris arms hole, hook the aperture arm onto the iris arm

4. Carefully insert the aperture into the chamber collar side hole, whilst simultaneously shortening the aperture arm to ensure the arm will fit

5. Bolt the aperture onto the side of the chamber collar

We performed this operation for both the downstream and central lids, we were successful in both cases on our first try. The irises were visually seen to open and close by looking through the beamline from the position of FC-E0.

When mounted the iris "closed" (18mm diameter) positions were:

Downstream = 82.5mm,

Central = 85.5mm (suggesting my 87.0mm quoted above is due to the iris arm hole size)

On the central lid I mounted the diode at 39 degrees and 22.0mm from the upstream side of the lid. This was very efficient due to the new mounting block. See pictures IMG_2823 and IMG_2839

I also mounted the S2 2623-26 in the central position with 12um mylar. See attached IMG_2841

In the meantime Raffaele mounted the target ladder with the following targets:

I need to check the CH2 target number at position 3. Target VII was removed from position 3 because it looked like it was peeling off -> needs additional epoxy

At 18:45 we left the lab.

Tomorrow morning we intend to mount the downstream S2 detector with 12um mylar and both aperture hooks. We then plan to pump down the chamber, setup the electronics, and perform an alpha source calibration.

If all goes well we will perform a beam measurement of 7Li at 5.0 MeV

On 04/04/18:

We mounted S2 2323-18 (496um) in the downstream position with 12um mylar

We mount the central and downstream iris arms, we set the central arm position to 14.0mm and the downstream arm position toe 1.5mm - opening both irises.

We started Ecodry 1 and 2, we checked T1600 Backing valves was closed. At 10:11 cracked open the gate valve. After several minutes of pumping Ecodry 1 cutoff, suggesting a big leak.

The downstream chamber lid was sitting at a slight angle, we vented the chamber and reset the downstream aperture to 82.5mm. We removed the downstream aperture. Lifting off the downstream lid we found the O-ring between the lid and collar was not correctly in its groove, this was corrected.

We closed the downstream lid and remounted the downstream aperture. We set the aperture back to 1.5mm (iris open). We closed the turbo vent valves ready for a second venting attempt.

There was a problem with the Labview DAQ, Guiseppe and Raffaele fixed this.

Ecodry 1 and 2 were switched back on, at 11:20 the gate valve was opened. At 11:28 the gate valve was fully opened and pressures read < 1mBar on all 6 gauges. T1600 Backing valves was opened allowing Ecodry 2 to pump on chamber. At 11:31 all 7 turbos were switched on.

During pumping we began mounting the preamp boxes onto the chamber. We discovered the preamp box to be mounted on the central lid needed its holes enlarging to accomodate the bolts for mounting the preamp box to the spacer box.

We mounted the downstream preamp box first, currently holding preamp 2 with the now copper taped preamp - feedthrough signal cable. We then mounted the central preamp box, holding preamp 1 with a copper taped preamp - feedthrough signal cable.

We daisy chained the preamp -15V and +15V cables for the two S2 preamp boxes and the diode preamp, we also daisy chained the +ve test input for the two S2 preamp boxes.

Originally, the central S2 HV was connected to the Silena module CH1, the downstream S2 HV was connected to Silena CH 2, and the diode HV was connected to Silena module CH3.

Powering the diode to -70V, leak I = 0.00uA.

However when powering the S2s to -130V, no leakage current was observed. It turned out the silena CH2 did not work, moving the downstream S2 HV to silena CH4 and powering to -130V reads 1.69uA leakage current.

For the central S2 the problem was found on the respective preamp box. It turned out the SMA connection of the BNC->SMA feedthrough was loose, tightening this and powering the central S2 to -130V reads a leakage current of 2.47uA.

The target ladder was set to position VII - alpha source facing downstream.

Checking RAL signals for central detector, RAL X, XII, XIII, R, S, Q all looked good, RAL U CH 6 was dead, RAL W channel 0 and 1 were dead.

We discovered RAL I (dowsntream S2) showed poor noise, for a 0.2V pulser setting, the amplified signal on RAL I CH 2 was ~5V with noise ~400mV. See 20180404_Pulser_PreampLid_ON.jpg

I emailed Tom Davinson for ideas, he suggested checking what happens when the preamp is disconnected from the vacuum feedthrough.

When the preamp was disconnected from the vacuum feedthrough the amplified signal on RAL I CH 2 was ~1.5V with noise ~40mV. 20180404_Pulser_DisconnectedFromFeedthrough.jpg

We collected vias vs leakage current curves for the two detectors. See attached S2_2323-18_496um_Bias_Vs_Leakage_Current.png for the downstream S2 and S2_2623-26_505um_Bias_Vs_Leakage_Current.png for the upstream S2.

I suspected the new cable between the downstream S2 and vacuum feedthrough is damaged - perhaps it has a short.

I also checked the central S2, again for a pulser setting of 0.2V RAL X CH 1 showed a 1.2V signal with ~50mV noise, which seemed a better case than for the downstream S2.

The time was 20:20, we decided to stop for the day. The turbos, Ecodry 1 and 2, both S2s, the diode, and the preamp power were switched off ready for dismounting the downstream S2 for a cable change the next day.

Today we began examining and modifying parts of the chamber shipped from CIRCE to Edinburgh.

The parts shipped were:

• Chamber Collar
• Central Lid
• Downstream Lid
• Two Metal Detector Masks
• Two Iris Mounts
• One Iris
• Two Actuators
• Pin Diode Mount

This morning we began piecing parts of the chamber one by one, starting with putting the lids on the collar and checking how easily the iris arms fit. We found whilst fitting together the chamber that the actuator used to control the downstream iris will need to be moved further away from the chamber than the current Al tubes allow. This is because currently the aperture arm hits the metal iris mount. By moving the actuator further away there will still be space for a thin wire to hook onto the iris arm.

Originally the S2 detector was intended to be mounted with the front strips facing downstream. We found in CIRCE during the recent February shifts that this was not possible, because the connectors on the detector PCB hit the lid.

Today Peter machined space for the downstream detector connector, so now the detector may be mounted as originally intended. We checked the vacuum feedthrough still fits too.

However, the ribbon cable between detector and feedthrough will still come between the actuator arm and the iris arm, so adjustment of the iris from outside the chamber is still not possible. We are still working on this.

Between 7th to 21st March modifications were made to the top of the CIRCE target chamber.

These modifications address the three key issues highlighted during the February shifts

1. The central S2 detector did not fit, because its connector pins came into contact with the chamber lid

Additional machining was performed on the central chamber lid, allowing the pins to fit

2. Mounting the iris hooks was non-trivial, resulting in the inability to open and close the irises whilst the lids were mounted on the chamber

Peter designed a guide rod allowing the hook to be connected to the iris arm before mounting the lid onto the chamber. Now the only difficult part is ensuring the actuator arms hook connects to this guide rod.

I checked the position settings needed for both actuators when mounting or removing the chamber lids, these are:

Central actuator setting = 85.5mm

Downstream actuator setting = 82.5mm

The actuator positions for a fully opened iris still need to be checked in CIRCE.

When mounting the lids the guide rods will be arranged such that the irises are in the mostly-closed position (~16mm diameter). This reduces chance of damage made to the S2's during installation and removal, and also means the iris is already setup ready for beam tuning.

See attached pictures for the downstream and central lids.

A sample S2 in Edinburgh was used to check the connectors fit with the lid and other parts. When mounting the lids onto the chamber collar it was decided to leave the S2s off the iris mount. This was to avoid the risk of damaging the S2 from slippage of the parts - note in CIRCE the collar will be on top of the target section with threaded rods and so the conditions are more stable than on tabletop. During lid mounting, the metal mask was used in place of the S2 to simulate its size.

Marks "CE" and "DW" were made respectively on the central and downstream lids and guide rods.

3. The pin diode angle was not reproducible

A mounting block has been machined in Edinburgh, it has an angled slant of 39.0 +/- 0.5 degrees for setting the angle of the plastic block used to support the diode. We noticed in the CAD drawing that mounting the diode at the closest distance to the target would mean the target frame is likely to cause scatter from the backscattered beam, therefore we chose to mount the diode in its central position. The angled block has a width 22.0 +/- 0.1mm for setting the distance between the back of the diode mount and the upstream side of the central chamber lid.

An excel spreadsheet was used to calculate the diode angle, see attached "DIODE_ANGLE_CALCULATION.xlsx"

The chamber parts were shipped back to CIRCE and arrived on 26/03/18, the mounting block and PB-5 Pulse Generator were also sent in the same shipment.

On 05/04/18:

At 9:00 the pressures read ~10mBar inside the chamber (static vacuum). We began venting the chamber.

During venting the cable daisy chain on top of the preamp boxes was removed, the downstream preamp box was removed. It was decided to also removed the central preamp box and its respective spacer box to avoid any collisions when removed the downstream lid.

By 9:20 the chamber was vented. The downstream aperture was removed. The downstream lid was removed, the S2-> vacuum feedthrough cable was changed.

The downstream S2 was remounted into the chamber, and the downstream aperture was remounted onto the chamber collar.

During this process the 4mm diameter collimator for the diode was ready, we decided to also mount this in front of the diode since the chamber was already vented.

The central aperture was removed and the central lid was then removed. The 4mm collimator with 0.9um mylar was installed in from of the diode. The 0.9um mylar was held in place on the collimator using an epoxy.

The central lid was reinstalled, followed by the central aperture on the collar.

The downstream aperture was set to 1.5mm, and the central aperture was set to 14.0mm, opening the irises ready for pumping.

At 10:45 Ecodry 1 and 2 were switched on, the gate valve was cracked open. During roughing the downstream preamp box was mounted followed by the central preamp box.

At 11:03 the T1600 Backing valve was opened and chamber pressures read ~1.0e-1mBar. The 7 turbos were started. (Turbos Up 1 and 2, Turbos Down 1 and 2, Turbo Entrance 3, Turbo CSSM, and T1600)

The +15V, -15V, +ve test, and HV bias cables were reconnected to the central and downstream S2 preamp boxes, and also the diode preamp.

Preamp power was applied, +15V draws 1.5A and -15V draws ~0.6A as expected.

At 11:15 CHUP = 2.6e-5mBar, CHDW = 6.8e-5mBar. Central S2 bias ramped to -130.0V, leak I = 3.51 uA. With the preamp lid off a 0.1V pulser on RAL X CH 1 produced a 1.5V amplified signal with ~200mV noise.

I recollected the bias vs leak I curve for the downstream S2. See S2_2323-18_496um_Bias_Vs_Leakage_Current_050418.png

RAL 1 CH1 looked clearer than yesterday, 0,6V pulser setting produces 3V signal. However there was a strange reflection on the end tail of the pulser peak. I mounted both preamp lids but still saw the reflection. This effect dissapeared after disconnecting the diode preamp test input from the daisy chain with the S2 preamp boxes. I replaced this with a 50 ohm terminator.

We adjusted the downstream S2 RAL thresholds (they were 15mV to begin with except RAL I (100mV) which was adjusted yesterday)

Now the thresholds are:

We tried starting the DAQ using an OR trigger from the downstream S2, however we were presented with an error: Too Many Bad Events

We tried disconnecting all of the RAL ECL outputs except RAL I's, we still recieved the error.

We tried increasing the pulser amplitude setting from 0.2 -> 0.6V, still recieved the error.

We tried restarting the FAIR DAQ, after restarting we got good events. We can see a pulser peak and two alpha peaks.

We reconnected the ECL outputs for RAL II -> RAL VIII to the DAQ OR trigger

Note: ADC 2 (module 3) still shows strange "TDC-Like" behaviour, so we avoid this module.

The first 8 channels of RAL I still show alot of noise, we increased RAL I threshold from 90 to 100 mV

Triggering DAQ only on downstream S2, we see a dead time ~ 9% and trigger rate ~3700Hz

I tried removing the downstream preamp box lid and checked both the preamp -> feedthrough and also the preamp -> IDC convertor box connections on the preamp PCB. I resealed the preamp box lid.

Check on RAL I CH1 (downstream S2) noise: for pulser setting 0.5V amplified peak has height 2.8V noise 100mV

Check on RAL X CH 1 (central S2) noise, for same pulser setting amplified peak has height 2.8V noise 100mV, signal to noise ratio ~ 3.57% for both channels

We then connected the central detector ECL signals to the 3rd and 4th MALU 4532 Logic modules in the CAMAC crate. RAL X -> "R" sent to ADC 4, RAL "S" and "Q" sent to first 16 channels of ADC 5. Channel 31 of ADC 5 is diode. Currently the diode CFD threshold = 200mV. RAL X -> "Q" thresholds were all kept at 15mV.

We set the DAQ trigger to an OR of all 129 channels (64 from each S2 and 1 from diode)

Started DAQ, dead time 9%, trigger rate 3500 Hz, event mult = 52.5.

We decided to run an alpha spectrum.

Pulser Settings:

• Fall Time = 100us
• Rate = 10Hz
• Delay = 250ns
• Ampl = 0.5V
• Pol = POS
• Pulse Top = Tail
• Atten = 1X
• Clamp = OFF

14:55 started DAQ for 15 min alpha spectrum, save name: /Remote_Objs/RData_1/180405140038

Dead time = 71%, trigger rate ~ 1600Hz, event mult = 57.6

On 05/04/18, after eLog #28

After the alpha calibrationrun Raffaele arrived and started beam tuning for 7Li 2+ at 5.0MeV using a LiO molecular source

We set the target ladder to position II (80.024 mm) - no target frame or target

Set downstream sperture to 82.5mm and central aperture to 85.5mm (iris diameter ~ 18mm)

Raffaele was struggling with tuning on accelerator end of beamline - decided to perform a pulser walkthrough

16:02 started DAQ for pulser walthrough, save name: 180405150720

Voltages covered: (Pulser amplitude setting)

• 1.4V
• 1.2V
• 1.0V
• 0.8V
• 0.6V
• 0.4V
• 0.2V

Note: pulser is daisy chained between both S2 preamp boxes (NOT diode preamp)

16:56 DAQ stopped, RT = 00:54:05, DT = 67%, Trigger Rate ~ 2kHz, Event mult = 53.1

At 16:57 the pressures read:

• TPU1 = 6.2e-6 mBar
• TPU2 = 9.1e-7 mBar
• TPD1 = 1.7e-6 mBar
• TPD2 = 1.7e-6 mBar
• CHUP = 9.4e-6 mBar
• CHDW = 2.5e-5 mBar

Checked detectors:

• Central S2 bias = -130.0V, leak I = 8.23 uA
• Diode bias = -70.1V, leak I = 0.00uA
• Downstream S2 bias = -130.1V, leak I = 2.44 uA

We opened the line valve, with no beam DAQ reads: DT = 8%, Trigger rate ~ 3400 Hz, event mult = 52.9

Beam tuning was performed with no target frame in the beam axis, beam voltage = 2.1358 MV

We had 1.2nA in FC 4, we supplied -400V to FC-E0 (FC immediately before chamber) for electron suppression

FC - 0 = 1.8nA (unsuppressed immediately after chamber)

We moved the target ladder to position I (3mm collimator)

FC-0 = 600pA

We need to use the ERNA magnet to bend the bean into FC-45 located after the chamber, this faraday cup is suppressed

CSSM magnet ON = 906.13mT, we noticed this section of the beamline needed pumping, switched on pump at 17:45

Received email from Tom D regarding noise, he suggested a few checks:

• Try changing preamp - feedthrough cable (currently have one spare which is wrapped in Al foil instead of copper tape - not ideal)
• Feedthrough - preamp cable drain wire should be connected to PCB at one end and the other left floating (both are attached to PCB at this time)
• Try isolating feedthrough - preamp cable from preamp lid - it may be touching

At 18:00 Raffaele needed to leave the lab, Jeremias could stay until 20:00, we decided to try and address the noise issue further

I switched off the detectors and preamp power, I removed the downstream preamp box lid and disconnected one end of the feedthrough - preamp drain wire (the end where the first 32 channels are connected on the PCB - connector P1)

I remounted the downstream preamp box lid.

The DAQ shows DT = 8% with 3kHz trigger rate.

To check the effect from the central S2, I tried setting the DAQ OR to only come from the central S2 and diode - I realised I could no longer see the pulser peak in the central S2 -> suggested central S2 trigger not working

I changed the cable between the 3rd and 4th MALU 4532 modules and the respective ECL-NIM converter -> now see triggers -> DAQ gives DT = 51% and trigger rate ~ 5.6kHz, but see pulser peaks in ADC modules 5 and 6

I considered only the downstream S2, setting pulser to 1.0V, RAL 1 CH1 amplifies this to ~5.5V with ~20mV noise.

Setting the DAQ OR to only come from this detector, the DAQ DT = 10% and trigger rate = 10Hz, event mult = 133.2. So it seems making one of the drain wires loose had a big effect on the noise.I decided to make the same change for the central preamp.

After switching off the detectors and preamp, I opened up the central preamp box and left one end of the drain wire on the preamp - feedthrough cable hangng loose. This is the same end as the one chosen for the downstream preamp.

Leaving the central preamp lid off, I reapplied preamp power and rebiased the detectors

• central S2 bias = -130.0V, leak I = 8.10uA
• diode bias = -70.0V, leak I = 0.01uA
• downstream S2 bias = -130.1V, leak I = 2.27uA

Checking RAL X CH1 (central S2) amplifiedoutput: 1.0V pulser output is amplified to ~5.5V with 50mV noise. Note the preamp box lid is off, so this looks promising.

I set the DAQ OR trigger to come only from the central S2, DT = 91%, trigger rate ~ 7kHz. High rate seen in ADC 6 channels 1->7 and channel 31

Changed DAQ OR to come from all 3 detectors, DAQ DT = 93% with trigger rate = 6700Hz, still see high rate in ADC6 channels 1-> 7 and channel 31

The time was 20:00 so we stopped for the day, it looks like the downstream S2 electronics is working fine, will check effect of mounting central preamp lid back tomorrow

All detector biases ramped down to 0V, preamp power off

Today we started by mounting the central and downstream preamp box lids.

The pumps were left running overnight, at 9:06 CHUP = 2.1e-5mBar and CHDW = 5.8e-5mBar.

We biased the detectors

• central S2 bias = -129.9V, leak I = 8.09uA
• diode bias = -70.0V, leak I = 0.01uA
• downstream S2 bias = -130.0V, leak I = 1.66uA

The DAQ OR is all 3 detectors, DAQ DT = 82% and trigger rate ~ 6kHz. So the preamp lids made little difference. We saw particularly high rate in ADC 2 channels 1->4, 6 and 7, ADC 4 channels 6 and 31 (both of these are the downstream S2, ADC 5 channels 4 and 5 (central S2), and ADC 6 channels 1->7 (central S2) and 31 (diode)

Decided to ensure copper taped cable is isolated from preamp lid. I cut two sheets of plastic ~ 31 X 19cm and (after powering down the detectors and preamps) placed these between the copper taped preamp - feedthrough cable and the preamp box lid, for both preamp boxes.

Now when the DAQ is triggered from all 3 detectors DAQ DT = 7%, trigger rate ~ 2600Hz, event mult = 49.7

I saw ADC 3 (the "TDC-like" module) was not disabled, i set its ADC threshold to 2046 (out of 2047 channels).

I opened the irises in front of both S2s and moved the target ladder to position VII - alpha source facing downstream.

Now DAQ DT = 7%, trigger rate ~4100Hz, event mult = 18.6

I reset all of the RAL thresholds back to 15mV, I also disconnected the 2nd ECL-NIM converter channel 13 from the Lecroy quad coincidence module because the last 16 channels werent being used for the DAQ trigger anyway.

DAQ DT rose to 85% and trigger rate 32kHz, event mult = 11.8.

ADC 2 channels 1->4, 6, and 7, and ADC 4 channels 30 and 31 still show high rates. Disconnected RAL I and RAL VIII analogue and ECL outputs to remove these.

Now DAQ DT = 22% and trigger rate 8.5kHz, still see a high rate for the diode in ADC 6 CH 31

I tried plugging in the pulser to the diode preamp, pulser setting 0.3V. Ultimately after checking DAQ rate and signals on oscilloscope I raised the diode CFD threshold from 200 to 226mV and settled on this new value.

I decided to keep the diodes analogue delay set to 1us (same used in Feb 2018)

I realised I do need the 2nd ECL-NIM convertor channel 13 in the DAQ OR, because otherwise i could not trigger ADC 6 channels 0->31. This was now remedied

So, the DAQ trigger OR was from RAL II -> VII (downstream S2), RAL X -> "Q" (central S2) and the diode

At 12:03 we started data collection for a 15 minute alpha spectrum, save name  = 180406110805

DT = 78%, trigger rate = 10kHz, event mult = 12.5. DT had dropped to ~ 50Hz during this run. Still saw high rate in ADC 6 channels 1->7 and 31

I was unable to see alpha peaks with good statistics for the downstream S2

At 12:35 I started a 30 minute alpha spectru, save name: 180406114011. DT ~ 73%, trigger rate ~ 9.6kHz

I noticed the central detector leakage current had risen to 10.01uA (bias -129.9V). The diode leak I was still 0.00uA and the downstream S2 leak I was still 2.21uA

When the downstream S2 was connected to the DAQ OR and ADCs a high dead time is seen. Because of this we decided to try and collect some beam data with the central S2.

On 19/11/2018:

During the vacuum test - see eLog entry #43 we performed noise tests on the preamplifiers with no S2 detector connected

First we performed a "tabletop" test - the preamplifiers were not connected to the chamber.

Preamp I for the near S2 is connected to RAL modules (reading left to right): X, XII, XIII, R, S, Q, U, and W, via long (~5m) preamp - ICU cables 13, 14, 15, and 16

Preamp II for the far S2 is connected to RAL modules (reading left to right): I, II, XI, IV, V, VI, VII, VIII, via long (~5m) preamp - ICU cables 0, 1, 2, and 3

We powered preamps I and II with + and - 15V. The power supply shows the correct currents drawn: ~1.5A for +ve and ~0.75A for -ve

We found our test 16 pin cables #45 had a dead channel 0, we switched to cable #46 which shows all 8 channels working.

We found the following dead channels on the RAL modules:

We mounted preamps I and II to the top of the chamber. We mounted preamp I on the far S2 mount and preamp II on the near S2 mount, however we were not connecting to the S2's and so continued with the electronics tests.

Now we saw the following dead channels on the RAL modules:

We checked the signal to noise ratio for RAL X CH0:

The input pulser amplitude = 50mV, the RAL analogue amplitude = 950 mV, the RMS noise = 40mV. The S/N ~4.2%.

We stopped tests for the day.

On 19/11/2018:

We mounted the following in the chamber:

• Iris frames - no S2 detectors or mylar
• Si pin diode mount with 0.9um mylar - no diode
• Target ladder with blank frames - same configuration as elog entry #41

At 15:20 we started roughing pumps Ecodry 1 and 2. We opened the manual gate valve followed by the T1600 valve.

After reaching ~10^-2 mBar we started the following turbos:

• Up 1
• Up 2
• Down 1
• Down 2
• T1600
• Entrance 3
• CSSM

By 15:30 the CSSM turbo had reached 56 krpm operating speed and the Up and Down turbos had reached 833 Hz operating speed.

At 19:51 just before we left the lab the pressures read, in mBar:

• TPU1 = 3.7e-6
• TPU2 = 6.9e-7
• TPD1 = 1.0e-6
• TPD2 = 1.2e-6
• CHUP = 6.0e-6
• CHDW = 1.4e-5

This morning whilst I was setting up the ribbon cabling (see eLog entry #8) I was considering collecting a voltage bias vs leakage current curve for one/two of the S2 detectors, depending on time available.

This would involve using the central section of the chamber and the straight spacer box to mount the preamp on top of the lid.

Note: we currently cannot use the angled spacer boxes for the chamber yet, because these need to be trimmed by Antonio (the technician) to allow more space for the ribbon cables and also allow the box to sit flat on top of the vacuum feedthrough. We will be seeing Antonio on monday morning to prioritise required machine work.

However, I then remembered that after we stopped the pumps last night we did not vent the chamber. This morning at 10:52 the labview program used to control the pumps read the following pressures:

• Turbo Up 1 = 1000mBar
• Turbo Up 2 = 389mBar
• Turbo Down 1 = 86.1mBar
• Turbo Down 2 = 49.7mBar
• Chamber Up = 1000mBar
• Chamber Down = 139mBar

I emailed Lucio to see if he could come in this weekend to vent the chamber and we could then mount and test two of the S2 detectors, one after the other. Otherwise we can do this monday morning.

As an aside there is some serendipity here, the pressures I read this morning indicate if there is a leak in the chamber it is more likely to be found upstream.

Excel spreadsheet showing CH2 and natC target thicknesses attached

(Note: Tom Davinson is now in CIRCE until the 21st)

On 19/02/2018 we performed the following:

This morning we gave CIRCE technician Antonio slanted spacer boxes (for mounting preamps) and also the Al box (used for target ladder) for minor modifications

I needed to unmount the Al box from the target chamber. As I did so some air rushed in as I unmounted the box, indicating the chamber was not back at atmosphere from pumping on Friday.

I also removed the central lid with the iris attached. I discovered the iris was broken beyond repair. We will need to order a new one and some spares.

Learning note: To avoid damage always have irises fully opened when pumping down the chamber.

In the meantime Tom D has been wiring up the rest of the electronics setup. He discovered the CAMAC crate was not supplying any voltage. We removed the crate and double checked the connection between the rear power box and the crate – it was not fully screwed in on the bottom. After fixing this we remounted the CAMAC and found all voltages read out correctly, including a current readout of -10A at the -6V voltage.

Next we set out to vacuum test the chamber. In total today we performed four vacuum tests.

In each test the diode mount, lemo feedthrough, 64-way vacuum feedthrough, and an iris mounts were installed in the central section (the iris mount in the central chamber had no iris due to damage mentioned above). The iris mount with an iris (fully opened) and 64-way vacuum feedthrough were installed in the downstream section. The Al spacer box for the target ladder was also installed, however the target ladder was not because Antonio was not finished with modifications. Instead of the target ladder a blank flange was installed. The two actuators for opening and closing the irises were also installed in the central and downstream chamber sections.

The first two vacuum tests showed a high leak rate, high 10^-4 mBar even without helium injected close to the chamber.

In the first test we found the two 64-way vacuum feedthroughs were causing a leak, after venting the chamber these were removed and mounted flat to the lid using four bolts, one at each feedthrough corner. We then started the second test.

In the second test we discovered a leak was caused by the lemo vacuum feedthrough on the central lid. After the test we put additional grease on the o-ring and around the flange section of the feedthrough (atmosphere side) to help it seal.

On the third vacuum test after ~ 20 minutes of running the up and downstream turbos the pressures levelled to:

Up = 3.6e-4 mBar, down  = 3.4e-4 mBar

We then switched off the turbos and continued monitoring the vacuum. After ~30 minutes from turning off power to the turbos, they had completely stopped spinning, the pressures read:

Up = 6.5e-1 mBar, down = 6.5e-1 mBar

We found during this test the lemo feedthrough was still not sealing, and in addition neither was the Al spacer box for the target ladder.

After venting we found a hair on the lemo feedthroughs o-ring. The hair was removed and additional grease was applied to the o-ring and lid. We also confirmed the nut was tight. For the spacer box we found the bolts used were slightly too long, and so removed ~ 2mm from the threads. The box was then remounted and looked flush with the side of the chamber.

After this test at ~16:00 Antonio had finished and delivered both the modified spacer boxes for the preamps and the target ladder mounting box, now it correctly uses a 16mm centering ring.

We then started the fourth vacuum test, now with the target ladder actuator mounted too (as well as all the other parts from tests 1-3):

At 16:43 the roughing pump in the leak test was started. At ~16:50 the turbos were started, both up and downstream from the chamber.

At 17:04 (~15 minutes later) the pressures in the chamber read: Up = 3.1e-4 mBar, Down = 2.8e-4 mBar.

We declared victory for the vacuum and decided to mount the detectors and an alpha source.

On the central lid we mounted the Si diode with 0.9um mylar and a 1mm collimator at ~28 degrees rotation (relative to the central lid) -> currently this is performed using an angle measuring app on my android phone and one screw, so this angle selection needs to be better reproducible for the future.

We discovered there was no way to mount the S2 detector on the central lid -> either the legs of its connectors collide with the lid, or when the iris mount is rotated 180 degrees (moves detector slightly upstream) the iris then collides with the target ladder. The way we mount the detector on the central lad needs redesign -> I will take the central lid with irises and diode mount back to Edinburgh to address and fix this problem. I can also use this opprtunity ot find a better mounting procedure for the Si diode.

We did mount the S2 2323-18 (496um) detector to the downstream iris mount. 12um mylar was placed in front with a central hole of max width <20mm to allow the beam to pass trough.

On the target ladder we mounted an alpha source at position 5mm. The alpha source specifications are:

Type: EAX

Prod No: 111109-1242002

Radionuclide: 241Am + 239Pu

Activity: 8.153 kBq

Reference Date: 25/3/2010

At 20:49 with the single S2 detector, diode detector and the alpha source mounted in the chamber we started the roughing pumps.

At 20:54 we started the turbos, including not only the up and downstream turbos but also the large turbo located below the chamber. The pressures read: Up = 1.3e-4 mBar, Down = 1.5e-4 mBar.

We went to dinner. Raffaele had fish.

22:19 The pressures read: Up = 1.4e-5 mBar, Down = 3.7e-5 mBar

We mounted the preamp box on top of the chamber with preamp 2 inside. The cables for the detector bias, +15 and -15V preamp power, +ve pulser, and ribbon signal cables were connected.

Biasing the S2 to -130V we read a leakage current = 1.16 uA. Powering the preamp we read currents: V=+15V -> I = 0.9A, and V=-15V -> I = 0.4A, as expected. The pressures read: Up = 1.4e-5 mBar, Down = 3.2e-5 mBar

Using DSO we observe

• RAL 1 CH 0->7 pulser OK, no alpha
• RAL 2 CH  0 no pulser or alpha
• RAL 6 CH 1->7 pulser OK,  no alpha
• RAL 8 CH 6 no alpha, pulser not connected to n+n ohmic strips

We declared today a good day and left the lab.

By 16:40 I mounted all three chamber lids.

The central chamber lid includes the diode detector mount, the iris, a vacuum feedthrough, and an actuator with no hook.

The downstream chamber lid includes a vacuum feedthrough and an actuator with no hook.

I then mounted the target holder actuator with its two aluminium boxes.

I noticed the sides of the boxes were not trimmed down since my last visit in July 2017. So again I needed to remove washers from the bolts directly below these boxes, which are used to connect the beamline and turbo sections of the target chamber. I also noticed the hole of the aluminium box which the target actuator arm sits was not enlarged since my last visit, and so is not large enough to fit a standard 16mm o-ring centering ring. Instead a smaller centering ring is used which means the actuator arm is not centered, for the vacuum tests this isn't a big problem.

Note: when I mounted the target actuator I replaced its 80mm (total length) Al tube with the 60mm (total length) Al tube, this should allow access to all target positions.

After some further setup on the beamline - including checking the turbo pumps were being water cooled - we were ready to start the vacuum test.

At 18:00 we started the roughing pumps

At 18:10 we started the two upstream and two downstream turbo pumps

At ~18:18 the chamber up and chamber down readings were at low 10^-4 mBar, we would've expected lower pressure readings due to the power of the pumping system. We suspected a leak and switched off the pumps.

Lucio noticed the 70mm screws mounting the lids and collar to the rest of the chamber were too long, and either needed to be shortened or additional washers used - otherwise we could heavily damage the chamber thread

At the end of the day we learnt Fillipo will be in the lab tomorrow (saturday 17th) between 9:00 - 13:00, and agreed that I (Thomas C) could come in and connect the electronics cables ready for next week.

The PB-5 Pulse Generator settings are as follows:

Fall Time = 100us

Rate = 2.00Hz

Delay = 250ns

Ampl = 0.50000V

Pol = POS

Pulse Top = Tail

Atten = 1X

PB-5 Pulse = ON

CLAMP = OFF

During the second beam tuning on 20/02/2018 we performed two pulser walkthroughs to determine the ADC offset.

For the walkthrough the rate was increased to 10Hz.

The first pulser walkthough was saved as: Remote_Objs/RData_1/180220170900

The pulser amplitudes used were: 0.1, 0.2, 0.3, 0.4, and 0.5 V

After this walkthrough I remembered channels 24 -> 31 of ADC 2 were unplugged from the RAL shaping amplifier, and so plugged these back in. I then repeated the walkthrough.

The second pulser walkthrough was saved as: Remote_Objs/RData_1/180220172838

Like the first walkthrough, the pulser ampltudes were 0.1, 0.2, 0.3, 0.4, and 0.5 V.

A sample pulser walkthough spectrum is attached.

After lunch we attempted to connect the actuator arms to the iris arm used to change the iris diameter. We need two iris diameters: fully open for experiment measurements and a 5mm diameter for beam focusing

To do this the original design uses a cuboid with two rods, one rod is connected to the actuator arm, and the other rod is connected to the iris arm with a hook - see attached. The rods were first cut down to fit into the chamber. This would allow us to change the iris diameter whilst under vacuum.

We decided to first check we could connect the hook to the iris in the central section of the target chamber, because this has the most amount of access space. We mounted the iris onto the bottom of the central lid using it's two legs.

After several attempts we could not find a suitable position for the hook and box combination, for several reasons:

1. The cuboid used to hold the metal hook will hit the iris legs at the limits of the actuator arms movement. This means the iris cannot be completely opened or closed. We could close the iris until a diameter of 13mm is left, however the active region of the S2 detectors starts at ~11.5mm diameter so this will not work when focusing the beam.

2. Also if we mount the hook facing upwards (above the cuboid) it is too high to reach the iris arm

3. If we rotate the hook such that it faces downwards (below the cuboid) it now reaches the iris arm. However, the hooks height needs to be adjustable because the iris arm moves vertically as the iris is opened and closed. This is currently not possible as the hook cannot be loosed without falling into the chamber

Note: when the iris is fully opened or closed its arm is at its furthest vertical distance from the chamber lid. When the iris is halfway between open (or closed) its arm is at its closest vertical distance from the chamber lid.

We noticed the cuboid used to hold the two rods is assymetrical and so attempted switching around the two rods, this did not solve the problem.

We then attempted a simpler solution: We replaced the cuboid and two rods with two long screws placed directly in the actuator arms threaded holes, see second attached photo.

The idea being that we can simply move the iris arm with these two screws - there will be some backlash to this technique because the distance between the two screws is larger than the diameter of the iris arm. This should not be a problem because we will only need two positions.

However we found the screws still collide with the iris legs close to the limits of the iris arm movement.

Most importantly: We realised that either the hook or screw technique will both hit the ribbon cable connecting the detector to the vacuum feedthrough, see attached sketch

The central collar of the chamber does have a second position for the actuator arm, however this will hit the Si diode we intend to use to measure the backscattered beam. So this will not work.

At 18:30 we decided to leave the lab and sleep on this problem.

The only solution seen so far is to focus the beam with no detectors mounted. Then after focusing is complete: vent the chamber, mount the detectors, repump the chamber down to vacuum, and remount the preamplifiers on top of the chamber. This would not be optimal but it may be a temporary fix until a better solution is found.

This morning both myself and Raffaele mounted the beamline section of the target chamber.

We also mounted the turbo located immediately downstream from the target chamber, see attached IMG_2621.jpg image - the beam will enter from the right of the image

We then mounted the collar section of the target chamber.

We mounted the two linear actuators onto the collar section, which are used to open and close the iris apertures.

These apertures will be essential to stop the beam during beam focussing

This morning we moved the electronics rack to within ~2.5m of the target chamber, leaving room for the ribbons cables of ~5m length which connect the preamps to the amplifiers in the rack.

Then I continued in attempts to solve the iris arm problem, see eLog entry #3.

When comparing to the CAD drawing (entry #4) David pointed out the iris was mounted the wrong way around, and that it should be rotated 180 degrees. I attempted this yesterday but decided to investigate more thoroughly.

I made three changes in total:

1. I replaced the lemo cable on the Si diode with a shorter one (~7.5mm) from Edinburgh, reducing the chance to catch the lemo between the lid and chamber wall

2. I moved the diode mount away from the upstream chamber wall to be certain there was no collision here

3. I removed the nylon screws intended to connect the S2 to the nylon standoffs on the iris

After each change I reattempted mounting the central lid with the iris attached, after the third correction it fit! This suggested the nylon screws were slightly too long.

I measured the thickness of the S2 detector serial no 2323-18:

PCB board thickness = 2mm

Yamatchi connector thickness = 8mm

Bond wire solder thickness on back of S2 ~1mm

The thread of one nylon screw was cut from 6mm to 4.5mm, after screwing this back onto the iris standoff I found the lid still fit onto the chamber. There will be enough thread to still mount the S2's.

At ~13:20 we then went to lunch

Between 9:00 to 12:30 this morning Filipo was working on the weekend 14C measurements, allowing me to work in the lab during this period.

I unpacked four MALU 4532 Lecroy Logic Modules (sent to CIRCE two weeks ago). Three of these modules were installed into the 1434A CAMAC crate, which already had one module installed from the July 2017 shift. One module remains out of the crate as a spare.

I connected 16 way - 16 way ribbon cables between the front of both IDC converter boxes to the back of the RAL shaping amplifier modules. They are connected as follows:

• RAL shaping amps I, II, XI, and IV are connected to top convertor box inputs 0->3, counting from the left when facing the rack
• RAL shaping amps V, VI, VII, and VIII are connected to top convertor box inputs 4->7
• RAL shaping amps X, XII, XIII, and "R" are connected to bottom convertor box inputs 8->11
• RAL shaping amps "S", "Q", "u", and "W" are connected to bottom convertor box inputs 12->15

In second electronics rack, "Rack B", at the top is a Nim bin used last July 2017 for the Si diode.

In this nim bin I installed two NIM-ECL-NIM converters and the PB-5 Pulser Generator

Currently in the Nim bin from left to right we have the following:

1. Silena mod 7710 Quad Bias Supply
2. Ortec 571 Amplifier
3. Ortec 863 Quad TFA
4. Ortec 935 Quad CFD
5. Two ECL-Nim-ECL Converters, Model EC1600 by EG&G-ESN
6. Lecroy Model 622 Quad Coincidence Unit
7. PB-5 Pulse Generator

The Nim bin was switched on in this configuration, I checked the supplied voltages with a voltmeter. All voltages, -6, +6, -12, +12, -24, and +24V read correctly.

Next I installed 5m 34 way - 34 way ribbon cables labelled 0, 1, 2, and 3 into the back of the top convertor box, including the grounding cable. Now:

• Cable 0 in back splits into cables 0 and 1 in front
• Cable 1 in back splits into cables 2 and 3 in front
• Cable 2 in back splits into cables 4 and 5 in front
• Cable 3 in back splits into cables 6 and 7 in front

Next I installed 5m 34 way - 34 way ribbon cables labbeled 13, 14, 15, and 16 into the back of the bottom convertor box, including the grounding cable. Now:

• Cable 13 in back splits into cables 8 and 9 in front
• Cable 14 in back splits into cables 10 and 11 in front
• Cable 15 in back splits into cables 12 and 13 in front
• Cable 16 in back splits into cables 14 and 15 in front

The other end of these eight 5m long cables are ready to be plugged into the EDI preamps

I then plugged 16-way ribbon cables into the Analogue Output of the RAL shaping amplifiers in the following order, where ADC ... is the label of the cables:

• RAL I -> ADC 8A
• RAL II -> ADC 8B
• RAL XI -> ADC 9A
• RAL IV -> ADC 9B
• RAL V -> ADC 10A
• RAL VI -> ADC 10B
• RAL VII -> ADC 11A
• RAL VIII -> ADC 11B
• RAL X -> ADC 12A
• RAL XII -> ADC 12B
• RAL XIII -> ADC 13A
• RAL "R" -> ADC 13B
• RAL "S" -> ADC 14A
• RAL "Q" -> ADC 14B
• RAL "u" -> ADC 15A

Some of the RAL "W" Analogue pins were slightly off center. Filipo needed to leave so I left cable ADC 15B unplugged.

I attach photos of the current status of both electronics racks

Checking SRIM 3.5 MeV 7Li is stopped in 9.06um mylar or 10.09 um CH2.

After the long run at 3.0MeV on 22/02/2018 we decided to put the 12um mylar back in front of the S2 and start tuning the beam for 3.45MeV 7Li. We also decided to tune the beam to a higher current of ~ 100pA to try and improve the detection rates.

At 20:54 The turbos were switched off. The manual gate valve for Ecodry 2, automatic gate valve for Ecodry 1, and line valves were closed.

Whilst the turbos spun down we removed the S2 preamp box from the top of the chamber.

At 21:31 the vent valve at FC-O was opened. At 21:49 we noticed the chamber was vented.

The S2 was removed from the chamber and 12um mylar mounted in front of it. The S2 was remounted in the downstream section of the chamber.

At 22:03 the manual gate valve was cracked open -> we began roughing the chamber with Ecodry 2. At 22:18 the manual gate valve was fully opened and we then opened the T1600 Backing valve allowing Ecodry 1 to pump the chamber. We then started the turbos.

During pump down the S2 preamp box was remounted on top of the downstream chamber lid. The lid to the preamp box was also mounted. The following pin diode preamp cables were reconnected: -15V, +15V, +ve test, input from lemo feedthrough, and HV.

By 22:34 the pressures in the chamber were CHUP = 1.3e-5mBar, and CHDW = 3.2e-5mBar. The S2 was biased to -130.3V, leak I = 1.54uA. The diode was biased to -70.1V, leak I = 0.00uA. The preamp power was switched on, correct currents were being drawn.

The target ladder was set to position 2, 80.024mm, no target or frame.

After tuning with no target, the GSSI slits were fully open at:

Up = 0mm

Down = 5.5mm

Right = 1.5mm

Left = 5.5mm

We found the following currents:

FC-4 = 740pA (upstream from bending magnet)

FC-0 = 430 pA (upstream from chamber)

FC-E1 = 70pA (downstream from chamber and GSSI magnet)

The target ladder was moved to position I, 95.010mm, blank 3mm frame.

After further beam tuning we obtain currents:

FC-E0 = 330pA

FCE1 ~ 200pA

Transmission: ~63%

At 23:33 CHUP = 6.5e-6mBar, and CHDW = 1.7e-5mBar. S2 bias = -130.3V and leak I = 1.66uA. Diode bias = -70.1V and leak I = 0.00uA.

We decided to call it a day. Both the S2 and diode was switched off for the night. The accelerator beam voltage ~ 1.4730MV.

Raffaele let a note for Liz to improve the transmission to 70% the next morning (23rd).

See next eLog entry #19 to continue the beam tuning adventure.