CARME pressure moved to

https://uoe-my.sharepoint.com/:x:/g/personal/cbruno2_ed_ac_uk/EVbsVOubNHRLj0PKiFIvd00BClMJ6WSJ1UDAyLA2F3R4jA?e=hEoZt7


Some photos overnight are over-exposed and pressure cannot be read. 
Might be able to fix with image processing if needed, but likely not worth the time. 

Baking started: 23/09/2021 14:00	
Photo frequency changed to once per 10 minutes.

Vacuum group confirmed temperature limits for the RGA of 200°C without the electronics removed and 80°C with electronics installed.

Ion pump, RGA and additional gauge installed on the chamber.

Our gauge is the one mounted above the NEG feedthrough, Michael's gauge is mounted below the feedthrough. Ion pump too large to install on the top of the vessel, sufficient room on flange immediately next to this top flange.  

Baking frame assembled around chamber, all required components fit within the baking volume. Insulation on baking tray and heaters set up and connected to pi controller. 

Pumpdown begun at 16:30. Turbopump on at 16:40. Unable to reach the required pressure for the hot cathod gauge before leaving the lab ~18:10. Suggests possible leak, pressure and leak check tomorrow morning.

MACOR strain relief and plastic strain relief tested in edinburgh compared (see image). MACRO hole size is 2x bigger than the test piece, suggests mix up between radius and diameter for MACOR strain reliefs. 

Accessed Pi remotely 

( 1. Without the Internet, the Pi has no way of knowing what time is it.
In case of power loss make sure to set correct date via

sudo date -s 'YYYY-MM-DD HH:MM:SS'

Or all log files will have wrong timestamp )

EDIT: Pi now connected to GSI time server.


2. Temperature logging OK


3. Commands/events logging enabled. Logging to file Log.dat, automatically moved to Logs/$(timestamp)_log.dat on
gracefuly quit.
Cannot test due to TC08 / RE72 not being connected. Please test when possible.


4. Plot_temps.gp gnuplot script currently optimised for 1080p monitor. Monitor in use is 1024x768. 
Please optimise first line of script (size, position) to make it fit while we search for a better monitor.


5. Hostname changed to CARMEbakingpi

Laser surveying team worked till 11. Chamber was too high by 3mm close to the jet target and 6mm by the turbo pump (towards the dipole). 

CARME lowered appropriate amount using bolts under CARME feet. Height checked using a laser on the centre point line of the middle flange turbo end and mid point on the front of the chamber (see images). During lowering front foot swapped and valve supports moved slightly to give more clearance for opening. ~13cm clearance for opening the chamber currently. 

Pipe and single bellow connecting CARME to dipole replaced by two bellows and new pipe. Rest of open ports on chamber closed. 

Thermocouple feedthrough flange installed under the motors flange in the turbo pump section. One internal thermocouple hooked onto kapton cable connected to MACOR and one thermocouple left floating in the chamber. Floating thermocouple corresponds to right output on feedthrough, thermocouple on kapton corresponds to left output on feedthrough. Image taken from flanges at the top of the chamber, quality is poor due to lack of maneouvering space.  

Baking frame partially assembled around the chamber. Full assembly requires removing valve airlines to get under the valve pipe, will be done Monday morning. Took measurements for the RGA around the chamber, should be able to fit in multiple positions without conflict with baking jacket. 

Chamber has been vented back to atmosphere. 

Motor was able to be pulled back to insert bloacking pin once at atmosphere.

Thermocouple feedthrough constructed with 4 thermocuples attached. Internal thermocouples each ~200mm in length with external thermocouples each ~1m in length. All were tested by plugging into TC08 and using baking code. Feedthrough will be installed under the motors flange tomorrow. 

Survey team come tomorrow morning, afterwhich ion pump and RGA will also be installed. 

Found CARME pressure at 2E-7 mbar

Raspberry pi took pictures until 4 am and 
then lost contact with camera. Reasons 
unclear. Plugging camera in / out solved 
the issue.  Will have to be investigated 
if it happens again.

Left list of things to do with RSS. In 
particular:
- Install xhv feedthrough and 
thermocouples. Prepared on desk
- Install blocking pin in right hand motor 
after pressure returns to atmosphere
- Cover bellows with aluminium to protect 
from fibreglass when mounting tent 
eventually
- Check all parts available for detector 
supports with mounting sketch Peter sent ( 
RSS to recover sketch)
- Investigate baking program logging. 
Unclear if logging is working - did we 
disable it for some reason? 

Valve on scroll closed. Scroll off. Turbo 
off. Pressure sensors off. Raspberry pi 
off.

Activities to resume on 16 September.



Attached are the pumping down plots
start point:  10.09.2021    19:00 
end point:    11.09.2021    16:00



10.09.2021    19:00    1.09e-6
10.09.2021    20:00    7.57e-7
10.09.2021    21:00    6.17e-7
10.09.2021    22:00    5.36e-7
10.09.2021    23:00    4.82e-7
11.09.2021    00:00    4.42e-7
11.09.2021    01:00    4.10e-7
11.09.2021    02:00    3.84e-7
11.09.2021    03:00    3.62e-7
11.09.2021    04:00    3.43e-7
11.09.2021    15:00    2.29e-7
11.09.2021    16:00    2.31e-7

Leak tested CARME via He leak tester + He 
bottle. No leaks could be found. Some 
indication of potential leak close to 
blind flange on bottom valve - disappeared 
after closer checks.

Outgassing from cables? If so second turbo 
may be required for baking.


Removed leak tester. Scroll on, turbo on. 
Pumping down overnight.


Tested baking system. Missing:
- Monitor (not shipped from DL??)
- Monitor power cable
- Power cable for USB hub (not shipped??)

Could find the first two, but monitor is 
of poor quality.

USB to RS232 cable damaged in shipping 
(poorly secured). Still works, may be 
worth repairing but not trivial. Only 
works plugged in to top USB 3 port on Pi.

System tested and still working aside from 
these issues.


Connected webcam to pi facing IE514 
display. Will take pictures every hour on 
the hour as in DL.

Will stop pumping tomorrow pm due to 
imminent shutdown.

Pressure at 5E-6 mbar decreasing.

Pressure found in 10-7 mbar range.

Decided to leak test anyway for safety.

Closed top valve. Vented turbo and backing 
line. Installed additional valve and T 
piece.

Pumping down again. Will begin leak 
testing soon.

Moved cables that touched neg pumps. See 
photos. Please bear in mind perspective.
Note avoiding touching chamber 
bottom is not possible.

Re-installed motor to fix alignment 
issues. All good. See attached.

Started pumping down. Reached 3E-1 mbar in 
20 minutes (ok). Turbo on. After one hour 
turbo stable at 0.85 A, pressure on Pirani 
at 1.1E-2 (not reliable). 

IE514 doesn't report pressure, i.e. > 1E-4 
mbar.

Leak likely. Will try to pinpoint 
tomorrow. We only changed 10 flangese.

Continued to mount flanges as per previous ELOG entry.
Peter also remarked that the wires look a bit loose inside the holes, although this is how they were during
dirty tests in Edinburgh.

Attempted to pack the wires in two holes rather than 4 - too tight. Cannot close.
Packed the wires in 3 holes instead. Better, but still a bit loose. Wire harness feels quite robust and no
obvious strain is placed on the connectors.

#1 flange left with four strands for now.
Connectors were tested in this way in Edinburgh for motion tests. No broken cables after several thousand
repetitions.

Had troubles with second flange being mounted. Glove was caught under MACOR connector. Managed to free it
somehow with a cleaned screwdriver. Other half of the flange refused to go in due to bent pins, until NB was
able to repair it.
Mounted as #5.

Most other connector pins OK. Had to adjust a very small number. Cause of damage unclear - transportation? See
photos.

Took some pictures inside the chamber. Most connectors look OK, but

- One harness is touching the NEG wires (#2). This could be an issue. Will try to move it tomorrow
- Wires hanging from the top may be putting a lot of weight on the connectors. See pictures. Given number of
wires this may be acceptable.
- One harness is touching the NEG cage. Definitely an issue. Will fix tomorrow

Top/bottom flanges mounted with short harnesses. Side flanges mounted with long harnesses.

Two flanges left to mount.
Only #1 has been tightened. Other flanges left finger tight for Grossmontage to tighten properly tomorrow.

Outside motor to be fixed tomorrow.

Assembled feedthrough flange with re-cut 
SS support bracket. M3 vented on the 
sides, M3x30 partly threaded on top. Top 
screws don't touch the 
Macor (by design? See first attachment). 
Macor affixed but has 
some play - good for thermal dilation.

Mounted long wire harness (side flange) on 
top. Unable to get one of the two 
connectors fully in. Unable to get it out 
as well. See last attach, it's a small gap. 
Tested few pins with multimeter - all ok. 
Decided to mount.

No issues mounting. Slowly and carefully 
threaded wires harness in CARME taking 
care not to hit negs.

Slightly concerned about weight of Macor 
connector on wires, but no obvious 
alternative.

Please have a look at attachments and 
comment. Will keep mounting in the 
afternoon.

Re-drilled SS supports in order to fit M3 screws.

Used hand tool cleaned with alchool for 6 of the 8.
Had to use machine for another two with thread damaged.

All 8 will be sent for cleaning. Should be able to start mounting at least the 6 ones tomorrow.

Reason for lack of fit still not clear. They were tested in Edinburgh. Possibly result of high-temperature baking?

Rudi and survey group representative came 
to CARME. 

Elephant feet confirmed to be out of 
position with respect to the models. Still 
unclear if CARME is where it should be 
with respect to the jet target and 
electron target. I.e. do they have space 
to mount? Do we have to move CARME again 
(hopefully not!!)

Valve to downstream confirmed out of 
alignment. Solution unclear but may be 
needed to ensure beam goes through 
detectors correctly.

Survey group will come to confirm CARME 
position on Thursday 16 September. We 
cannot mount baking tent until then, but 
no major issue given delays due to strain 
relieves and shutdown on 12-15 September.

Tested every strain relief. M3 screws get stuck at different points in each hole. Only one good hole out of 16
(2x8).
They were tested in Edinburgh - unclear if this is a result of baking at high temperatures, or bad screws, or
something else.

Easiest solution may be to re-machine holes and clean the connectors here in GSI.
Now investigating options.

Mounted both  motors on CARME. Started 
with outside motor.

M6x160 screws proved slightly too long to 
fix bellows to rod and seal the vacuum. 
Had to be manually shortened by 1 cm 
roughly.

Mounted whole assembly but could not fix 
actuator rod to moving plate due to holes 
mis-alignment.

Mounted motor on CARME but could not align 
holes on flange and chamber without 
turning motor to the side... see attach 3 and 4


Got in contact with Peter over lunch and 
understood issue.

Mounted second motor with experience 
acquired. No issues. All screws go in. 
Motor box level. See attach 5

Will need to dismount / remount first 
motor box before pumping down. Currently 
secure on CARME thanks to adjustable 
support.

Note we forgot to buy DN 40 CF gaskets. ML 
supplied a few. Will buy more. Needed 
between bellows and rod.


Strain relief mentioned in previous ELOG 
entry was taken to workshop and hole re-
drilled. Workshop reports other hole is OK 
for M3 screw. Either screw used or hole 
were poorly machined. M3 screws should be 
correct. Re-machined part sent for 
cleaning.

Will try with clean parts again tomorrow 
unless we remount motors.

Activities for the day cut shorter due to 
SPARC meeting this week.

Prepared the motors to be mounted and 
found all components required. No issues. 
May have to partly dismount again to fix 
bellows depending on ease of mounting.

Note extremely long screws mentioned in 
previous Elog inventory are to be used for 
the motors.

Attempted to swap Pirani gauge with ours 
but realised it's CF and needs appropriate 
port without risk of baking. That port 
would be on the interaction chamber which 
is not going to be mounted. Will probably 
keep using ABD gauge for now.

Peter spotted mistake in mounting front 
foot of CARME which is back to front. 
Correct mount will gain us 2-3 cm but will 
conflict with current CARME position fixed 
by tube on the back. Will discuss options 
with ML.

Motor mounting to take place on Mon with 
local technical support. No activities 
tomorrow unless a solution for the clamps 
mentioned in previous entry is found (i.e. 
what size bolts go on top). Peter will 
check step file and we will consult with 
local engineers on Mon pm.

Attempted to mount cable clamps for strain 
relief on inner part of Allectra 
feedthroughs.

SS support mounted with M3 vented screws 
OK
Cannot figure out which bolt type to use 
to fix Macor on SS support.

M2 falls through

M3 screws in partly and gets stuck...
Will need to figure out how to get unstuck

No M2.5 bolts available.

In contact with Peter. Not clear what the 
solution was

Will try again tomorrow or move to motors 
first.

In order to install the detector supports 
and eventually install the detectors 
themselves CARME must be split and 
separated via rails.

However it does not look like we have 
sufficient clearance on either side. See 
pictures attached.

I suggest we mount the strain relievers 
wires and motors, not split CARME, and 
pump down + bake to give us time to figure 
out what is different from the models and 
how to fix it. Outgassing from detector 
supports should be negligible wrt wires.

One possibility would be to install rail 
extension on the back side of the Al frame 
where the cross was meant to be supported, 
temporarily remove the tube connecting 
CARME to the YR09 valve and slide CARME 
back. Will need to discuss with Peter.