SYSTEM DESIGN NOTE

SDN0004 - Review of On-Instrument Wavefront Sensor (OIWFS) Optics in GNIRS

Prepared by	Date	Approved by	Date	Rev.	Rev Date
Jay Elias	2/19/99	N. Gaughan	3/4/99	I	3/4/99

1. Introduction

Ming Liang was asked to carry out two investigations of the On-Instrument Wavefront Sensor (OIWFS) optics in GNIRS. These were an examination of the temperature sensitivity and an examination of the effects of the GNIRS dewar window, which acts as a field lens for the Offner relay.

For the investigation of the effects of the GNIRS window, we told Ming to carry out the analysis without varying spacings of elements within lens groups, where the 7 elements preceding the detector were considered a group, and the 6 elements immediately before and after the gimbal mirror, plus the gimbal mirror itself, were also considered a group.
The reason for this was to determine whether the performance would be acceptable using the IfA-designed lens mounts. The performance in the NIRI configuration was also calculated as a check.

For the investigation of the temperature sensitivity, the IfA Zemax file for the "cold" design was used, with all temperature coefficients "as is". This procedure is certainly incorrect in detail, in that the elements mounted on the detector can be presumed to have their temperatures controlled. Also, Ming says that the coefficients that were in the file
appear incorrect, and may not have been intended to describe differential temperature performance at 65K. (Buzz Graves told us during our visit that he had not performed calculations of temperature sensitivity, as NIRI's workbench temperature is being actively controlled.) In fact, there appear to be no coefficients at all for most of the lens materials.

2. Results

GNIRS Window

When the GNIRS window replaced the NIRI window, the main effect was to create a pupil image ahead of the gimbal mirror rather than right at the gimbal mirror. It was possible to maintain focus on the detector while maintaining good image quality.

Specifically, Ming evaluated the performance with the gimbal mirror positioned at the guide field center, the 3' field edge, and halfway. The RMS image radius for the three
positions was as follows:

Ming evaluated a 10" x 10" FOV at any given gimbal position; the quoted RMS values are just for the field centers.

There are two effects of the pupil not being located at the gimbal mirror. First, if the gimbal counterweight is sized exactly to the NIRI ray bundle, there will be vignetting in GNIRS. A rough calculation based on footprints on the gimbal mirror indicates a maximum effect of 6%; it will be less if the counterweight is at all oversized. Second, the mapping of gimbal tilt to sky location will be somewhat different than for NIRI. (Ming suggests that the relation between tilt and position will not be exactly linear for either instrument.)

The pupil image could be brought back to the gimbal mirror by moving the OIWFS field lens a few cm, but Ming states that the image quality on the detector is no longer satisfactory, unless adjustments are made within the lens group ahead of the detector.

Temperature Sensitivity

The imaging performance on the detector was evaluated for temperature shifts of + and - 10K. There is substantial defocus, with the resulting geometrical image diameter of almost 100 microns (actually, about 71 ?m for +10 degrees, and 107 ?m for –10 degrees). In fact what would happen is that the OIWFS would "correct" the focus of the telescope so as to bring itself in focus, thereby producing defocus at the spectrometer slit. The resulting blur at the spectrometer slit would be ~500 microns, which is clearly unacceptable. An improved analysis is not likely to show performance >10x better (recall that the standard slit for the long cameras is 62 microns width).

3. Options

GNIRS Window

With regard to the pupil image, there are several options. The first is to do nothing. If the gimbal mirror counterweight is at all oversized, the overall effects on performance will be small, and probably acceptable. A different conversion of gimbal tilt to position will be needed, but this would almost certainly be required already simply because the layout will be somewhat different in GNIRS than in NIRI.

The second would be to examine the drawings for the lenses at the detector to determine whether appropriate adjustments are feasible; if so, the adjustments required after moving the WFS field lens could be calculated and then made. This would require some interaction with IfA (Jeff Douglass and Buzz Graves).

The third option would apply only if the second were not viable, which would be to modify the detector lens mount design to accommodate the calculated shifts.

A fourth option is to design and fabricate a new field lens.

Options 2-4 require additional effort by NOAO, with some help from IfA. Since it appears that the first option – “do nothing” – produces acceptable performance, this appears reasonable as a default. If option 2 can be implemented with minimal effort, it might be worth doing. In order for this to be feasible, though, the drawings must be available and the analysis must be done before the layout proceeds much further.

Temperature Sensitivity

The only temperature-sensitive optical system in GNIRS itself is the camera (actually, all four). Although at one point thermal control was discussed, the review committees recommended use of a focus drive instead, which is being done. However, the substantial thermal sensitivity of the WFS creates a problem. There are, again, several potential solutions.

Note that the performance "as is" is not likely to be acceptable as it is highly unlikely that instrument temperature can be passively maintained to ~1K under all conditions. A particular concern is that alignment will be done under much warmer conditions than occur on the summit of Mauna Kea, and the instrument will be operated cold for lengthy periods. The instrument will therefore tend to be systematically colder during use than during alignment.

One option is to calculate (using correct temperature coefficients) the defocus with temperature. These calculations would be verified during instrument testing. This information would be supplied as a look-up table or as a function to the A&G unit, which could then apply a focus correction when supplied with the instrument temperature. This solution is likely to result in degraded performance for larger deviations in temperature, because at that point the image quality on the WFS array is degraded; for smaller variations it should be adequate.

Another option is to adjust the WFS detector focus position as a function of temperature. This again requires a careful calculation and a check in the laboratory. It is not clear whether this is best done by the OIWFS with corrections communicated by GNIRS, or whether it makes more sense for GNIRS to control the OIWFS focus stage directly.

A third option is to control the bench temperature much as NIRI does and much as GNIRS planned to before the focus drive was implemented. One issue is whether the ±1K control achieved by NIRI is truly adequate for GNIRS; NIRI can after all compensate using science detector focus for most applications, whereas GNIRS cannot. Again, a more complete analysis would answer this question. Examination of the IfA-supplied file suggests that a proper analysis will show less temperature sensitivity, so that
±1K control would be adequate. All of these options require additional effort by NOAO, and all but the last require additional effort by other groups.

The analysis shows that some means of compensating for temperature sensitivity is needed. The pros and cons are discussed below:

Solution a: applying a correction to telescope focus, based on bench temperature.

Pros: does not require additional thermal/mechanical/electronic work.

Cons: probably most complicated solution in terms of software interfaces. The performance of the OIWFS degraded for large temperature excursions because image quality of WFS detector gets worse. Optical analysis required plus validation in lab for completed instrument; latter is likely to require extended operation in environmental chamber.

Solution b: adjusting focus of WFS, based on bench temperature.

Note that there are two possible ways to do this: direct control of focus by GNIRS and
passing focus corrections to OIWFS (solutions b1 and b2).

Pros: does not require additional thermal/mechanical work. Additional electronics (b1) is straightforward. Focus adjustment testable in lab. Correction for filter thickness (where small) also possible.

Cons: additional software, whether for b1 or b2. Same amount of optical analysis and validation as (a). Some degradation of performance because aperture stop will be out of focus as temperature departs from nominal.

Solution c: thermal control of structure

Pros: does not require adjustments for thermal effects in WFS or cameras. Performance somewhat more robust against degraded cryohead performance. Focus corrections do not require testing in lab. No interfaces to A&G or WFS for this purpose.

Cons: additional mechanical/thermal design/fab. Electronic/software effort for thermal control; some testing required. Can't adjust for filter thickness.

Discussion

Solution (a) does not appear acceptable: it provides reduced performance and may well take more work. Unless it turns out that we have to interface in software to the OIWFS anyhow (as opposed to modifying the code slightly for our geometry), solution b2 looks messy as well. The choice is then between solution b1 and solution c. The attraction of solution c rather than b1 is that any of the solutions involving correction imply an extended period of testing after final integration of the instrument; eliminating the effect is likely to be much more straightforward then correcting for it. Note also that this is then analogous to what NIRI is doing.
 

---

---

If you have any questions or suggestions regarding this website, please contact Melissa Bowersock.

NOAO Intranet Services

NOAO Copyright  Statement