20220801
========
setup 2 on pulse generator: Sinuswaveform
input on even channel, highest gain (3 = 27mV/fC), longest shaper (0=120ns).
Amplitude/frequency adjusted to give all values between 0 and 1023 (pulse saturates ADC).
Free running (internal trigger on pulse generator, i.e. unsynched with ADC clock). 
Purpose: to see how the ADC values are populated, ideally a smooth distribution.

Run with 10000 events.

Single event:
plotchip-0-7-2-38.pdf   single event with all channels
ch8-7-2-38-full.pdf     single event for one channel
ch8-7-2-38-zoom.pdf     single event shows detail of pulse

Analysis:
Frequency of ADC values (ADC value on X-axiz, entries åer ADC value on Y)
protomcm-adcvalues-38-even-sinus-all.pdf  raw plot with 0 < Y <  5000)
protomcm-adcvalues-38-even-sinus-l10000.pdf raw plot with 0 < Y < 10000)
protomcm-adcvalues-38-even-sinus-zoom0.pdf enlargement between ADC values 0-200.
protomcm-adcvalues-38-even-sinus-zoom200.pdf enlargement between ADC values 200-400.
protomcm-adcvalues-38-even-sinus-zoom400.pdf enlargement between ADC values 400-600.
protomcm-adcvalues-38-even-sinus-zoom600.pdf enlargement between ADC values 600-800
protomcm-adcvalues-38-even-sinus-zoom800.pdf enlargement between ADC values 800-1023
protomcm-adcvalues-38-even-sinus-norm.pdf normalized to number of events

Frequency of bits:
Number of times a bit (0-9) is set (normalized to number of events).
protomcm-bits-38-even-sinus-all.pdf

20220809
========
Reanalysed the bit frequency since it was using all samples in the analysis above.
This analysis do not use zero and 1023 valued samples since they are not providing any
useful bit frequency information.
Count number of 0 < value < 1023 samples (= nbsamples) and weight the bit frequency histogram filling with 1/nbsamples,
for each channel and event by event.
Scale with 100/(number of events) to get 0-100% frequency.

Result:
protomcm-bits-38-even-sinus-percent.pdf

The frequency is now around 50% as one expect.
It goes up at higher bits due to the shape of input sinus signal.
Bit 0 deviates for most channels from 50%.


Zerosuppression
---------------
Have run with zerosuppression.
Data not analysed.
Got problems that intermittently the data taking stopped.
Solved temporarily by inserting a delay efter data has been recieved and next trigger sent.
There is one UDP packet in this case. Without zerosuppression it is three.
Guess:
SRU is still busy so it do not see the softtrigger command???
or something else????