DWARF:Wishlist
From IactWiki
|
Mechanical Construction
- Automatic Lid (e.g. 270 deg flap, shutter, roller blind or accordion-type)
- Little shadowing of Camera housing + opened Lid
Arrangement
- Pixel arrangement hexagonal
- focal plane = Winston Cone entrance plane
Electronics
- DC measurement in central region (about 1 Hz or triggered)
- innermost region (radius ~ 3/5 * mirror PSF): per G-APD!
- intermediate region (tbd): per Pixel
- not necessary in outer region
- With a slow rate for all APDs as starguider'
- actual applied voltage per Pixel
- data rate of 100 Hz (random triggers)
- isochron to 1 ns independent of input shape
- adjustable discriminator thresholds
- single PE peak detectable at low NSB
- dynamical range - from simulations
- switch off the APD signal to be able to record pure electronic noise (if we can measure the single photon peak this might be necessary to determine the base line. Another option would be something which electronically keeps the base line constant so that we don't have to calibrate it)
Ideas for additional readout
- number of indiv. G-APDs above threshold
- sum of clamped and total sum within Pixel
- identifier for brightest G-APD in pixel
- Store information about the trigger signal, e.g. time and length.
- We should have a few "pixels" for which we readout and control all individual APDs (for crosschecks of their bahaviour). Maybe six pixels in the outer ring?
Trigger
- Trigger dead time of the time of at least the readout window
- Usage of the signal length as trigger condition or anti-coincidence condition
- If, after a trigger took place, the signal is still raising we can use the maximum signal height as readout position.
General
- temperature readout
- High modularity for reuse of components in 5 deg Camera
- Humidity monitoring
- atmospheric monitoring, lidar
- Anti-coincidence sensors for bright light (e.g. cars) to switch off the readout (we would store events which are thrown away later anyway)
- Intensity calibratin to calibrate the (non-)linearity of the APDs.
Readout
- the number of readout samples per channel should not be fixed, neither in hard- nor in software
- The effective observation time, resp. the dead time.
- Single photo-electron resolution while not saturating the highest signals (expected between 300 and 1000) (Note, that less APDs per pixel would increase our dynamic range dramatically)
- the shorter the signals the less noise we (have to) integrate!
- write out trigger condition for each event
