1 Latency and pedestal test Mikko Nykänen
2 Purpose Find latency that gives best pulse height spectrum and get close the best latency for this setup Also to get feeling to electronics Examine how chamber affects to pedestal Measure 3 setup for pedestal and ped. sigma
3 Info Test made Gas mixture ArCo2 Temp 22.8 C, Air preassure hPa, huminity 43% Detector triple-GEM, at 3.9kV voltage Used source, strontium
4 Setup
5 Changing latency Latency we change is delay detween write and tigger pointers So 1 latency is 1 clock cycle, 40MHz clock so 25ns
6 How test was done First I took pedestal run, put strontium source to unintrumented part of detector Next I set latency value, check register values in APV, checked the baseline and then took data run After data run I had to exit the program, set new name for apv ”run40x..” – Because program writes new test data after the previous one After data runs, I took a control run, so we can sure that baseline didn`t change during tests Last step was to process the data in ROOT and make a landau fit for PH spectrum, from landau fit I took MPV
7 Data Latency scan – 25ns, 50ns, 75ns, 100ns, 125ns, 150ns, 175ns, 200ns Pedestal and ped. sigma – APV without detector – APV with detector, but no HV – APV with detector, HV 3.9kV
8 In ROOT First we use ”line” to make ”filtered” datafile, this phase is done before entering root Next we run ”APV_an.C” macro and put data on the Tree and save it as a root-file Root-files then are run in ”ARCanalysis.C” which collects mean and sigma of the pedestal values from channel, values are saved in ”_Pedestal.tmp” ”totalcharge.C” makes PH histogram from values that are over ”channel pedestal + x times pedestal sigma” Next slides shows histograms we got
9 Channel 65 histogram
10 All channels histograms
11 Pedestal distribution of channel 65
12 Pedestal and ped. sigma
13 Pedestal subtracted spectrum
14 Latency scan Next we show Pulse Height spectrum with different latencys Following PH spectrum are made by sigma multiplier value 2
15 Latency 100ns
16 Latency 125ns
17 Latency 150ns
18 Latency 175ns
19 MPV Next slide MPV vs. latency Two peaks at values 3 and 6 latency? Four different kinds of PH spectrum made in every latency and MPV was collected Was done by changing range which defines if certain value is included in Total charge -> simply increasing sigma multiplier
20 MPV vs. Latency
21 Pedestal & ped. Sigma test To observe how chamber affects the readout by checking the pedestal and pedestal sigma value in these tree settings
22 Pedestal
23 Pedestal sigma
24 Pedestal & pedestal sigma Calculated mean values of both pedestal and ped. sigma No chamber, sigma 1.53 and pedestal Chamber, no HV, sigma 1.66, pedestal Chamber, HV 3.9kV, sigma 1,65, pedestal No difference when chamber has/or not have HV. But there is a difference if there is a chamber connected to electronics or not
25 Results Highest MPV is get with latency 150ns Pedestal avg. differs in pedestal test Chamber connected showed same behavior when HV on/off ”No chamber” is different from two other
26 Remainder These latency values ns were based on gut feeling when observing the display. So we can be careful to say absolutely this 150ns is the best latency -> solution is to make latency scan
27 Bugs & fix Problem before was baseline was in wrong place -> over flows at high signals – >Solved, by changing VPSP value in register VPSP change APSPs(FIR filters) voltage level adjust, digital filter Second problem, how to read multible APVs Solution may be in I2C bus, not sure but I2C bus automaticle define APVs that are connected in bus -> Can work in series, so cable that can connect multiple APVs could be solution Now trying to confirm that is the case
28 Step by step Make new set of measurements, best latency and better baseline Increace the baseline -> not many over flows anymore Revise the analysing program Make setup that can do X/Y- strips reading X/Y –strip scan/reading, 2 APVs Also start writing the user manual