EUROPEAN SOUTHERN OBSERVATORY
La Silla Observatory
SCIENCE OPERATION DEPARTMENT
| 3.6m Telescope M2 Upgrade:
Minutes of the Final Design Review Meeting
| 2003-06-04 first draft, ohainaut
||2003-06-11 release after review, ohainaut
- 1.- Introduction
- 1.1- Purpose and Scope
- 1.2- Stylistic Conventions
- 1.3- Acronyms and Definitions
- 1.4- Documents and References
- 2.- Minutes
1.1- PURPOSE & SCOPE
This document contains the minutes of the 3.6m telescope f/8 M2 project Final
Design Review (2003-06-03).
More than a formal FDR, this meeting was a progress report of the project,
as many items are already completed or well into construction phase. This document
also makes references to the "Delta Review" that took place on 2003-05-06.
1.2- STYLISTIC CONVENTION
1.3- ACRONYMS AND DEFINITIONS
SciOp: La Silla Science Operation Department
LED: La Silla Engineering Department
1.4- DOCUMENTS AND REFERENCES
-  3P6-SPE-ESO-36022-1: f/8 top-end overhaul: user requirements, Sterzik
-  LSO-PLA-ESO-nn-mmm-zzz: 3p6 Telescope M2 electronics control, Macchino
-  3P6-TRE-ESO-80030-0003: Software Feasibility Report, Kastinen
-  3P6-SFS-ESO-80030-0004/1.3: Software Functional Specification, Kastinen
The meeting took place at LSO and Vitacura on 2003-06-03, 08:30 till 10:30.
It was attended by G.Andreoni, I.Kastinen, J.Alonso, W.Eckert, R.Parra, O.Lavin,
N.Goncalvez at La Silla, and F.Franza, G.Ihle (Chair), G.LoCurto, A.Gilliotte
and O.Hainaut in Vitacura.
G.Ihle reported on the mechanical design of the upgraded M2 unit, and on the
advancement of the work. This was done in detail to the benefit of F.Franza,
guest from ESO/HQ and specialist in opto-mechanic design.
- the design is modified compared to that of the NTT. Care was taken not to
have backlash, the rotation of the mirror is prevented by guiding sliders.
- The full range of motion is +/- 20mm, i.e. much larger than the +/-6mm in
UsrReq, to ensure that the actual position of the focus will not be too close
to one of the limits. The full encoder resolution is available over the whole
range. FF notes that the init/limit switch(es) could be made adjustable, so
that the whole range does not need to be scanned at each init. Nevertheless,
IK point out that we have to make sure that the mirror is never actually beyond
its init switch. GIH will investigate this.
- The system is balanced, although it is designed in such way that it will
work with significative unbalance (this could potentially be used in case
some weight has to be removed).
- The design is directly inspired by that of FF for the NTT, with modification
to work with modern motors.
- Most parts are manufactured, only the XY support flange in under machining
afer its stress release by heat treatment process..
- At previous meeting, there was some discussion about the best configuration
for the Pantograph, i.e. parallel or conical. After investigation, it conical
is much better, indeed it permits corrections without changing the pointing.
A parallel pentograph would correct the coma in a similar way, but would introduce
a collimation and depointing error.
- The legs of the pantograph should match the radius of curvature of M2.
This is mechanically not possible bc of space constraints. Therefore, they
are as long as mechanically possible. This will introduce a small depointing
error, that shall be corrected by the pointing model, and refined on-the-fly
by the auto-guider.
- The current spider is operated with pre-stress. As this pre-stressing is
performed using only the fixation bolts, it is inaccurate, resulting in an
uncontrolable system. As the upgraded M2 will be installed on the current
f/8 top ring, this cannot be completely solved. However, as the new system
will permit control of the coma (either closed loop, or at least open-loop),
this is not critical anymore.
- The thickness of the spider is 20mm, i.e. equal to the current one (10mm
+ 2x5mm reinforment plates).
- The design consider either "open" spider, or "filled".
The filling could be performed by installed plates in the triangular opening
(without causing additional light losses). The filled spider has slightly
better rigidity, but the sensitivity to wind would be much higher. It is therefore
decided NOT to install the filling plates and to work with open spider.
- GIh will design gutters to be installed
on the top part of the spider to guide the cables and cooling fluid pipes.
These gutters must be water-tigh, so that cooling fluid drips or water condensing
on the cooling pipes will flow to the edge of the top ring and not fall on
the mirror [question: when the telescope is inclined, some of these gutters
will point toward the M2 unit, not toward the top ring. What do we do with
water at that level??].
Construction under way - there were some delays caused by other projects interfering.
GAn will take care of assigning priorities
in order not to cause unacceptable delays.
The outer body is designed with some stiffening rims, that can take the load
of the current full preload of the spider while remaining within specs.
Cooling of the top control unit:
- A full LCU is located on the top of the M2 unit. As it uses standard VLT
components, it dissipates significant heat, so that a cooling unit must be
installed. This will be a standard VLT cooling unit, with liquid coolant.
The total dissipation is ~500W, roughly distributed as following: cooling
unit: 150W, fans: 90W, control electronics and drives: 200W. JAl remarks that
a non-VLT compliant system could have been designed with very low power consumption,
and that the VLT system is "an atomic bomb to kill ducks" for what
we have to do, but this is the price for full VLT compliance.
- The presence of cooling fluid on the M2 unit is a source of concern. Some
- the only joint will be at the Top Ring to Serrurier connection. These
must be auto locking joints of high quality (only one droplet of fluid
is lost when opening them).
- the temperature of the fluid must be controlled. For instance, VLT uses
a Delta-T of 6deg between the fluid and the air.
- Instead of insulating the pipes (which is problematic considering the
total budget of 20mm for the spider thickness), it may be worth to install
the pipes in the gutters on top of the spider, the gutter taking care
of evacuating the condensation water.
This upgrade will re-install the current M2 baffle. A secondary project will
take care of replacing it with an advanced baffle with air circulation (like
NTT). Indeed, the improvement of the 3p6m Image Quality is likely to make the
aberrations caused by the current baffle significant (while they are currently
lost in the noise of the other sources).
2.3- CONTROL ELECTRONICS
Note: The presentation of the control electronic by JAl at this meeting
was very short, essentially addressing the points that remained pending after
the 2003-05-06 meeting. This section includes part of the discussion that took
place at that previous meeting.
The control electronics is designed as fully VLT compliant, in a thermally
neutral enclosure. It consists in a harmonic motor - encoder units, forming
with the controller a closed velocity loop. The system includes precision limit
switch, backed up by heavy duty safety limit switches.
As mentioned before, the total dissipation is about 500W, requiring a cooling
system using cooling water. This, in turns, has some constraints on the mechanical
design (see above), and implies that the water temperature must be controlled
and adjusted wrt the dew point.
At the time of the previous meeting, the CPU, Maccon and amplifiers had not
been purchased yet. This has been done by the time of this meeting.
The motors had originally been purchased equipped with brakes. However, the
presence of these brakes does not make sense in the closed loop design used,
therefore the motors had to be returned to the manufactor for removal. This
has been done. GIh will track the motors
and make sure they are not stalled in some remote storage.
At the previous meeting, NGo presented detailed simulations of the full control
system, in order to study the precision achieved in positioning the mirror.
Indeed, as the position of the motors (as opposed to that of the M2) is measured,
this is critical. It appears from these simulations that damping of the system
is critical in order for M2 to reach its position in reasonible time and not
to oscillate around that position.
2.5- VARIAE and SUMMARY OF ACTIONS
- GIh (with AGi) will prepare procedure to
center the new M2 unit in the top ring while the top ring stands on the
dome floor. This includes the preparation of required tools.
- Emilio Barrios (with help of IKa and AGi
if needed) will optimize, streamline and automatize the aberration measurment
- GIh will communicate the final schedule to OHa before the end of June,
so that this info can be communicated in due time to J.Breysacher for schedulling
Actions listed earlier:
- GIh will investigate the possibility of having adjustable init switch for
the focus, to restrict the range available in operation.
- GIh will design water tight gutters for cables and water pipes, to be installed
on the top edge of the spider legs.
- GAn will keep track of the work assigned to the workshop in order not to
delay this project.
- GIh will trak the motors and make sure they return to la Silla in due time.