Page History

Introduction

The follow following discussions apply to the CSpad family of detectors, but all of the specific discussions and tests were done with the CSpad140k detector.  If, while reading this page, you would like to refer to the CSpad 2.3M or CSpad 140k configurations, an image of each configuration GUI screen is given at the bottom of this page in child pages.

...

The weighting function shows the response at various flux levels vs time.  These curves are for a 40 microsecond integration window.  The 40 microsecond digital integration window starts with the command to release the amplifier reset, at -40 microseconds in the above curves.  Since we now use the analog reset, several microseconds elapse from the time we command the release of the reset and when some response is seen in the curves.  In the discussion below we are talking about the digital window, but remember that the actual performance of the detector is determined by the curves show shown above.

Note also, that since the all the FPGA times start with a count of zero, we should add one to the settings before using them in calculations.

...

When the stale stand-alone evr issue raised it's head, CXI scientists scanned and then used these numbers.

500 400 + ((1102022580+1)*0.008) + ((280+1)*1.024) = 868.3912 microseconds, with a 40 microsecond window ending at 908.3912

...

The data readout time is related to "Read Clk Set", "Read Clk Hold" and "Row/Col Shift." Every block has to be read out serially: a total of 26x185 pixels for every pixel we have to clock 14 bits of data and do at least one row/col shift. So every pixel needs at least 14*(read_clock_set + read_clock_hold +2) + 2x2*(row/col_shift+1) + 4 clock cycles.

...

For the CSpad 2.3M, because the data is read out over copper LVDS links, we need to use a "Read Clk Set" value of 2.

Timing for the DAQ trigger

Both detectors need a DAQ trigger to label events that the DAQ system wants to save.  All the timing of the integration window position depend only on the run trigger or the arrival of the specified event code.  The DAQ trigger only specified that the DAQ system wants this frame, so ti's arrival time is not critical.  The DAQ trigger must arrive after the run trigger, but no more than 500 microseconds after it.  The software looks at the configured delay of the run trigger, and adds 250 microseconds to that for the setting of the DAQ trigger delay.  This scheme means that you set the run trigger delay to set the position of the integration window and the set the DAQ trigger delay to the same value, so the software will place the daq trigger in the middle of the window.