This guide may be slightly outdated (2016)
This page describe the procedure of bringing a single HBU to operation. Some very basic component specification is summarized by Mathias: Operation and Instruction Manual
Further commissioning steps are described as well. 2 manuals have been written in the past: Operation manual written by Oskar in 2013 written for the USB labview control and commissioning procedure written by Eldwan.
The HBU needs 3 modules for the operation: The Power board (POWER), Calibration board (CALIB) and Detector interface (DIF). These 3 modules are inserted in the Central Interface Board (CIB), which has 3 pairs of flexlead connectors, through which it connects to HBU (1 pair connects to each slab)
POWER, CALIB and DIF modules can be replaced by careful pulling of the module out without tilting. There are 2 different flexleads: signal and power and they are not interchangeable. The flexleads should be both oriented with the printed label towards the CIB, as shown at the picture TO DO.
The connector and the flexlead has to be aligned and then pressed. A click might be heard during pressing on each corner of the connector. The most common reason for fail is CIB vs. HBU misalignment. If the flexlead is connected / disconnected too many time, the contacts in the connector might get loose and the flexlead needs to be replaced.
The power suppies adn typical currents: +6V (0.6 ~ 1.4 A), +12V (0.1 A), High Voltage (Power supply must be set to 3 V above the configured voltage, +90 V max)
Boot LED blinks for ~1 s when the power sitches on. The LED indicates a phase where the firmware is downloaded from the flash to FPGA. If the light stays on, something is wrong.
Firmware can be update via Xilinx impact tool. TO DO
Easiest test if the DIF is working: Run the USB labview and check, if the "USB Open" indicator and "RESET_PROBE_REGx" indicators go green. The first connection after the first power on is not successful (the "RESET_PROBE_REGx" goes red). The Labview program needs to be stopped by the blue square stop button "Program Exit" and started again.
It is a good idea to test the POWER board first before connecting to the HBU. This only applies when the power board is replaced or reinserted.
Voltages generated by the POWER board can be measured on the CIB board on dedicated test points: TO DO The power board by default doesn't generate all power. Only the DIF is powered after first power-on. Volages at test points TO DO ....... must be 0V
Therefore the DIF communication has to be established and power has to be enabled (as described here) in order to measure generated voltages:
- TO DO button enables TO DO voltages
- TO DO button enables TO DO voltages
- TO DO button enables TO DO voltages
POWER board provide 3 independent high voltages. The selection of which will be used in each ASIC is given by the resistor configuration on the HBU (each ASIC can have different HV). The high voltage is never completely switched of: a pre-bias of ~10V is always present. The high voltage is enabled by ''Pre_Bias 2'' switch confirmed by the ''enable! set'' button in the ''Calibration'' tab of the labview, as shown at figure TO DO.
Slowcontrol files are text files, which contain text decimal representation of the binary configuration of each asic. Each SPIROC needs a separate file. The modification of slowcontrol files is performed via a labview program
Common settings for all modes
- Must be unique among all SPIROCs connected to single DIF! Otherwise data will get mixed up. It is represented in binary form with LSB on left and MSB on right. The DIF ID on the picture would be 1*0^2^ + 1*6^2^ + 1*7^2^ = 1+64+128 = 193
- Defines how long will the trigger circuit wait until the memory cell is stores the data. /!\ 3 fields have to be set together: ''Trigger'', ''!ValidHold'' and ''!RstColumn''.
- When this field is ON, then the TDC memory cells are used for conversion. When OFF, ADC_!LowGain is used instead of TDC. This is useful for intercalibration between HG and LG
- When ON: automatically decides which memory cell will be converted (HG or LG).
- Threshold for the autogain selection. The level needs to be set safely below the non-linear region of HG ADC.
Trig Ext (OR36)
- Enables a forced trigger generated by internal LED calibration system
- Defines the minimum signal size, which is needed for the trigger to be generated and signal to be stored. Typical value is around 230.
Lab mode (forced trigger)
The Trig Ext (OR36) should be on and 'DAC1 : Trigger should be at some high level (>500 is usually sufficient)
Testbeam mode (autotrigger)
The Trig Ext (OR36) should be off, DAC1 : Trigger should be set to some value obtained by the autotrigger threshold scan.
USB Labview control
3 indicators should green if communication through the USB is established properly:
USB Open, USB Init and
USB Open can fail in following conditions:
- HBU doesn't have power
- USB cable was not plugged properly
- The previous connection was interrupted and the power is still open or windows driver failed.
RESET_PROBE_REGx fails for the first time, labview program needs to be restarted. If it keeps failing, the DIF probably 'crashed' and HBU needs a power cycle .
There are 3 stages of power control: HBU2, 3.3V and SPIROC.
Power ON sequence
Power OFF sequence
Slow control files have to be prepared before and should be place in a dedicated folder. The expected filenames are:
SC_SP2b_ASIC01.txt SC_SP2b_ASIC02.txt SC_SP2b_ASIC03.txt SC_SP2b_ASIC04.txt. The folder path has to be copied into the SlowControl Folder Path.
Check if the High voltage is switched off before slowcontrol loading! The High voltage can be controlled from the calibration tab.
For single HBU should be the configuration left in defaults (as shown at the screenshot). Then press buttons in order
- SC_toDIF Write
- RESET_SC RESET
- Load_SC Load
Following buttons should be pressed or (sequence order is not important)
- 'Set in Delay Lines section
- DAC1 ON or OFF in the DAC section. ON is needed for LED calibration
- Set_DAC Write transfers the DAC bias voltage value to HBU
- Enable Set in the left bottom part, which transfers all the configuration from the Enable section. Configuration takes few seconds. Important enable switches:
- Pre_Bias_2 enables the High voltage to be applied on the SiPM, otherwise keeps just small prebias 10V
- TCALIB1 (5) enables the trigger for LED
- Trig_Ext synchr. (1) forces a trigger via external line. Should be on all the time except for the LED autotrigger threshold scan
LAB mode means to do measurements without particles. Just with internal LED system. Following settings are needed for LED measurements
- DAC must be ON
- DAC1 setting should be set to some reasonable voltage (>3700 mV for the new boards)
- TCALIB1 (5) must be on
- Trig_Ext synchr. (1) should be on
- DAC should be OFF
- DAC1 setting should be set to some low value voltage
- TCALIB1 (5) should be off
- Trig_Ext synchr. (1) should be on, since we want to force the dummy trigger.
There are 2 modes, in which the data can be taken. These modes differ in the run parameters. Filenams are given with the full path and without extension.
If the filename remains the same, the data is appended to an existing file, not overwritten. Data and Time is added to the filename in the Testbeam mode.
The LAB mode configures just the number of external triggers (No. Ext. Trig) and the distance among them (Distance_Trigger 2).
The Testbeam mode sets 1 dummy trigger only and the distance is set to minimum or to the value of (Distance_Trigger 2) in the LAB tab (depends on the labview version used). The Meas.-Length defines the maximum time, for which the acquisition is allowed.
The resulting length is calculated and transfered to the HBU when Def_Meas NOW is pressed. If the duration is longer than 16 ms, Bunch-X will start counting from 0 each 16 ms.
Validation should be switched on in the Testbeam mode only when there is an external lemo signal connected, or when the LED light is recorded in the autotrigger mode. Internally generated LED triggers are validated automatically.
InfoCycle defines how many readout cycles will there be among the temperature readout. The temperature is saved in the text files.
The gain is defined as a distance between 2 photopeaks in the single pixel spectrum. The gain is an individual property of each channel and each gain mode (high gain and Low gain). We assume, that the gain is same for all memory cells belonging to a single channel (and gain mode). However the pedestal depends not only on the channel and gain mode, but also depends on the memory cells, therefore data from all memory cells can not be simply merged together. We also assume, that the signal in memory cells does not deteriorate too much during the acquisition time.
SPIROC waits fixed amount of time from the moment, when a trigger from fast shaper is detected and sampling moment. This time is called hold value. This value needs to be set experimentally. To do so, a set of slow control files needs to be prepared. There is actually a generator in the slowcontrol editor, which creates the set automatically.
There might be a difference between a hold value in autotrigger mode and a hold value in forced trigger mode (LED Lab mode), therefore the data have be taken in appropriate mode.
The default hold value is
The ratio between high gain and low gain values need to be properly calibrated, if data from both gain modes need to be combined (operation in autogain mode). The gain can not be usually extracted directly from the low gain mode, because the resulting Single Pixel Spectrum does not see the individual peaks. Therefore we calculate the gain only for the high gain mode.
The SPIROC allow operation in a special mode, where it converts and sends data from both memory cells, from high gain and low gain. This mode can be configured by the slowcontrol Switch TDC option set to off.
This way we record exactly the same events in high gain and low gain memory cells and we can determine the slope from the ADChigh gain vs. ADClow gain plot.
Autotrigger threshold scan
Autotrigger threshold is set for each ASIC individually and is common for all 36 channels. Channel-wise trimming is possible, but it affects the global threshold and never been tested in beamtest. The autotrigger threshold can be measured in 2 modes: in forced External trigger mode and in autotrigger mode. Both modes should be used. The threshold typically lies between 200 and 240.