To do

Abstract or introductory sentence on procedure

Outline of steps in the procedure

List of outstanding tests, if any, to be performed to optimize this procedure and a proposed schedule for the tests

List of online monitoring diagnostics which are essential for this
Indication of which exist and which are new programming tasks

List of any offline programs required for these procedures:
Indication of which exist and which are new programming tasks:
Identification of the authors of the code

Identify any DAQ development required for these tests:

LI pulser and cable tests

Intro:

Need to test pulser boxes and light injection fibres bundles upon arrival at Soudan
Pulser boxes are installed, one in each electronics crate on the middle level, in the space above the Wener power supply and below the RPS unit
Light injection fibre bundles are installed to the planes before each plane is lifted, (see installation document written by Richard White, Sussex for more details)

Testing of pulser boxes involves:

checking that electronics boards inside pulser boxes have not been dislodged in transit
checking that all LEDs are working
checking for broken fibres
measuring light emitted from each fibre in each connector at the pulser box via 1m test fibre bundle

Testing of fibre bundles involves:

checking for damaged fibre bundles and broken fibres before installation
measuring % transmission of fibres compared to a 1 metre test fibre (from pulser box tests)

All measurements are made using Sussex testing apparatus.

Outline:

Sussex Test Apparatus

Consists of a set of 20 PIN diodes, calibrated to have identical responses, connected to a single amplifier.
A fibre bundle is connected at one end to the pulser box via a 20-way connector and at the other to the box containing the PIN diodes. The fibres have metal connectors at the end and can be plugged directly into recesses in the PIN diode box, one fibre per PIN.
Only a single fibre is pulsed at a time by the pulser box, (a result of the fact that only a single LED can be pulsed at a time in the pulser box), and so only light from one fibre at a time is amplified.
The signal from the amplifier is sent to a picoscope which has a PC interface. The picoscope is triggered by a signal from a BNC connection on the front of the pulser box. This sends out a constant 4V signal when the box is on. When the pulser pulses the signal dips to below 0V for a short time.
The PC controls both the flashing of the pulser box and the analysis of the pulses as sampled by the picoscope. A set of three LabView programs can be used for controlling the pulser box and for sampling the pulse heights of the signals measured by the PINs, (see separate document for use of the LabView programs).

Testing a Pulser Box

A pulser box contains 20 LEDs and is controlled via an RS232 cable from the serial port of a Laptop or over ethernet via an EDAS box. The boxes have 64 regular connectors, each with 20 fibres, one from each of the LEDs in the box. There is also one spare connector and two connectors for PIN fibres that go to PIN diodes on the front end electronics boards mounted on each MUX box. The fibres in the PIN connectors are made to be significantly brighter than the other fibres, as PIN diodes are not as sensitive as PMTs to low light levels.

Using an Allen key the front of the pulser box can be opened. Open carefully as there are wires from electronics boards within the box soldered to the front panel. Check that the electronics boards have remained firmly in place after shipping. If not, reseat the boards.
Remove the blacking connectors at the back of the pulser box and use the LabView sequential flasher vi program, to check that there are no broken fibres in the box and that all the LEDs are working. If there are broken fibres, contact Phil Harris for advice.
If the same fibre position in all connectors always looks dark, this implies an LED problem. Burnt out LEDs can be replaced inside the pulser box. Again if this has to be done, contact Phil Harris for advice.
To test the light output from each fibre from each connector, connect the 1 metre test cable to one of the positions on the back of the pulser box and connect the single fibre ends to the appropriate position in the 20-PIN diode array, (always connecting fibre 1 to PIN 1, etc. removes any possible error due to variations in the PIN diode responses. This is important later when measuring the transmission of the fibre bundles compared to the test fibres).
Run the appropriate LabView program using the settings stated in the LabView document and save the text file produced. The file lists the average value for the peak of the pulse in millivolts as measured by the picoscope, for each LED, i.e. for each fibre in that connector.
Once all connectors have been measured, use root macro to write all the information into a tree.

Testing Fibre Bundles

Each fibre bundle consists of 20 clear fibres of varying lengths with a 20-way connector at one end and individual metal connectors at the other. The 20-way is plugged into the back of the pulser box, to the spare connector and the 20 fibre ends into the PIN diode box ,(fibre 1 to PIN 1, etc).
Use isopropyl alcohol to clean the 20-way connector and it's mate before connecting the bundle at the pulser box end. Also dab the tips of the fibre ends before inserting them into the PIN diode box. Only use special lens cleaning papers for cleaning.
Unlike in the case of the test cable, the connectors on the individual fibre ends are not designed to fit perfectly into the PIN diode box, so ensure that the fibres are pushed in as far a they can go.
Once again the LabView program can then be run using appropriate settings and a text file saved for each fibre bundle.
The data for each bundle can be added to an existing tree containing all the values for the % transmission for all the LI fibres.
ENSURE THAT PULSER BOX DOES NOT CONTINUE TO PULSE AFTER TESTING. THIS CAN CAUSE PREMATURE LED DEATH! If blue LED on front of pulser box is on when you expect pulser box to be off, hit the red reset button on the front panel.

LI cables hook up

Once a plane has been raised, need to uncoil the LI fibres and securely rout them to the appropriate pulser box, (each pulser box serves one side of 64 planes). The mine crew will be trained to undertake this task and also to connect the 20-way connectors to the pulser box. Grease will be used when installing to help reduce the transmission differences observed over a connector.
Phil A. expressed the need for recording the installation date of each fibre bundle to the pulser boxes. Mine crew will be asked to fill in the installation date for each plane+side on a form which can then be entered into a database somewhere (to be done by Robert Hatcher, Phil Adamson...?).

Also need to install PIN fibre bundles and trigger PMT fibres, see separate sheet for instructions.

Unsparcified running at low light level, one pmt at a time

Using liMaster operating under run control, can define a runType to take this data and to set light level, repetition rate, etc. required.
Some tests required to determine what these setting should be; need to wait until step dowels have been installed before these can be carried out.
RC config files need to be edited to accommodate new runType(s).

Sparcified running, step through the light levels

liMaster already set up to take a full calibration run using runType "LightInjection". Refinements to number of points to take on gain curve and pulser settings to be used may have to be made, (if someone hasn't already decided this.... ).

Test LI mapping

Need to understand what information is currently in the LI summary files... If we know which LED is being pulsed on which plane we know where we expect to see a signal in the electronics. If the signal pattern or mean looks unreasonable, it may be that the LI fibres were installed incorrectly. Use highest light level data from gain curve mapping to investigate this.

Online tool: May be useful to have a canvas displaying histograms for all electronics channels for a single plane (2 planes,...?), showing only LI hits from a particular LED and highlighting pads where the signals are expected to be; or a macro which looks for deviations in the checkerboard pattern for quick checks. Fingers crossed we don't find any problems... retro-fitting is painful!

Test conductivity

Non-linearity calibration