-- Public.MarkWard - 27 Sep 2017

Previous Logbook


  • Website up and running - 99% thanks to Patrick, I mostly was present trying to debug based on what little I knew of m3slow.
  • Preparation work for cables for valve actuation, see Zac and Adriana
  • Installation of all safety hardware and valves on the HPTPC vessel
    • Gas inlet/outlet valve
    • Protective valve for vacuum gauge
    • 5 barG burst disk installation
    • 6 barG burst disk installation
  • Should likely check with Sam if it was Gauge Pressure, or absolute... if the operating pressure is gauge pressure then we can run at 5barA quite easily and still be within the rules. Scientists typically like to work in Absolute pressure... for safety devices however, industry standard is gauge pressure.
  • Solenoids for piping compressed air to our valve actuators have all been tested and operate.
    • Note, these valves are gas pressure return, not spring return. It means that we must have outlets connected to a closed system in order for the valves to successfully close
    • This is confusing initially as all valves iv ever worked with were spring return.
    • Safety wise it is fine as,
      • If a valve is open, and we loose pressure, the solenoid valve will remain open, and the valve will close as result of pressure lost (same as if spring loaded)
      • If we reset the system in the above configuration, with the solenoid unpowered (but latched open) the solenoid valve will reach 2bar and snap closed.
      • Our valve actuators move at 5bar+
    • State : Have Pressure, no power
      • Valves close
    • State : No pressure, with power, valve open
      • Valves close
    • Results -> Lose power, or pressure, valves close -> Safe
  • Re-routed compressed air delivery -> It was leaking (Who'd of guessed that!), it is now tidied up and looks good rather than looking like an explosion in a piping factory.
  • Tested all valves... they all work
  • Connected HPTPC to the Gas board
  • Connected an N2 Bottle to the gas board at V1
  • Tested putting 2 bar pressure into the gas board and purging it, isolating and venting.
  • Need to do some finishing touches to the electrical cables for the valves and some relabelling of channels on the netbooter and we are basically good to go for control of the valves big grin YAY


  • hpdaq and hpslow have a fresh install of SL6 to match m3daq and m3slow. Online and SSH accessable at hpdaq.pp.rhul.ac.uk , hpslow.pp.rhul.ac.uk respectively. Pat is undergoing the process of compiling everything and making modifications to point things at the correct addresses, bring up web-services etc (update) is is now serving... a page, Patrick is ironing out any issues
  • Hydraulic clamps are fully operational, no leaks thus far, pressure is at roughly 150bar in the lines.
  • Energize the left hand side to open, right hand side to close.
  • Extra force on the flange improves ultimate pressure from 1.7x10^-4 to 1.45x10^-4 mbar, occurred within about 1hr of clamping.
  • Leak rate/rate of rise improvement from 3.7x10^-4 mbar l / s to 5.9x10^-5 mbar l / s ~roughly factor of 10
  • Still need to assess if the vacuum seal is fully sealing
  • Investigating Readout/logging options for pressure gauges, WIKA (brand of our pressure gauge) have a RS232 addressable controller that has two channels capable of current loop operation (pressure gauge) and voltage monitoring (Vacuum gauge) - waiting on quote
    • Short term readout can be done via labview - we have a couple of different options to make things look presentable
  • Investigating use of a USB addressable Relay to control our valves. While the netbooter method works, it is overkill and we require more channels to actuate valves, it also requires an independant PSU per valve. I found a product for ~70 which will do exactly this, ships with its own front end for windows and Linux so in many ways can operate the same as we currently have but with 16 channels. Would require 1/2 PSUs to provide 24V power to it and all valves.
  • Waiting on Quote for Gas fittings, going to quickly scrummage around our lab to see if we can get something quick setup.
  • Need to make up a couple cables for valve acutation, minimum of 3, we already have capacity to do this though need a more perminent solution (see above)
  • Need to aquire a 27mm long socket for applying correct torque to the secondary clamps... DONE! Got the Socket, next for the torque.
  • Need readout method for our 4-20mA Pressure transmitter. We have possibiliy of 3 options
    • WIKA dual channel box, would be good as a final goal as it is RS232 addressable (like the DMTPC box) and can do both 4-20 and 0-10V input
    • National Instruments Card (already down in the lab, the DAQ card is in m3slow, command line addressing and m3 slow doens't have the software, doable requires a bit of moving computers around etc)
    • Borrow a NI MyDAQ from Lab, will allow us to run a screen with a big pressure display, can be turned around quickly.


  • Wednesday last week - Opened vessel, removed, inspected and cleaned the inner vacuum seal o-ring. Made notes of several problematic locations on the sealing flange.
  • Pump down was successful but reached a similar ultimate pressure as previously.
  • Thursday - Software face to face at imperial, outlined software tasks etc. My main short term goal is to get two PCs in place for the purpose of hpdaq and hpslow. - I have assigned rowan and lab-dm-009 for the task, and am finalising details of having independant control and new addresses for these PCs with PP-sysadmin, should be done soon and then we can do what we like with said machines.
  • Friday last week - installed a pump out line from the seal pump out port, set the system in motion
  • More discussion with workshop of parts for the field cage, hammered out a few final touches.
  • Minimum Pressure after weekend pumpout is 1.83x10^-4 mbar
  • Deavtivating the pump and opening the pump out port causes similarly large rise in pressure as observed before. So, still leaking badly frown
  • Pump out re-initiated - Ultimate vacuum 1.83x10^-4 mbar
  • Performed a 3 stage test
    • Isolate Pump out port
      • No change in pressure after 5 minutes
    • Isolate Whole system
      • Rate of rise 3.8x10^-4 mbar l / s
    • Pump on port but not the vessel
      • Rate of rise 6.9x10^-5 mbar l / s
  • Indicates that the impact of the silicone seal being the primary vacuum seal is a rate of about 3x10^-4 mbar l / s unless there is another o-ring that is causing an issue.
  • We have our hardware for mounting 2 rings on the test stand
  • We have our commando power plug - has been handed off for mounting (Along with Cable), need to poke Hakim to mount things
  • Iv left the two remaining hydraulic lines off so the cable can be easily mounted.


  • Turned around drawing updates for a smaller TPC volume (slight reduction of radius), communicated the changes with the worksop
  • More workshop communications and test fitting of components
  • Hydraulic pump for external clamps is waiting on its power plug - has been ordered, cable has been identified
  • Computing
    • Between Patrick and I, we figured out which copy of the DMTPC code is the most recent, turns out (somewhat naturally) that m3daq has the most recent versions of all the code except...
    • DmtpcControl, which is most recent on m3slow
    • We have now merged the two in an external source so that we can start correct version control
    • Identified some of the workings of the SQL database
  • Several rounds of leak checking etc... here is the fun part
Date Starting Pressure Rate of Rise
13 / 11 / 2017 2.06x10^-2 mbar 2.20x10^-3 mbar l / s
15 / 11 / 2017 1.69x10^-2 mbar 1.36x10^-3 mbar l / s
16 / 11 / 2017 4.76x10^-4 mbar 7.59x10^-4 mbar l / s
20 / 11 / 2017 3.27x10^-4 mbar 4.00x10^-4 mbar l / s
21 / 11 / 2017 3.37x10^-4 mbar 5.00x10^-4 mbar l / s

  • Some numbers and discussion
    • Turbo was added 15 / 11 / 2017 , type is a HiCube 67 l/s turbo molecular, reaching an ultimate pressure of 3.27x10^-4 mbar
    • An effective pumping speed of 33.7 l/s predicts a massive gas load of 0.011 l/s... Id not trust this... too many assumptions made about the pumping configuration
    • Opening the vacuum seal pumpout port causes a large rate of rise... approximately 3.2mbar l/s!!! <--- AAAGH!
    • Above suggests there is no effective seal between the Silicon (pressure seal) and the Inner vacuum seal. Sam sort of ummed and arrrd his way through telling us that during one of the phone meetings, I expected a partial seal, not an enormous leak like that... thats more than a virtual leak
    • Other option for our observed rate of rise is... Permiation
    • Silicon is the worse material to make a high vacuum seal out of due to Permiation constants being quite high.
    • Qperm = Kperm * Pressure (external) * Area (exposed to pressure)/ Thickness
    • Silicons N2 permiation is roughly 2.7x10^-10 m2/s
    • Put in the spec for our seal and you find a number roughly 6x10^-4 mbar l / s
    • To put it bluntly... it is likely the majority of our rate of rise is permiation through that seal. frown frown frown
    • Clear we need to make that inner seal work, or concider a procedure for backfilling which involves minimum time at vacuum.
    • Viton and Buna are factors of between 800-4000x better, which would put us into 10^-7 mbar l/s region, if that volume is pumped out, the issue reduces to zero
  • Workshop have been test fitting our components, I saw a few final products this moring big grin we should have multiple in our possession Soon


  • Complete DMTPC and HPTPC internal comparison drawings, for design consistancy
  • Adriana and I spent a full day preparing/repairing HPTPCs feet, removing burring on the screw threads such that the feet can be put down.
  • Learnt about being a contortionist
  • Levelled the vessel and adjusted the height of the end cap such that it can be married up to the vessel
  • Sealed vessel using screw clamps only
  • -> Hydralic pump is 3 phase -> in the process of aquiring a plug and cable for it to get wired up
  • -> outer hydralic line is the return (open) so for now that cut is less critical
  • Worked on friday with the team to do final cleaning of flanges and sealing of vessel for pumpdown
  • Friday pumped down (at 6pm) to approximately 600mbar -> pump overheat -> Suspected and confirmed from manual
  • Monday, DMTPC slowcontrol and DAQ online for Will and Kulveer, not ideal locations and need to relocate the NIM crate to give cooling to the digitizers.
  • Monday, completed pumpdown to 3x10^-2mbar by throttling the inlet of the pump and opening fully when at about 50mbar
  • Ultimate pressure after an overnight pumping period 1.7x10^-2 mbar, roughly matches the pump spec or what is expected
  • 3x10^-3 mbar l/s rate of rise from first test.
  • 1.4x10^-3 mbar l/s rate of rise second test after an over night pumpdown.
  • have attached Turbo from the one shooter camera - pump has not been used for a long period so needs to be 'dried/conditioned' cant reach full speed while isolated.
  • Found documentation for RGA, will prepare to use it/give a tutorial. The system is a pumping stand with independant RGA so will need a little read-up.


  • Set a new personal best, passed 1,104,516,000s breathing in and out... (thank you for the cake yesterday!)
  • Fixed a couple small bugs with rawAna on friday, does not require there to be triggers in order to export CCD images (weird logic mistake in one of the flags)
  • Made a new Drawings page for DMTPC -> its a good repository and source of drawings
  • There are a few different versions of the parts, although I have converted a batch of them into pdfs that i think are the latest/important ones
  • We have 3d models of some of the parts, these have been incorporated into our DMTPC + HPTPC drawings
  • A week of drawing and lab work planning (See Zac and Harrison)
  • Rails have been smoothed (a little) and ready to be re-mounted
  • Hydraulic pump is to be relocated on the windowsill, (to be done)
  • Made a drawing of DMTPC Amp Assembly with rail mounts, inside the HPTPC profile for comparison DMTPC_in_HPTPC.pdf
  • Made new rail designs for the support stand
  • SupportStand_Drawingv1.pdf
  • Machine Shop status
    • Machinest who works the CNC is back next week, lots of the prep work has already been done, (so lots of blocks have been cut etc) just the features need putting on them
    • Support stand parts have been pushed to the top of the list, so we can finish the stand and then start mounting other parts (when they come)
  • HPTPC arrival video - sorry - no music, but... like and substribe to my channel wink (shameless plug) https://youtu.be/bcbzPKV3uCg
  • Plan for the rest of the week is to prepare to close, we have i believe all the parts we need now, and can get going on it (Lifting hardware, lube for the legs... etc)


  • Prepared the lab for arrival of HPTPC, Involved taking apart DMTPC and making space for requipment.
  • Latter part of last week was a bit of a disaster having to make two trips to London to see the immigration lawyer
  • Got the vessel into the building! YAY, everyone knows what happened... it was an adventure.
  • CLEANING! lots and lots of cleaning, the vessel was extremely dirty on the inside (and outside)
    • Two soap washes
    • Two water washes
    • 1 Alcohol wipe
  • Put curtains back on clean tent (required cleaning also)
  • re-plummbed the hydrolic lines on the vessel, had to tighten a few other fitting
  • Made inventory list of parts we recieved from Cryovac
  • Notes
    • All gas regulators have TIPOC inlet fittings, we will have to change this for relivent DIN/CGA fittings for use in the lab
    • Captured o-ring kf fittings are good for 10bar, note, we appear to have no spares! Though we do have enough.
    • Vacuum valve... if that is rated to 5 bar id be surprised. Looks like some random valve they found somewhere in the bushes. Lesker Part number is SA0150MVQF, technically rated between atmosphere and Vacuum. Internal construction is probably capable of holding back 5 bar, but... yeah... dodgy.
    • Chamber has 3 missing caps - they are plastic caps that have been taped on. All are KF40s, I think these were the ones used to pressure test with the HV feedthroughs.
    • We will probably want to put a steel brush pipe cleaner through the water drain and air release valves (top and bottom penetrations on the vessel) Both show corrosion and are generally hard to clean.


  • Lots of meetings with Sam / Cryovac, good news is that the vessel should 'should' be here today, keep fingers crossed.
    • We aquired quotes from kingsbury for cranage in preparation for our own shipping options
    • Speciality Spanish Transport, quote was lowered and we accepted
    • Southern Cranes and Access gave an insane quote, I think this was in response to bad communications between SST and cryovac and themselves which basically involved SST demanding a quote without giving them any information about the Job. So... knowing or believing they had no options, they quoted 1500 ish... which Cryovac (as much as they are not in a position to complain really) identified as crazy.
    • We provided Cryovac with our quote from Kingsbury
    • Cryovac contracted Kingsbury
  • Got my simple diagnoistic tool working on m3daq
    • I made a sym link into /home/dmatter/bin (i think, i cant check, m3daq is offline for the delivery)
    • This means the program can be used from any command line location
    • Invoked by rawAna.exe - if given with no options it will print the usage directions and flags
    • For looking quickly at a root file and rumping the number of events/triggers point rawAna.exe at a raw root file with no arguments
    • -d dumps waveforms to screen ->really fast, much better than using a CINT macro
    • -c dumps CCD images
    • -d -c dumps both to screen
    • It can be used for offline analysis, currently it will do baseline cuts, peak time cuts, find peakheights for the charge channels and output 4 histograms. baseline sigma cut and time cut can be added using appropriate flags (see useage)
  • Wrote small script for exporting fits file using the DMTPC image library for Kulveer to use
  • Lab cleanup, tidy with the team, My task was mostly disconnecting and tidying away all cables and moving the DAQ computer and electronics rack.
  • Shoring props are in place, clean tent has not fallen over... yet


  • Practiced some mesh stretching, (mostly a helper with Zack and the team)
    • Lots of intersting things learnt
  • Started writing a tool for doing quick diagnostics on data files from DMTPC,
    • Very very fast at dumping waveforms to screen, which is initially the most useful use of it, to check what triggers look like
    • Optionally can dump CCD images along side the waveforms. So we can visually pair up waveforms and Events
    • Slowly updating it to perform baseline cuts given a list of files and generally have it do all of what we usually do in Macros
    • Motivation... CINT IS BAD AND SLOOOOW
    • Already this code appears much faster
  • TPC rail mount designs -> Not got a full assembly drawing yet, though have had a few communications with machine ship, the rails are advancing and hopefully should be out soon.
  • Reviewing Drawings with Zack.
  • Chamber Shipping Drama... mostly Zack doing the waving of sticks at Sam, Adriana speaking to various spanish transport companies.
  • I have spoken to the company wh lifted the DMTPC vessel, they are coming tomorrow to scope the area and will quote us for a UK lift.
  • c7eb6fc7c3dd4bdfc5dcd3f6c97209da.gif


  • Wrote a "How too and getting started page" for people to use to do basic dmtpc anaylsis etc it is located here
  • Took data sample for purpose of CR-112 calibration, See Adriana's analysis
  • Discovered Trigger issues with the DAQ
    • Appears to ignore trigger and pulse polarity settings (at least for the charge sensitive pre-amp)
    • Settings in the hptpc_neutron template (copied from one of the ones that shipped with the setup) has triggers on charge channels effectively disabled.
    • Thus it appears we are triggering only when we have a signal on the fast amplifers, ONLY negative going
    • Need to investigate this, possibility of hard coding or bug in the DAQ initialisation.
  • Helped (Tiny bit) with the Ring Turning -> See Zacks page
  • Helped (Tiny Tiny bit) with putting clean tent curtains up
  • Created a 50ohm terminator for SHV connectors for grounding the anodes
  • Grounded All anodes and ground meshes
  • Ramped voltage - breakdown reoccurs at same voltages (630-650v)
  • Cycled the gas, first pumping down to ~3x10^-6 mbar and then back to 40mbar
  • Channels now ramp to 670 without sparking, sparking comes in closer to 680V now
  • Cathode performance is identical, sparking occurs at 3200V
  • Slowly ramped up all channels to 680V, sparking is moderate, though approximately 0.1Hz so we can get a few events between sparks
  • Need to write some error handling into my code, currently segfaults if the first event has no triggers... Thus if the first event is after a spark, it sometimes contains zero triggers.
  • Going to migrate to Linappserv and make a set of compiled programs to improve analysis speed
  • Have 3 data samples in different configurations for analysis


  • Worked with Adriana on developing the analysis to accept multiple raw files
  • Can identify each peak in the spectrum based upon behavour of the CSP channel.
  • PeakHeights.png
  • hptpc_neutron00_000_018.pnghptpc_neutron00_020_078.png
  • hptpc_neutron00_047_024.png
  • All this is within a single run, a spark appears to trigger the change between each state. Looking through later files, the behavour returns to normal and then back to bad again
  • Applying peak time cut effectively removes this bad behavour from the files.
  • See Adriana's logbook for final plot showing combined files with baseline and time peak time cuts (Also preliminary gain)
  • Shakedown of Amplifiers was good, using the test input, scope traces looked identical for all channels.
  • 3 instances of bad grounding on the boards which has now been fixed. Extra insulation on the inside of the cases has been added.
  • Vessel backfilled with CF4 to 40mbar(30 torr), pressure reached before backfill was 8.44x10^-7 mbar
  • Drift field achieved and stable 3200V (compared to 2700V previously) Getting better smile can at do 100V/cm field at least.
  • Anode voltage frown Bottom left starts to spark at approx 0.5-1Hz when at a voltage of 650V, 50V short of where I usually try and take data to see how everything is behaving.
  • All_Alphas.png


  • Baseline Analysis,
    • Initially performed my own baseline extraction in order to learn a few things about the raw TFile structure
    • Rewrote the same analysis in the space of a few hours using the dmtpc framework
  • Process is to,
    • Extract the baseline from every trace and push it into a histogram for each channel
    • Fit gaussian/find RMT and define a cut of accepting only waveforms that have a baseline within 8sigma
    • Remove whole events that do not make the cut, Replot and refit.
  • BaseLines_bfandaf.png
  • Pulse height analysis
    • dmtpc framework is mostly defined in Pulse.hh (for anyone who is interested)
    • The getPeak() function extracts the peak and subtracts for baseline offset
    • Further acceptance can be performed based on getPeakTime()
    • Code (and a tutorial on what each part of my macro is doing what has been passed down to Adriana for further modification/development (getPeakTime() cuts which shouldn't take much development
    • Currently my analysis works only one a single file though it should be realitively simple to make it work on a string of files.
    • Quite slow... might concider a compiled version rather than a macro, shakes fist at CINT
  • PeakHeights.png
  • PeakHeights_Cut.png
  • Friday - Adriana and I did a full vacuum system and DMTPC operations tutorial,
    • Made modification to HV feedthrough, changing the connection type to a push on one
    • Applied Kapton insulation over any bare wire, the push on connector and length of the pin.
    • Proved the HV feedthrough operates up to the full capacity of the supply (8kV) in Nitrogen, no sparks or issues
    • Evacuuated vessel
  • Started Testing performance of the Cremat boards, Initially tests look like boards are healthy, but we have devised a full set of tests to be performed in order to check board health and operation of different Amps in each
  • Should be completed in next couple of days, if all checks are good we move onto the grounding and backfill the detector for sample runs
Topic attachments
I Attachment Action Size Date Who Comment
PNGpng All_Alphas.png manage 810.8 K 04 Oct 2017 - 13:07 MarkWard  
PNGpng BaseLines_bfandaf.png manage 94.9 K 27 Sep 2017 - 12:39 MarkWard  
PDFpdf DMTPC_in_HPTPC.pdf manage 120.2 K 08 Nov 2017 - 12:31 MarkWard  
PNGpng PeakHeights.png manage 36.8 K 27 Sep 2017 - 12:39 MarkWard  
PNGpng PeakHeights_Cut.png manage 15.7 K 27 Sep 2017 - 12:49 MarkWard  
PDFpdf SupportStand_Drawingv1.pdf manage 170.2 K 08 Nov 2017 - 12:54 MarkWard  
GIFgif c7eb6fc7c3dd4bdfc5dcd3f6c97209da.gif manage 376.9 K 18 Oct 2017 - 13:31 MarkWard  
PNGpng hptpc_neutron00_000_018.png manage 185.2 K 04 Oct 2017 - 10:39 MarkWard  
PNGpng hptpc_neutron00_020_078.png manage 186.8 K 04 Oct 2017 - 10:39 MarkWard  
PNGpng hptpc_neutron00_047_024.png manage 187.8 K 04 Oct 2017 - 10:39 MarkWard  

Physics WebpagesRHUL WebpagesCampus Connect • Royal Holloway, University of London, Egham, Surrey TW20 0EX; Tel/Fax +44 (0)1784 434455/437520

Topic revision: r19 - 13 Dec 2017 - MarkWard

This site is powered by the TWiki collaboration platformCopyright © 2008-2017 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding RHUL Physics Department TWiki? Send feedback