ITk Module tests in the single-module testbox

NEW 9Nov23, Juergen Thomas (jpt@hep)
UPDATED 16Oct24  
For: epldt126 (new PC)
Account: itkuser2

- Switch-on the small chiller under the desk (Grant EcoCool), it's
 set to 18C, no need to change.
- Install module in the testbox (make sure the
 power supply outputs are off !)
- Make sure the nitrogen flow is on. It should be. 
 To confirm, you can turn the white
 round dial to see the little ball in the flow metre. The flow metre
 has only 5l/min range, so our usual 8l/min means the ball is invisible.
 When you're sure, dial it back up again. The nitrogen flow is actually
 audible in the box. 
- Wait for about 10min, until the chiller is at +18C and the humidity is >5%
 The humidity sensor values are here, scroll down to 'Module Testbox - Arduino'.
 the 'NTC' is taped to the chuck:
  
 http://epweb2.ph.bham.ac.uk/user/thomas/tracker/trh/trh_bilpa.txt
 
 There is also a hand-held RS thermometer: T1 is the chuck (while T2 is taped
 to the coolant tube). Remember to switch off this thermometer, 
 otherwise the battery  will be empty next time you try (it has no auto-off).

- Switch-on LV power supply (TTI) and HV power supply (Keithley 2410).
- LV: power-up the Nexys (centre output of the TTI), wait until the
 row of green LEDs flashes intermittingly. Then power-up the module
 (left output of the TTI).
 Note: The HV is fully remote controlled, the front panel controls
 of the Keithley 2410 don't need to be and shouldn't be touched (!)

Before starting the DAQ, check in this file that the module is
correctly set in the DAQ configuration:

 gedit /home/itkuser2/itsdaq/itsdaq_v29Aug24_FMCDP_local/sctvar/st_system_config.dat &

The default is a LS module which has 'crossed HCC
bonds' (XHB). All modules like this have a sticker 'XHB' on
the transport box. Indeed all new modules are 'XHB'.
Only the uncommented lines are indeed
applied, namely this line, which is an LS with XHB:

 #        id CCR  s0  Name     (type)
 Module    0   0   0 JaneDoe0 Barrel_xHCC

The hybrid config now can be the same (!) for any module. Only
'LS' (one hybrid) and 'SS' (two hybrids), with xHCC or not,
needs to be adjusted in this file, see comments in the file.
Note all new modules are of type 'xHCC', 
meaning crossed HCC wirebonds.
Make sure only the lines for the hybrid(s) required
are un-commented, ie one for LS, or two for SS.

Now ready to start-up the DAQ.
Go to the ITSDAQ installation:

 cd /home/itkuser2/itsdaq/itsdaq_v29Aug24_FMCDP_local

Set-up the environment (this is required every time)
and start ITSDAQ:

 source setup_al9_local_root6.26.sh
 cd itsdaq-sw/
 ./RUNITSDAQ.sh

Check for possible error messages. Esp if you see 'No AMAC found !',
this is a problem and the tests can't continue. Ask experts
for help (eg Juergen, Karol).
Two windows will pop-up: 'ScanData' (which shows test results),
and 'BurstData' which contains the GUI interface to start the tests,
and shows data during the test. As a rule of thumb, 'something' ie
blue data entries need to be visible in 'BurstData' while running
the test. If nothing is there, again ask the aforementioned experts.
The Keithley 2410 HV will make a 'click' noise during this process,
this means it is in the correct 'REM' remote control mode.
Again, if it doesn't click, something is wrong (possibly somebody
has borrowed the adapter cable for the Keithley 2410). Ask the
experts then.
Compared to earlier installations, this extra step is now required.
Without doing this, the setup is not ready, namely the serial number
and the correct 'chip tunings' won't be applied:

 AutoConfig(false,false,false)

It will then query you for your production database login, usually
two passwords, but no login name.
If everything looks ok, namely the various ABC and HCC chip serial numbers
are showing, the test can continue. Otherwise, stop and ask
experts (eg Juergen, Karol).
The power supply should then show (left-hand display) 
about 300mA for an LS, about double this for an SS module.
An (optional) quick check to see if the HV wires are correctly
connected is to ramp the HV to just 10V:

 HVSupplies[0]->Ramp(-10,2)

There should be non-zero value showing on the Keithley 2410.

Next (regular) step is to ramp the high voltage 'at earnest' to
the nominal 350V which will be used in the detector, this is
done in two steps:
 
 HVSupplies[0]->Ramp(-200,25)
 HVSupplies[0]->Ramp(-350,40)

The beep from the Keithley 2410, the blue light on its front panel,
and the ramping values indicate that things are ok.
Its display will show the 'leakage current' of about 33muA 
when the ramping is finished.
Note that while having the HV on (blue light and a muA value
showing on Keithley 2410), 
the black box _MUST NOT_ be opened. This would be damaging to our very
valuble sensor, meaning the module may be broken.

Next step is running 'Full Test': 
Click 'Full Test' button in 'BurstData' GUI window of ITSDAQ under 'Test'.
This contains all the tests we require in the correct sequence and
the correct settings. The other options in the 'Test' menu can
be ignored here.
It takes about 10-15min. 
When done, ramp HV (IMPORTANT: Never switch-off
LV power while HV is 'ON' - blue light on the Keithley 2410)
This is done in steps (the first number is the applied high voltage, the second
is the 'compliance' current limit in muA):

 HVSupplies[0]->Ramp(-200,40)
 HVSupplies[0]->Ramp(-200,22)
 HVSupplies[0]->Ramp(-100,22)
 HVSupplies[0]->Ramp(-100,11)
 HVSupplies[0]->Ramp(-0,11)

Power-down the LV on the AMAC:

 AMACStar_DCDC(-1,0)

and leave ITSDAQ:

 .q

Now you can switch-off the LV supplies, first the module
(left-hand output of TTI) then the Nexys (centre output of TTI).
Don't swap the module while the Nexys board is running !
This would be a 'hot swap' with active connections.

And switch-off the chiller if there are no more tests to do.
Leave the nitrogen untouched at its current flow rate
(above 5l/min, ie the little ball not being visible).

The results are in:
 
/home/itkuser2/itsdaq/itsdaq_v29Aug24_FMCDP_local/sctvar/ps

and

/home/itkuser2/itsdaq/itsdaq_v29Aug24_FMCDP_local/sctvar/results

It makes sense to have a second terminal window open to 
check the results even while the test is still running.

 cd /home/itkuser2/itsdaq/itsdaq_v29Aug24_FMCDP_local/sctvar

 ls -ltr ps/* 
 ls -ltr results/*
 
The hybrid serial number is visible there, eg: SN20USBHX2001681.
(Note: If the files are called 'JaneDoe', the earlier 'AutoConfig'
step hasn't been done. Ask experts if this happens). 
 
The 'headline' test result is the 'full response curve' noise
test, which is the second of the two 'RCPlot' tests.

 ls -ltr ps/*RCPlot*pdf

Check esp the bottom-left plots of p.1 and p.3, this is the input
noise. This ideally should be continuous and flat at about
800ENC for an LS module. Outliers above are (likely) shorted
channels (bonds or on the sensor), noise of about 400ENC means
the sensor is not connected on this channel. Those shold only
be very few channels. 
The values from this test are here, the newest one is the
test which has just been performed:

 ls -ltr ps/*_RC_*txt

Finally, it makes sensor to zip-up the test results, only
the actual directory 'sctvar/results' is required, and copy
that onto /disk/moose/bilpa/atlas-itk/DAQTestsStar.
Use the module name in the zip-file. 
The 'Json' files from 'sctvar/results' contain all the information
and are uploaded to the database.