Current Status of the MOIRCS Upgrade Project
After the replacement of the detector to Hawaii2RG in late last year, we have completed some engineering observations for the chracterization by July 2016. The success of the first engineering observation was appeared in a Subaru Web Topic article.
We have confirmed that the new detectors and electronics provide a significant reduction in electronic noise in the data. The new detectors have 14 and 15 electrons respectively of read noise in a CDS frame, which is about a factor of 2 improvement compared with the old system. In multi-sampling exposures we confirmed that the readnoise reduces roughly proportional to the square-root of the multisampling. However, the degree of reduction slows relatively quickly at the number of multi-sampling larger than 20, possibly due to the ambient noise. In fact, we see some streak pattern across the slow-read direction in dark frame.
The new detectors have a better sensitivity than the old one, especially in Y and J bands. The rise of the sensitivity is from a few % (red side) to 27% (blue side). Note that the quantity largely depends on how you treat the QE variation across the detector.
The new system offers the 32-channel readout (cf. old system was 4 channel), which offers faster readout speed than the old system. The measured overhead time during a data acquisition, however, is about 15 sec. This is somewhat longer than expected, and we are still working on this.
Latent and inter-channel cross talk is quite low in channel-1 detector, while some degree of recognizable latent is seen for channel-2 image. We thus recommend to put high-priority target on channel-1 side.
The radiation hit events observed after replacement of the new detector was completely suppressed by the installation of a new alpha-paricle blocker window we have installed just after the last lens. A short report can be found here.
The spectroscopic performance was evaluated in April and May. Overall, the trend of the improvement is as expected from the imaging observation. The improvement of YJ band is quite impressive. Thanks to the reduced readout noise, we can reach the background-noise-limited performance in all grisms we have within reasonable amount of time. This has helped better subtraction of the sky, as we can keep the unit exposure time relatively short, This resulted in the better quality of the final spectra.
We are also planning to introduce the IFU spectroscopy function to MOIRCS, as well as the upgrade of the current MOS system. The date of the downtime is still TBD, though currently starting from early 2017, when MOIRCS might "hibernate" more than a year for possible SWIMS operation, is expected.
We will update the information via this website.