Mon., Oct. 31st

Need to change my FAMMail password again soon.  Remember this time to also change it in Gmail!

Ray called Mentor about the status of the donation and they asked him to forward the licensing agreement to someone who has legal signing authority on behalf of the university.  Ray is sending it to the General Counsel's office.

Ray said I can go ahead and start testing coincidence detection.  His pulse height peak in the current configuration (90 degree overlapping orthogonal paddle ends, 30" apart, 3 ns arrival time filter) was 241 mV for the paddle with the higher peak, and 236 mV for the lower peak.  (Bias voltage still 1,200 V.)  He suggests we set the first threshold at 200 mV.

I should perhaps change the firmware to count and periodically report the number of non-coincidence peaks, so we can see how many of them we are skipping.

I could perhaps integrate the (56-bit) changes by Juan and Michael Dean into the Quartus design before I test my code, but perhaps I'll wait; it might be better to just test one thing at a time.

Started "Nios II 9.1 Software Build Tools for Eclipse;" switched it back to my main workspace for the FEDM project ("Q:\eclipse_workspaces\mpf_workspace").

Opened pulsebuf.h and uncommented "#define FILTER_COINS"; this enables compilation of the code for coincidence filtering.

Edited pulsebuf.h and pulsebuf.c to define macros for normal & error return codes from pb_get().

Modifying server_stream_pulses() in pulsebuf.c to just count up the non-coincidence pulses, and output the counts periodically.

The guys helped me hook up the paddles on opposite sides of the room and tested.  That is working now, as far as it goes, except that for some reason I am seeing isolated pulses reported as coincidences - this should never happen, right!  Anyway, debug later.

Got David & Darryl started on calculating the standard deviation of the triangle distribution representing the analytical distribution over actual pulse widths for a given measured pulse width.

Found the bug in my coincidence detection code - forgot to update last_pulse_time[] when pulling pulses from the queue.  That's fixed now.  Started an overnight run.