Detectors
Near-read-noise-free detectors let the TDT spread the light of many telescopes over many pixels without a read-noise penalty.
Combining the light of many unit telescopes in one spectrograph means recording it on many detector pixels. With conventional detectors the read noise of all those pixels would add up and limit the depth for faint targets. The TDT therefore uses near-read-noise-free detectors, with read noise below 0.1 e− per pixel. Today’s qCMOS detectors already come close, at about 0.3 e−.
With specialised controllers, each detector is split into dedicated rows or pixels for each unit telescope. Units then do not need to start and stop their exposures together, which makes it easy to rearrange the array from one observation to the next.
Towards photon counting
Detector technologies on the horizon, such as microwave kinetic inductance detectors (MKIDs) and single-photon avalanche diodes (SPADs), go further: they record the arrival time of individual photons, and in the case of MKIDs their energy, with no read noise at all. Within the TDT they are developed as MKID spectrographs in the Instrument Cluster. Photon-counting spectrographs would make optical spectroscopy time-resolved on sub-second, potentially millisecond, timescales:
- No fixed exposure times: photons are streamed continuously and can be binned in time after the fact.
- Fast variability, such as flickering, quasi-periodic oscillations and impulsive transients, that integrating detectors smear out becomes directly observable.
- High dynamic range, so that bright and faint components of the same source can be studied together.
Optical astronomy would gain an equivalent of the X-ray “event list”, opening up spectral-timing techniques that are central to powering the Universe.
See the sign-up form if you are interested in joining the TDT effort.