Muon Unscripted Testing: HIFI

Introduction

These tests looks at data from the HIFI instrument. The first test will introduce dead time corrections. When a detector receives a signal, there is small period of time afterwards when it can not record a new signal. This is called the dead time. We can model and adjust the data to compensate for the dead time. When looking at the raw counts, which look like an exponential decay, the correction makes the biggest difference at large count values. The second test will use multiple periods, this is when the same data is recorded twice at slightly different time intervals.

HIFI Transverse Field Simultaneous Fitting

Time required 5 - 10 minutes

  • Open Muon Analysis (Interfaces > Muon > Muon Analysis)

  • Change Instrument to HIFI, found in the Home tab

  • In the loading bar enter 134028-39

  • Some new data will appear in the plot

  • Go to the grouping tab

    • In the group table, tick the Analyse option for fwd

    • In the pair table below, untick the Analyse option for long

  • Go to the Corrections tab

    • Set the plot to Counts (combobox above the plot), you will now see an exponential decay

    • Change the dead time correction to None and notice that the counts at small times decreases

    • Change the dead time correction back to From Data File

  • Go to the Grouping tab

    • In the Pair table, click Guess Alpha

    • In the resulting dialog, change the run to HIFI134034 to be used for the calculation

    • A value close to 1.3 should appear

  • Go to the Fitting tab

    • Check the Simultaneous fit over checkbox, and change from Run to Group/Pair

    • Right click the empty table area; Select Add Function

    • Add a FlatBackground (Background > Flat Background)

    • Similarly, add ExpDecayOsc (Muon > MuonGeneric > ExpDecOsc)

    • Set all parameters to Global, except Frequency

    • Click Fit

  • The fit might work but with a large Chi-squared Squared value (>100)

  • Now to try the fit a different way.

    • Click Undo Fits

    • Click the value for the Frequency parameter; A ... should appear next to it, click it. A new window should appear

    • Enter a value of 0.01 for the first run in the table

    • For each of the other runs in the table, enter values from 0.1 to 1.1 in steps of 0.1

    • Click Ok

    • Click Fit

    • This time the fit should work with a significantly lower value for Chi-squared (<10)

  • Tick Tiled plot by in the plotting window

  • Change the combobox to Run and each subplot will now contain just 2 lines

  • Check that the lines look similar to the results below

HIFI-TF-Result.png

HIFI MultiPeriod Data

Time required 5 - 10 minutes

  • In the loading bar enter 84447-9

  • Go to the Grouping tab and press Default

  • Go to the Fitting tab

    • Untick Tiled plot by in the plotting window

    • Untick Simultaneous over

  • Go to the Sequential Fitting tab

    • Press Sequentially fit all

    • When its complete the table should update

    • It is ok if some of the rows say “Failed”. This happens because these rows need more iterations to converge. However, if you go to the Fitting tab and cycle through the datasets, you can see that the fits are still reasonably good.

    • Selecting rows will show you the data and fit for those rows

  • Go to the Grouping tab

    • The Period column in the group table will show a series of numbers (1, 2, 3 or 4)

    • Press the Periods button and a pop up will appear with 5 rows (one spare)

    • Close the pop up

    • Tick the Analyse box in the grouping table for “fwd1”

    • Untick all of the Analyse boxes in the pair table

    • This will leave you with 3 lines (one for each run)

    • In the Periods column you can sum periods byusing a comma seperated list or a range using a dash

    • When you change the periods summed the plot will change