GMOS North and GMOS South are currently available for use and are offered in both classical and queue mode for imaging and long-slit, multi-object, and integral field spectroscopy.
Nod-and-Shuffle long-slit and multi-slit spectroscopy can be performed with both GMOSs. Nod-and-Shuffle with the IFU is only available at GMOS-S.
Neither GMOS is equipped with an atmospheric dispersion corrector (ADC) and there are no plans to install such devices. The absence of an ADC in GMOS has several implications on blue imaging and spectral data. A discussion of what GMOS configurations are potentially impacted and what you can do to minimize the effect of the Atmospheric Refraction on imaging and spectroscopy can be found in the Impacts of no ADC webpage.
To check current GMOS configurations:
Mar 19, 2020
As Gemini South is closed for operations from last Monday due to the Covid-19 outbreak (see https://www.gemini.edu/news/gemini-south-shutdown-until-further-notice), GMOS-S has been warmed up and switched off. This shutdown will last until further notice
Jan 10, 2020
After the short thermal cycle on Dec 18, 2019 to replace the coldhead's displacer, the charge smearing in GMOS-S CCD1 has re-appeared as shown in the screenshot of the N&S dark from April 2019. Any programs using the N&S mode are affected and the programs set up with IFU-2 or IFU-R observations are possibly affected by this as well. PIs of those programs are contacted to assess the impact on their data quality and the programs are set on hold. A controlled thermal cycle is scheduled to resolve the issue at the end of January.
Sep 16, 2019
The Gemini-North telescope shutdown was used for a GMOS-N intervention to refill index-matching oil in a (previously inaccessible) lens group through a newly drilled hole in the main optical support structure. Air bubbles in the interfaces of this lens group had previously obscured the bottom ~20% of the GMOS-N imaging field-of-view (see e.g. the update from May 1, 2019 or https://www.gemini.edu/sciops/instruments/gmos/imaging/air-bubble-issue). The lens interfaces are now thoroughly refilled which significantly improves the flat fielding performance in the area previously affected by the air bubbles. The refill lines were retrofitted to make them accessible for future refills as needed. We recommend to not mix science data taken after August 29, 2019 with flats taken before that date and vice versa.
Jun 24, 2019
GMOS-S is back on the telescope and fully operational in all science modes
Jun 10, 2019
GMOS-S is in the CPO Instrument Lab where the ESD board replacement work is starting today. The instrument is scheduled to go back to the telescope on the week of June 24.
May 20, 2019
A thermal cycle was performed on GMOS during last Wednesday/Thursday in order to get the transitory CTE problem fixed.
May 15, 2019
A new controlled thermal cycle will be performed on GMOS-S today in order to address the charge smearing problem reported on April 11, 2019. Therefore GMOS-S will be unavailable till April 12 night.
May 1, 2019
As reported June 20, 2018 (see below) GMOS-N has been affected by loss of index-matching oil between lens interfaces, presumably in the collimator. The remaining air bubble shows up as a darker region on the bottom 20% of GMOS-N imaging. It is not easily corrected for since the bubble moves with telescope position, and at a slow rate due to the viscosity of the oil, so twilight flats are unlikely to match the features in the bottom of science imaging. The affected lenses in the collimator are not accessible without major intervention. Work needed to refill the lens interfaces is currently being discussed and planned, and may happen as early as the July-August telescope shutdown.
April 11, 2019
The charge smearing has reappeared in GMOS-S CCD1 (screenshot of Nod&Shuffle Dark is attached). We have set the GMOS-S MOS N&S programs on Hold since it can severely affect the data quality. The daily monitoring is ongoing and the data are being investigated.
February 21, 2019
Starting from semester 2019A, GMOS on-sky arc lamp exposures for blue setups which do not cover the 5577 A sky line are provided as baseline calibrations (see Baseline Calibrations for detailed information on the corresponding grating/central wavelength configurations). As before, on-sky arc calibrations for other setups, if required, can be requested as part of the program time. Daytime baseline arc lamp exposures will continue to be provided for all spectroscopic observations that do not include on-sky arcs.
October 1, 2018
GMOS-N is back on sky after the Gemini North Shutdown which took place between 17 and 27 September 2018. Since the GMOS-N filters and imaging mirror were cleaned as part of the instrument maintenance work during the shutdown, it is recommended to not use flats taken before the shutdown for science data taken after the shutdown.
September 3, 2018
GMOS-S came back on sky after the major optical intervention (in which the collimator was removed for the first time since the instrument was built). Everything looks very good and GMOS started delivering science data again. After the fix, the flatfielding performance is substantially improved.
July 27, 2018
GMOS-S came off the telescope on July 18, and after preliminary masurements on the CP instrument lab, the work for the collimator service started this week with the removal of the masks frame and its supporting structure, in order to free the way for the collimator removal.
June 25, 2018
This is a reminder that GMOS-S will be off the telescope between July 17 and August 31. During this period we will perform a major optical maintenance on the collimator in order to address the effects of the inter-lens oil loss that have been developing over time. Please see the GMOS air bubble issue page (https://www.gemini.edu/sciops/instruments/gmos/imaging/air-bubble-issue) for more details.
June 20, 2018
Recent GMOS-N data have been affected by an accelerated loss of index-matching oil from one of the collimator lens interfaces. The issue was diagnosed based on an analysis of twilight flats which show that an increasing area in the lower part of the frame was affected by a partial obscuration. A visual inspection of the instrument on 6 June 2018 revealed a major oil leak in one of the collimator lens groups which had led to an air bubble affecting almost half of the field of view. The lens group was repaired and refilled with oil on 7 June 2018. The same intervention was also used to clean the GMOS-N imaging mirror. While this particular collimator lens group could be refilled, less extended air bubbles resulting from the gradual loss of index-matching oil from other GMOS lens interfaces have been a long-standing concern. Only some of the lens refill ports are accessible, while other lens groups cannot be refilled without a major intervention.
Whether the air bubbles have a limiting effect on the data quality depends on the science goals. A detailed assessment of the effects on photometric and, possibly, astrometric accuracy is currently ongoing. More information on this issue, including suggested data reduction strategies, can be found here. GMOS-N users who encounter issues with the photometric or astrometric accuracy in their data - especially due to the recent extended GMOS-N air bubble which might have started to affect data since late 2017 - are invited to contact the GMOS-N team (gmos_n_science 'at' gemini.edu).
June 1, 2018
GMOS-S is back to business!!
GMOS-S came back fine after the controlled thermal cycle performed this week. In the biase images taken after the cooldown yesterday, bright structures are completely gone and readout noise is back to its nominal value (~4e-). The CTE is ok in the three chips and charge smearing is no longer present in CCD1. Therefore GMOS-S is ready for science in all modes.
May 28, 2018
GMOS-S will be warmed up on Monday, May 28 in order to perform a controlled thermal cycle and improve the vacuum quality. It will NOT be available for science operation during this week. It is scheduled to start cooling down on Thursday, May 31 and will be back on sky by Friday night (June 1st).
May 14, 2018
The charge smearing has reappeared in GMOS-S CCD1 (N&S dark frame attachd below). We have set the GMOS-S MOS N&S programs on Hold since it can severely affect the data quality. The daily monitoring is ongoing and the data are being investigated.
May 4, 2018
On April 16, 2018 we noticed that CCD1 started showing charge smearing. We checked the earlier data and it first happened on GMOS-S data taken on April 3, 2018 at a very low level (1-2 counts). It becomes more evident in long exposures caused by numerous cosmic ray hits, but still stays at the 3-5 counts level.
On April 28, 2018 GMOS-S was warmed up for the coldhead displacer replacement. After the full thermal cycle of the dewer, the charge smearing problem disappeared but the bias pattern which existed from September 30, 2016 to February 21, 2017 reappeared in CCD2 although they are not as strong as the previous ones in 2016. PIs with data taken after April 28, 2018 should use biases taken after that date. We recommend to use the bias frames taken on the same day to properly remove the pattern shown in the figures. We are currently analyzing the bias data and performing a quantitative analysis on the impact of the enhanced noise levels.
Left: GMOS-S r-band raw image taken on 4/29/18, middle: Bias pattern in CCD2, right: reduced image after applying bias subtraction
April 24, 2018
GMOS-S will be warmed up today in order to perform the scheduled coldhead service. It is expected to be back online for thursday night. The problem reported on April 16 persists, it is being monitored and is stable.
April 16, 2018
GMOS-S CCD1 developed a charge smearing a few days ago, similar to the one affecting the detector in May 2015 (see May 7 , 2015 update for more details). This can severely affect N&S programs data quality, due to this all N&S programs are being set on hold (unless PIs decide to continue the execution despite this effect). The problem is currently under investigation and the behavior is being monitored.
September 8, 2017
GMOS-N science observations were resumed on August 31 after completion of the Gemini-N shutdown and instrument on-sky checkouts. As the GMOS-N OIWFS potentiometers were replaced during the telescope shutdown period, the on-sky checkouts for GMOS-N included a recalibration of the OIWFS mapping solution. Work on re-establishing the supplemental fine mapping solution is currently ongoing. The fine probe mapping allows for improved accuracy during large blind offsets.
The pre-combined GMOS-N biases and flats in the Gemini Observatory Archive are now overscan-subtracted. (The corresponding GMOS-N frames for data taken prior to the July/August 2017 Gemini-N telescope shutdown are not overscan-subtracted.) This change was implemented, because the time-variable and curved bias levels are best removed by using overscan-plus-bias subtraction. For optimal reduction results, we recommend users to create their own master biases and flats from raw frames instead of using the archival pre-combined biases and flats.
July 20, 2017
A new version of the Gemini IRAF package (v1.14) has been released. This release provides support for the new GMOS-N Hamamatsu CCDs. Installation instructions and further information can be found here.
Gemini-North is currently in scheduled shutdown until August 25. The shutdown period is used for GMOS-N maintenance tasks, including the replacement of potentiometers in the OIWFS base and pickoff stages.
June 20, 2017
GMOS-S is back on the telescope and the scheduled mechanical maintenance, and already taking science data again.
June 2, 2017
GMOS-S was removed from the telescope yesterday for a scheduled mechanical maintenance period at the CPO Instrument Lab. It is sheduled for going back on sky on June 15.
April 7, 2017
First science observations using GMOS-N with the new Hamamatsu detector array have been obtained since March 26, 2017. The final detector characterization and update of the web pages and full data reduction package is ongoing. A beta version of the reduction package for the main observing modes will be released in approx. 2 weeks. (Please note that the beta version can be used for a quick-look reduction, but will not provide science-quality results, as the detector characterization is ongoing and important detector parameters are therefore not yet available.) Updates will be announced on this web page.
March 14, 2017
The installation and alignment of the new Hamamatsu detector array in GMOS-N was successfully completed during February. GMOS-N was mounted back on the telescope last week and the daytime and on-sky commissioning work for the new CCDs is currently ongoing.
March 2, 2017
The Hamamatsu detector array has been installed in GMOS-N. GMOS-N is expected to be mounted on the telescope next week to start the commissioning of the new CCDs.
February 21, 2017
The GMOS-S bias structure reported previously (see notes from November 3, 2016; February 6, 2017) are spontaneously gone as of today, after an unscheduled thermal cycle (yesterday the instrument was cooled down after it had warmed up on sunday due a failure on a helium compressor). Biases are mostly clean (some residual vertical fringes are apparent on binned data, but only about one or two counts over the bias level on average). Only the known bad columns persist (amps #3, #5, #8 and #11). Noise values are the normal ones (~4e- rms).
February 10, 2017
A new R150 grating has been available at GMOS-N since December 2016. This grating replaces the previous R150 grating which had developed an issue with the coating causing a loss of throughput. A throughput comparison of the new and old grating and more details on the issue with the previous R150 grating are available here.
February 7, 2017
GMOS-N was removed from the telescope on February 3rd, 2017 for the scheduled installation of the new Hamamatsu CCDs. The instrument will be unavailable during February and parts of March for the installation and commissioning of the new CCDs.
February 6, 2017
This is an update on the GMOS-S bias structure (see note from November 3, 2016). Several engineering tests have been performed, together with consultation with the manufacturers as well; but a solution has not yet been found. Currently we are carefully monitoring the situation, whaich has not changed since the last thermal cycle on late September 2016.
January 16, 2017
This is an update on the recent GMOS-N noise issue (see note from December 2, 2016), which had been intermittently affecting all amplifiers of the detector mosaic since November 22, 2016. Regular bias monitoring shows that the noise has been stable and back to a consistently low level since December 27, 2016 (rms of ~3 e- in the standard science readout mode). The noise issue reported on December 2 was possibly caused by loose bolts in the cold head mounting brackets, which were discovered in the course of the investigation into the problem on December 27, 2016.
December 02, 2016
The GMOS-N detector mosaic has recently shown incidents of high noise levels associated with a time-variable chevron pattern. The issue was first noticed in frames taken on November 22, 2016 and affects all amplifiers. Bias frames taken over the last two weeks show that the noise level is variable. The most affected frames have an rms of up to ~12 e- (i.e. about 4 times the normal rms level of ~3.2 e- in the slow/low readout mode), while other frames show lower noise levels or no evidence of any increased noise or pattern. Efforts to understand the origin of the noise and to resolve the issue are currently ongoing.
November 03, 2016
This note provides an update on the situation that occurred one month ago when the GMOS-S detector mosaic developed a new structure on its bias frames, especially on CCD2 (see previous post). This matter is under active investigation and will not be easy to resolve. Here follow some further details that have been obtained recently.
In the most affected areas of CCD2 the detector is significantly noisier. After bias subtraction, noise can be enhanced as much as five times with respect to unaffected areas. For example, if the normal noise on a bias-subtracted 1hr dark is ~3.7 counts rms, in the region with these new features it can be 16 counts rms. When the noise is dominated by the sky background, this is less of an issue. For CCD3 the effects are milder, and CCD1 is almost unaffected.
The dominant bias features are not uniform but highly structured (bright, vertical lanes), and they do not always subtract perfectly. Some residual vertical features may remain after bias subtraction and can be of the order of a few counts to tens of counts. If one does not succeed in removing the bias structure satisfactorily, stacking individual exposures can make it actually worse. The cause for this is still not understood, yet it is under investigation as more data are being acquired.
This bias structure is more harmful in the following cases:
- Imaging of very faint targets;
- Imaging of low surface brightness structures, as the residual features can interfere with the detection of low signal levels;
- Spectroscopy of faint targets. If spectral features fall in the affected regions they can get confused with residual detector features, and possibly an exposure time that was good before may not be sufficient anymore to reach a desired signal-to-noise for affected particular spectral intervals.
Additional complications may arise during data reduction. For example, fringing removal in the z' and Y bands is complicated because of the added noise and residuals from the subtraction. Also the number of bad columns has increased with each thermal cycle.
Clearly the situation is far from optimal, and we are concerned as to how the situation will develop. PIs who have recently obtained data (after Sept 30, 2016) or will obtain data from GMOS-S during the coming weeks and months, are invited to carefully inspect their data and report any problems to their Contact Scientist. We also reiterate our advise to use bias frames taken after Sept. 30th for data taken since that same date, preferably as close as possible in time.
October 03, 2016
On Friday, September 30, after a service on the GMOS-S coldhead (which requires a full thermal cycle of the dewar), new features appeared on GMOS-S raw data, the most notable features being vertical fringes on CCD2 and CCD3. This is more pronounced on CCD2 where the peak counts of these fringes now climb to ~100-400 counts above the background. For CCD3 is milder, CCD1 is the least affected. Additionally there are horizontal 1-pix stripes across CCD2 and CCD3. These features can be removed by proper bias subtraction (see first screenshot for an example on imaging). PIs with data taken after September 30, 2016 are advised to use biases taken after that date. The noise will be enhanced somewhat at the location of these features which can affect spectroscopy of faint targets, particularly in MOS and IFU observations. The second screenshot shows an example of a MOS Nod&Shuffle observation (zooming into a section of the frame close to the gap betweeen CCD1 and CCD2). The raw frame is on the left, the bias-subtracted only is on the center, and the reduced (bias-subtracted, A-B differenced and cleaned for cosmic rays and bad columns) on the right (larger image here)
May 05, 2016
The mask cutting machine at Gemini South is back in service as of Friday, Aug 26; and mask cutting for both GMOS-S and GMOS-N has been resumed in the south. PI's please mind the time needed for shipment when submitting GMOS-N mask designs (see the PII deadlines page for more details)
Aug 3, 2016
The GMOS-N general maintenance, originally planned for the second half of July, has been rescheduled for the telescope shutdown period Aug 10-31. The new R150 grating is foreseen to be installed in GMOS-N after the telescope shutdown period; the exact dates are to be confirmed.
Jul 18, 2016
The mask cutting machine at Gemini South is offline due to a hardware failure, and mask cutting for both GMOS-S and GMOS-N will be done in the north. It is currently unclear when it will come back online. PI's please mind the time needed for shipment when submitting GMOS-S mask designs (see the PII deadlines page for more details).
May 19, 2016
GMOS-N will be available through June, and will be removed for a two week period during July 11 and August 1 for general maintenance, the exact dates are to be confirmed. In addition, we will be replacing the R150 grating in mid-June and it will therefore be unavailable for a couple weeks.
May 05, 2016
The mask cutting machine at Gemini South is back in service, and mask cutting for both GMOS-S and GMOS-N has been resumed in the south. PI's please mind the time needed for shipment when submitting GMOS-N mask designs (see the PII deadlines page for more details)
March 16, 2016
The GMOS-N Hamamatsu upgrade project has met with further challenges. At this time we do not believe there is any problem with the CCDs themselves, but various components of the detector controller electronics are presenting issues. Unfortunately this means the planned June removal of GMOS-N from the telescope for installation of the new CCDs and commissioning in July will be delayed, and likely the new CCDs will not be available for semester 2016B as previously expected.
February 3, 2016
The R150 grating for GMOS-N has been degrading for about 4-5 years. A new R150 grating is now expected to be installed in the middle of the year between semesters. The degradation is occurring only at certain wavelengths centered on 550nm and at most the reflectivity is degrading at ~ 1% per month. More details are available here.
November 2, 2015
GMOS-S is being installed today back on the telescope after the scheduled telescope shutdown period. During the shutdown the instrument was removed from the ISS and taken to the CPO Instrument Lab. The dewar was opened and inspected as planned. No signs of problems or anomalies were discovered, however the connector cable for CCD1 (suspected to be related with the intermittent electronic problem described on May 7) was replaced by a new one. The detector was then cooled down to operational temperature, and alignment checks performed. GMOS-S will be back on sky into normal operations within the next few days.
October 14, 2015
GMOS-N has developed a noise problem. The problem was discovered about a month ago, just prior to the Gemini North shutdown. During the shutdown, emergency work on GMOS cold head and grounding issues was conducted and the noise levels appeared to return almost to normal. Unfortunately the noise has proven to be variable and in recent days is once more unacceptably high. The problem appears to have started (initially at low levels) in the last weeks of July coincident with the GMOS-N cold head failure and subsequent displacer replacement. Whether or not the two events are related is conjecture at this point. In addition to higher read noise (variable from ~10% above normal to 3 x higher: readout noise plot) the images also exhibit a chevron pattern (typical of pickup noise) with amplitude that varies with time and location on the detector (chevron noise example image). The chevron pattern does not subtract off though it may average away to some extent when multiple images are summed. Both types of noise will most significantly impact dark time imaging and blue spectroscopic data with low sky background. The problem is seen to effect all 6 amplifiers equivalently as well as both gain settings and readout speeds. Efforts to identify the source(s) of the noise and eliminate them are actively ongoing.
September 28, 2015
The mask cutting machine at Gemini South is currently out of service due to an unexpected technical problem, and hence GMOS-S MOS masks are currently being cut at Gemini North. PI's please mind the extra time needed for shipment when submitting GMOS-S mask designs (see the PII deadlines page for more details)
September 03, 2015
GMOS-S is back into nighttime operation as of September 02. Data taking is being conducted normally and observed datasets are available through the GSA as usual. The values for the new gains will be available soon (not before next week though) for the use of PIs willing to reduce their science data promptly. General updates on the Gemini Iraf package for GMOS can be expected to be available by late September, as mentioned on the August 14 update.
August 26, 2015
The new 'Revision E' video boards have been installed in GMOS-S yesterday. The saturation 'banding' effect is gone and now the re-charaterization is ongoing. It is expected that GMOS will be operational within a few days.
August 14, 2015
The GMOS-S Science and Engineering teams outlined the remediation plan to be implemented starting on the last week of August. Meantime, decision was made to resume Nod and Shuffle observations, as long as the behavior remains normal, and with the assumed risk that the CTE issue could potentially be back any time. Daily monitoring will continue though, to ensure no science data are compromised inadvertently.
The timeline for the remediation plan comprises two stages. On the first one, the new 'Revision E' video boards will be installed in the instrument on the last week of August, during the GPI campaign. This intervention is done without removing the instrument from the telescope, and the goal is to fix the saturation 'banding' effect. Prospective video boards had been configured and tested between May and July and it was verified that they successfully eliminate the effect completely. Installation (hardware and software) will take a full day. Once verified that the video boards work without trouble, regular observations can be resumed immediately; it is to be noted however that full characterization -i.e. deriving new gain, linearity, full well and noise values- will need several hours of data taking and therefore the full characterization will only be completed after several days. Therefore and assuming no unexpected issues, updates on the Gemini Iraf package for GMOS can be expected to be available by late September.
The second stage contemplates the investigation of the CCD1 CTE intermittent problem reported in May. For this the instrument will be removed from the telescope, and taken to the CPO instrument lab during the telescope shutdown scheduled for October. Once there the cryostat will be opened for intervention, addressing the electronics inside the dewar first, as preliminary tests had indicated that the most probable cause lies inside the cryogenics; while problems with the actual chip seem very unlikely. Simultaneously, the fix of a minor vacuum leak that was detected several weeks ago will be tried in October.
There exists nevertheless, the risk of not finding the source of the CTE problem, or finding that a repair in October will have too big an impact on the planned science. In this case it may be decided to close up without actually having solved the problem in a permanent way,
The remediation plan for the GMOS-S CCD has been designed so as to minimize the impact on science operations as much as possible, within the existing constraints and complexity of the issues to be addressed. It is expected though, that this will result in one to two weeks of effective downtime, depending on the nature of the underlying problem and the timescales for troubleshooting.
August 06, 2015
GMOS-S was warmed up after a power outage on the night of August 5 UT. As severe winter weather is forecasted for the CPO summit the next four days, the instrument will be unavailable until normal operations are resumed. Also further information about the remediation plan for the detector issues will be posted next week.
July 28, 2015
The CCD1 CTE problem is now spontaneously absent, without any intervention, as revealed by test images taken during Monday. Menwhile the team met as planned and the remediation plan outlined, final details still to be decided. The situation is a complex one given that GMOS is in near continuous use. The instrument will be scheduled for a shutdown period in which the new video boards (which fix the saturation 'banding' problem) will be tried on the full detector, and the potential causes for CCD1 CTE issue will be addressed in the lab. This period will last from 3 to 6 weeks, depending on key decision points. Therefore an impact on science can be expected during August and September.
July 16, 2015
The GMOS Science+Engineering team will meet next week to define the repair plan, in order to addres both the saturation 'banding' efect fix implementation and the CCD1 CTE issue troubleshooting.
June 19, 2015
The HBF Lab Controller arrived in La Serena earlier this week. It will be tested on Tuesday, June 23. This test will aim to address the charge transfer issue reported on May 7; in particular, to discern whether the root cause lies on the cold electronics or the CCD1 detector itself.
May 28, 2015
The problem with GMOS-S CCD1 reported on may 7 continues unchanged, after more hardware tests have been performed on the warm electronics last week. A plan has been devised this week for the troubleshooting and it execution will begin within the next weeks as new hardware arrives at GS.
May 7, 2015
A new video board prospect was tested in order to address the saturation banding in binned data problem on GMOS-S Hamamatsu CCDs (see the 'Detector array' section for Hamamatsu for details on this issue). The test board was used with only one of the CCDs and was successful in removing the effect completely for all binnings. The next step will then be the installation the new definitive boards on all three CCDs to have this problem fixed definitely. The approximate schedule is about two months.
Another problem has developed during the last couple of weeks, namely a charge transfer issue affecting CCD1 in Nod and Shuffle data. The snapshot below shows a cutout from a 20 minute MOS exposure using 10 N&S cycles. As can be seen, CCD1 (left) is affected by a smearing of counts (~10% for a 15cycles N&S) along the shuffle direction with an amplitude equal to the shuffle distance. On the other hand CCD2 (right) is not affected, as well as CCD3. PIs of N&S programs have been made aware of the effect, and N&S programs are being set on hold (unless PIs decide to continue the execution despite this effect). The problem is currently under investigation and the behavior is being monitored (so far it has been stable).
February 26, 2015
This note is intended to raise the awareness of a recently discovered problem with one of the Hamamatsu CCDs installed in the GMOS-S instrument.
Currently the readout of the GMOS CCD shows a “hot” column, which for some reason accumulates spurious charge. This hot column is located on the second CCD and is read out through amplifier #5, at x=1175 when binning is 2x2, so just off the center of the field. With increased exposure time and binning the saturation of this column becomes worse. The effect of this hot column, in case of severe saturation is that the background level for the section of the CCD that pertains to the same amplifier can significantly deviate from the normal level by up to 25%. Thus the usual bias subtraction and flat fielding does not result in a good uniform background. The noise characteristics of this effect are as yet unknown. Under some configurations, subtracting the local background may correctly recover the signal, possibly with a somewhat higher noise. But if the background level is strongly affected then the quality of the science data will also be affected. Hence the best strategy is to avoid this problem as much as possible.
A severe example is shown below.
An investigation of the historic trend of this effect shows that the problem started as a hot pixel of little impact. But particularly during the past few months the problem has gotten progressively worse, now affecting most of the columns and hence most of the sector that is being read out through the same amplifier.
Depending on the type of observation and the science objective the effect may be limited and can possibly be corrected for, or even disregarded. But particularly for long exposures, and full frame binned readout the effect may have a big impact. This has to be assessed on a case-by-case basis.
Our advice to minimize the effect of the hot column when preparing observations is as follows:
- · Preferably use the detector without binning, as this much reduces the saturation effect on the hot column.
- · Avoid exposure times in excess of 900 seconds in unbinned mode.
- · Avoid critical parts of the spectrum or image to fall on the section that is read out through amplifier #5.
- · If possible, carry out dithered observations, although this reduces much the observing efficiency.
In case of doubt, don’t hesitate to get in touch with your Gemini contact scientist.
Our expectation is that new controller electronics that are being prepared will resolve this problem, although at the time of writing we cannot be certain of this.
Example GMOS-S raw mosaic image of the three Hamamatsu CCDs, in inverted grey scale. Each CCD has 4 readout amplifiers, resulting in four broad vertical bands per CCD. The horizontal dark band is a stellar spectrum. The leftmost section of the central CCD shows spurious uneven intensity levels that are the result of severe saturation of a CCD column.
December 19, 2014
GMOS-S is back on sky!. After a period of maintenance on the components, last night the engineering on-sky checks were performed successfully and queue observations were resumed.
November 28, 2014
GMOS-S will be removed from the telescope on Dec 2 for two weeks, and a maintenance work will be performed on the mask assembly and the OIWFS system, in order to address som reccurring faults. An assessment of the optical system will also be done. The scheduled back-on-sky date is Dec 18, 2014.
November 19, 2014
The R150 grating for GMOS-N is degrading. The degradation has only recently been discovered, but has actually been going on for about 4 years. A new R150 grating has been ordered, and delivery is expected in April 2015. The degradation is occurring only at certain wavelengths centered on 550nm and at most the reflectivity is degrading at ~ 1% per month. More details are available here.
August 22, 2014
GMOS-S is back on sky, after the Gemini South shutdown period. The problem with the electronics that affected one of the CCDs was successfully fixed and the instrument already started taking science data.
July 22, 2014
GMOS-S will be removed from the telescope today in order to fix some issues with the controller electronics. These were discovered during the first phase of the commissioning. GMOS-S will be back on sky after the scheduled GS shutdown (see telescope schedule for more details on this).
June 25, 2014
GMOS-S Hamamatsu CCDs began taking science data! While weather has been mostly poor, the final on-sky commissioning tasks are ongoing, intercalated with queue observing. Imaging, longslit spectroscopy and IFU spectroscopy have been enabled. MOS spectroscopy is still undergoing commissioning, however MOS programs that have had their preimaging taken with E2Vs can be also scheduled as soon as their masks are submitted, cut and installed.
June 10, 2014
GMOS-S was reinstalled on the telescope last friday and cooldown of the new detectors started immediately. The first daytime commissioning tasks were performed on Saturday 7th, and the daytime activity will continue during the week. On-sky commissioning is scheduled to start after June 17th.
May 22, 2014
The Hamamatsu CCDs have undergone an intensive reliability testing at CPO for the last month. GMOS-S came off the telescope on Monday, May 19 and is now on the CPO Instrument Lab, being prepared for the E2V detectors removal and for receiving the Hamamatsu CCDs.
February 26, 2014
The Hamamatsu CCDs will begin the last reliability testing in Hilo this week, and if all goes well would be shipped to GS by the end of March. The installation in GMOS-S is currently scheduled for on the second half of May. The expectation is to finish the on-sky commissioning by mid July, so that the instrument is ready for normal operation starting 2014B.
March 01, 2013
As mentioned in the August 31, 2012 news item below, the Observatory will install the new Hamamatsu CCDs in GMOS-S during the 2013B semester. The work is planned to occur during October-November, simultaneously with other instrument maintenance, and will take the instrument offline for ~6 to 8 weeks. The new detectors will provide significantly improved sensitivity, particularly in the red, and ~15 hours have been reserved in the 2013B time allocation for Demonstration Science observations after recommissioning The expected improvement over the current CCDs for GMOS-S is substantial, with significantly higher QE long ward of 550nm and essentially unchanged sensitivity in the blue (GMOS-S Hamamatsu QE comparison). Meanwhile, we are developing a plan to install a second set of detectors into GMOS-N in 2014.
October 22, 2012
Repairs to the GMOS-S mask mechanism are complete, and the instrument has been released for full spectroscopic science use. MOS and longslit capabilities were restored on October 08, and installation of the IFU is scheduled for today. Several GMOS-S program with early targets accessible from Mauna Kea have had those targets transferred to be observed with GMOS-N. This not only relieves the backlog on the GMOS-S queue resulting from this incident, but also helps the Gemini North queue which is exceedingly thin for the first few hours of the night.
September 22, 2012
The GMOS-S mask assembly suffered a serious failure the evening of September 17. Work is underway to repair the mechanism and understand the cause of the faiure. Currently the instrument is available for use in imaging mode only
August 31, 2012
The following announcement made today impacts the installation of Hamamatsu CCDs into GMOS-N and GMOS-S:
- Gemini has decided to stretch the schedule for the GMOS CCD upgrades. This decision is driven by the combination of scarcer resources from next year on and the will to focus the efforts in the first half year on FLAMINGOS-2 and GeMS.
We understand that the upgrade is eagerly awaited for by the users, but Flamingos 2 and GeMS are even more so. Furthermore, the E2V Deep-depletion devices are currently providing a significant boost in performance in GMOS-N.
The new timeline foresees an upgrade of only one system in 2013, and this is expected to be GMOS-S during the third quarter. GMOS-N will be offered with the e2v DD CCDs until its upgrade, expected in early 2014. A new decision point is set to January 2013 in order to complete the 2013B call.
February 01, 2012
The following e-mail was distributed to all PIs with GMOS-N program time awarded in semester 2011A and all 2011B semester PIs with data taken after the GMOS-N CCDs were upgraded. The data process development team recommends that only investigators reducing data taken with the GMOS-N e2v Deep Depletion CCDs download this beta patch as it is not intended as a full release.
- Data reduction support for the new GMOS North e2v Deep Depletion CCDs is now available. The gmos package has undergone many changes to ensure compatibility with 6-amp data from the new CCDs, including improved gain values and the addition of new BPMs. Please note that only data taken in slow read, low gain mode is supported. For other modes, please contact the GMOS team. Full support for the other modes will be included in a future release. The latest version of the Gemini IRAF package (v1.11) must first be installed before the beta patch can be installed. The Gemini IRAF package v1.11 is available for download at http://www.gemini.edu/sciops/data-and-results/processing-software. The beta patch, which is required for processing data from GMOS North e2v Deep Depletion CCDs, is available for download at http://www.gemini.edu/sciops/data-and-results/processing-software/releases/downloads/prereleases. Full installation instructions for the beta patch are available at http://www.gemini.edu/sciops/data/software/gemini_v111_patch1.txt. Please report any problems and / or provide feedback on the beta patch via the Gemini HelpDesk: http://www.gemini.edu/sciops/helpdesk/helpdeskIndex.html.
November 22, 2011
Gemini is pleased to announced that GMOS-N has been released for science use. There are, however, some caveats. The upgrade to e2v Deep Depletion devices has not been without it's complications, and was never intended as a permanent upgrade to GMOS-N. The detectors have been characterized as well as the limited time allowed, but not optimized. One controller re-cabling optimization was enacted in order to reduce spurious noise effects, but an unanticipated side effect of this was the unfortunate elimination of the "best amps" controller option. Efforts are underway to restore the "best amps" option, but due to time constraints and resource unavailability we do not expect to have this recovered until after the Thanksgiving holiday, if ever. In the meantime, all science data will be obtained with all six amps. Every effort is being made to ensure that Gemini IRAF data reduction scripts to properly handle the new detectors as well as six amps will be made available as soon as possible. In addition, due to the thickness of the new detectors, the GMOS team recommends exposure times not be longer than 40 minutes. Gemini staff will work with PIs to adjust exposure times and the number of amps used as necessary. The exposure time limitation will be added as a "Phase II check" in the December OT release. Detailed characteristics of the new detectors will be added to the GMOS-N Array (e2v DD) webpages as they become available.
November 01, 2011
GMOS-N was installed back on the telescope last Wednesday, October 26. On Friday a new detector temperature controller was installed in order to raise the operating temperature of GMOS-N from -120C to -100C, the manufacturer recommended value. By operating at slightly warmer temperature the QE improvement afforded by the new CCDs will be maximized. Tests over the weekend revealed an an issue with signal clipping, and today a parallel clock voltage was adjusted to correct this. Tests are on-going to assess the effect of both of these changes on dark current and to confirm we are now taking advantage of the increased full well that these detectors also possess compared to the original EEV detectors. On-sky checkouts have already been completed, and once we are happy with the detector controller operating parameters we will take our first science data with the new detectors.
October 17, 2011
GMOS-N has new detectors! The e2v deep depletion devices have been successfully installed and dark images (with the instrument still warm) have confirmed that all amps are functioning within expectations. The first test images with the GMOS-N dewar integrated back onto the instrument and cooled (but still off the telescope) are expected to be taken Thursday. If all goes well and if enough progress can be made on NIRI queue programs we could see a return to sky with the new and improved GMOS-N toward the end of next week!
October 14, 2011
GMOS-N was removed from the telescope Tuesday October 11 and is currently being disassembled. The dewar is being transported to the instrument lab on Mauna Kea where e2v DD CCD exchange will take place this weekend.
September 29, 2011
The new e2v 42-90 deep depletion devices have been received at the Gemini North Operations Headquarters in Hilo, HI. So far we are still on schedule with GMOS-N removal from the telescope scheduled for October 11.
August 12, 2011
Further testing of the Hamamatsu detector with the bad output has suggested that this CCD is no longer science worthy. While testing is continuing, plans to purchase a replacement (or spare) CCD from Hamamatsu (likely an HSC-type detector) are proceeding simultaneously. In order to bring an upgraded GMOS-N focal plane to the community as soon as possible, while allowing time for Gemini-HIA to develop and adopt safer ESD procedures and insure that a healthy Hamamatsu focal plane will be installed in GMOS-N, Gemini announces the following modification to the GMOS-N CCD upgrade plan:
- Gemini has signed a contract with e2v for the purchase of three deep depletion (DD) detectors (device designation: 42-90 with multi-layer 3 coating). These CCDs are "plug and play" replacements for the current EEV detectors, requiring no modification to the GMOS focal plane and very minor adjustments to the detector controller software. The CCDs are expected to arrive at Gemini in mid/late September of this year; note, however, that the delivery dates (and the subsequent installation and commissioning dates) are not yet confirmed.
- The current GMOS-N CCDs will be replaced with these new devices. These CCDs have improved sensitivity in the blue and the red compared to the original detectors, and extend the sensitivity to 0.98 µm. The fringing with these detectors is also much improved compared to the current GMOS-N CCDs, and is expected to be ≤ 1% peak-to-peak. More information will be made available on the GMOS-N Array (e2v DD) webpages.
- Aside from the QE and fringing, the new e2v DD devices are essentially identical to the existing CCDs. The only other upgrade planned at this time is an investigation to see if the "dust bunny" can be removed.
- Pending confirmation of the delivery schedule, GMOS-N is expected to be removed from the telescope October 11. The installation of the new devices is expected to take 5 weeks, with GMOS-N going back on-sky November 16. Barring complications, we expect to be taking science data again immediately with commissioning data necessary for characterizing the new detectors obtained in the queue.
- A patch to the Gemini GMOS data reduction scripts supporting the upgrade to e2v DD devices in GMOS-N will be made available to GMOS-N PIs for "shared risk" use as soon as possible after the upgraded GMOS-N starts obtaining queue data. After a sufficient period to finalize the CCDs characterization, complete testing and address issues, the patch will be made available on the Gemini webpages, nominally by the end of the year.
Gemini is still fully committed to delivering an upgraded focal plane array populated with Hamamatsu devices, yielding unsurpassed sensitivity in the red and extending the useful wavelength range of the GMOS-N instrument to 1.04 µm. It is expected that the e2v DD devices will populate the GMOS-N focal plane for ~6 months to 1 year maximum. Once the upgrade to GMOS-N is complete Gemini will then proceed with plans to upgrade the GMOS-S focal plane.
June 15, 2011
The Gemini-HIA project to upgrade the GMOS-N CCDs with red sensitive Hamamatsu CCDs has encountered an unexpected issue. One of the new detectors recently developed a bad output. Although the direct cause is unknown, we suspect the CCD may have been affected by an electrostatic discharge (ESD) event. While it appears that the CCD will still perform to specification for science use after adjustment of certain operating voltages, it must still be fully tested; such tests are underway at present. Discussion with other groups working with these CCDs and other similar fully-depleted devices reveals that they are much more sensitive to ESD and more easily damaged than previously appreciated. As a result of this new information, modifications to Gemini-HIA handling procedures are being put in place to ensure that the new focal plane can be safely installed in GMOS-N without risking electrostatic damage.
In addition, other recent evidence indicates that this class of devices may suffer a permanent reduction in well depth if exposed to bright light for a sufficiently long time. The need for over-illumination protection (within GMOS-N) will need to be evaluated, and possibly implemented.
The potential for significant delays resulting from one or both of these parallel tracks of investigation is high, and at this time we feel it is prudent to accept that it is unlikely the Hamamatsu GMOS-N CCDs will be commissioned in October as previously advertised. The new telescope schedule estimates commissioning will not happen before mid-December, but re-evaluation of various options for delivering improved GMOS-N CCD performance is on-going. Proposers for GMOS-N programs should continue to define their phase II observation sequences using the existing e2v CCDs until further notice.
April 28, 2011
The Region of Interest (ROI) for Central Spectrum and Central Stamp have been redefined to correct a small error whereby these ROIs were not exactly centered on the detector as had been advertised. Prior to the update the Central Spectrum was one row too low, and the Stamp was both one row too low and one column too far to the left. With the change there are several header keyword values that have changed accordingly, including DETRO1X, DETRO1Y in the primary header unit , and CCDSEC and DETSEC in the science data extensions  - . The change occurred on UT 2011 April 27 for GMOS-North and UT 2011 April 28 for GMOS-South. When reducing data using these ROIs taken after these dates (in particular spectrophotometric standard stars normally use the Central Spectrum ROI) users should take care to use bias exposures from the same date or later so as to be assured of not mixing ROIs.
February 28, 2011
A replacement for the defective CCD has been ordered from Hamamatsu and is scheduled to arrive by the end of March. Upon confirmation of delivery at Gemini, the new CCD will travel to HIA to be incorporated into the new focal plane with the other two Hamamatsu CCDs. This third CCD has a different anti-reflective coating (developed for the Hyper Suprime-Cam instrument on Subaru Telescope), presenting much improved blue QE between 400-650 nm while approximately matching the improved red response of the other two Hamamatsu CCDs (originally designed for Suprime-Cam). This new CCD will be inserted at the right end of the focal plane array so that the blue end of the spectral dispersion will land on this blue enhanced detector.
- Expected CCD characteristics including quantum efficiency are available here.
- These quantum efficiencies have been incorporated into the GMOS-N ITC, and when proposing for semester 2011B PIs should choose either Hamamatsu Red or Hamamatsu Blue depending on their spectral regions of interest when estimating exposure times.
- Work is underway to incorporate the relative QE into the GMOS data reduction scripts within the Gemini IRAF package in order ease data analysis since the difference in QE between the Hamamatsu CCDs is greater than has previously existed between any of the CCDs in either GMOS focal plane.
- Barring further unanticipated delays, the new focal plane is expected to be installed in GMOS-N at the beginning of semester 2011B. The scheduled down-time for GMOS-N is ~6 weeks. For this reason the call for proposals limits the RA range for GMOS-N proposals, see the 2011B call for proposals for more information.
September 15, 2010
Update on the Hamamatsu CCDs for GMOS-N: Efforts to address controller and noise issues have progressed at HIA, and have now confirmed that one of the science CCDs has a readout problem. One of the amplifiers is unresponsive, which means that one quarter of the columns in the CCD mounted in the middle of the focal plane assembly cannot be read out. Teams at HIA and Gemini are formulating strategies for how we proceed, which include replacement of the faulty CCD and/or repositioning it in the focal plane array to minimize science impact until a replacement CCD is available. Inevitably this does mean that the CCD replacement project is again delayed, and it is quite difficult to anticipate when a science quality focal plane will be ready for commissioning. Currently we estimate that the GMOS-N CCDs replacement will happen no earlier than very late in semester 2011A. Upgrading the GMOS-N CCDs with high QE red-sensitive devices remains high among the Observatory priorities. Every effort is being made to bring the task to a conclusion that meets or exceeds the scientific community expectations. We will issue another update when we have more information and a better idea of the likely schedule.
August 06, 2010
The CCD project continues to encounter hardware difficulties, involving various detector controller components and issues with the CCDs themselves. Testing continues at HIA, but it unfortunately is no longer possible for Gemini to support an upgrade of the GMOS-N CCDs in semester 2010B. The next possible window of opportunity to swap the CCDs is in January 2011, with on-sky commissioning taking place in mid-February. We will know better the likelihood that we can take advantage of this window after the various hardware options can be further explored. We plan to issue an update in a couple of months.
June 06, 2010
The CCD testing continues at HIA, and good progress has been made on several electronics issues. However, the delivery schedule has slipped, and we now expect we will be commissioning the new CCDs for GMOS-N in the first half of November.
February 04, 2010
New broad band Z and Y filters have been ordered for GMOS-N. These filters were chosen to match the filters currently installed in the UKIRT Wide Field Camera (WFCAM) and will enable imaging and band limited spectroscopy in the new far red wavelength regime made available by the new red-sensitive CCDs. We anticipate being able to offer these filters to the community for science use in semester 2010B, contingent on the delivery schedule. More information is available here.
January 23, 2010
New red sensitive deep depletion CCDs manufactured by Hamamatsu Photonics have been purchased for GMOS-N. The detectors are currently in Victoria, BC Canada being integrated into a new focal plane array at HIA. Delivery of the new focal plane array is expected at Gemini in June 2010, with commissioning scheduled for August 2010. We anticipate offering these new devices for science use in semester 2010B with the understanding that schedule delays may result in some science programs not being executed (particularly those exploiting the enhanced far red sensitivity). The popular Nod and Shuffle mode will still be available with the new detectors. More information is available here.
May 04, 2009
A new B600 grating has been successfully commissioned and is now being offered for queue and classical observing with GMOS-N. Users of the Gemini IRAF package should update their GMOSgratings.dat file. The full announcement is available here.
January 27, 2009
The original B600 grating delivered with GMOS-N has been removed from service as it has recently been damaged. A replacement grating will be ordered. The full announcement is available here.
December 12, 2008
The GMOS Photometric Standards web page was updated on November 28. The photometric zero points for GMOS-N and GMOS-S were updated. The web page includes accurate zero points and color terms values for each of the GMOS South CCDs.
October 25, 2007
Gemini is now providing GMOS users the option to design MOS masks without requiring GMOS pre-imaging of the field. This capability is offered at your own risk, and is not recommended for use by all programs. For MOS programs using slits narrower than 1.0" or for programs requiring very long observations of faint targets, Gemini recommends pre-imaging of the field with GMOS prior to designing the MOS mask. Please see the GMOS Multi-Object Spectroscopy webpages for an overview. Detailed instructions for MOS design using object catalogs are available. Please be sure to read the recommendations to ensure good mask design when using object catalogs. We thank the UK NGO office for its assistance in enabling this enhanced MOS capability.
October 23, 2007
Attempts to remove the GMOS-N flatfield features have not been successful. We are waiting for a suitable opportunity in which to more aggressively attack the situation, but that might require significant amounts of GMOS-N downtime which is not easily accommodated. Until then, users of GMOS-N imaging data should continue to employ twilight flats taken as close in time to their science imaging data as possible. The flatfield features do appear to divide out nicely and do not impact the sensitivity of GMOS very much or over a large area.
September 24, 2007
The GMOS Mask Design Instruction web page was updated on UT September 10. New instructions for PIs about the standard naming convention for mask submissions have been added. A link to the mask design check document used by the National Gemini Offices for mask design checking has been also included. The PIs are welcome to read these instructions to check and improve the mask design.
March 20, 2007
The GMOS-N flatfield features have recently changed in appearance. Observers reducing imaging data obtained since UT February 27, 2007 should take care to employ twilight flats also taken since this date. Attempts to clean the optical surfaces within the GMOS-N camera have not resulted in changes in the feature's appearance, yet the shape changed noticeably after a power outage caused the dewar to warm up. We believe this confirms our theory that the feature is actually the result of ice located on the dewar window. During the upcoming Gemini North shutdown period (March 26 - April 2, 2007) GMOS-N will be warmed up and attempts will be made to remove any contaminants from the dewar. We are optimistically expecting this new flatfield feature to therefore disappear. Watch this space for updates in early April.
December 22, 2006
The acquisition sequences for longslit observations has been updated for both GMOS (N/S). For 2007A, the longslit acquisition sequence for all Baseline and Program standard stars (flux standard, velocity standard, like-standard, telluric, etc) should use a ROI Central Stamp (300x300 unbinned pixels) to image the field, to measure the slit center and to confirm if the target is within the slit. Similar modifications were introduced in the longslit acquisitions for science observations. For Point Source, the ROI Central Stamp should be used to measure the slit center and to confirm if the science target is within the slit. For the extended objects, double source, and off-axis sources, the ROI Central Stamp should be used only to measure the slit center. We recommend to use the updated GMOS OT library as a source of example for these observations. See also the GMOS OT Helpful Hints web page for detailed information.
December 22, 2006
The GMOS-S World Coordinate System has been updated on UT November 28, 2006. With the new WCS, we have improved the large offsets ((RA~5", DEC~4") which has been presented in the the images. The WCS still has offsets of the order of ~0.8"-1.0". Users, please send your feedback about the accuracy of the new GMOS-S WCS since Novembre 28, 2006.
September 12, 2006
On UT September 9, 2006 the GMOS-N World Coordinate System was updated. The WCS is still only a first order correction (pixel scale and rotation) but we believe we have improved the large (~5") offset which has been present in GMOS-N images prior to this and the pixel scale has been improved. The WCS will still have offsets which should be on the order of ~1", we believe these errors originate not from the WCS calibration but from remaining errors in the OIWFS probe mapping (steps to improve the GMOS OIWFS probe mapping are continuing for both GMOS-N and GMOS-S - watch this space for updates). The WCS for GMOS-S is expected to be similarly updated shortly. We are interested in feedback from users as to the accuracy of the GMOS-N WCS since September 9, 2006.
August 30, 2006
Since the recent Gemini North shutdown new flatfield features have appeared on GMOS-N. Observers should take care to employ the correct Twilight Flats in order to remove these features from their data. We are investigating their removal and will post updates when available.
August 30, 2006
We are in the process of updating the GCAL configurations and exposure times for GCALflats and CuAr calibrations. The current table will be replaced with the new table including example spectra and more possible GMOS configurations once we have completed this task.
August 30, 2006
We have recently discovered that GMOS spectra suffer from scattered light which we believe originates from the classically ruled diffraction gratings. Information as to how much scattered light as a function of wavelength for each grating will be made available here as we progress with the characterization.