[mmaimcal] Re: ALMA requirements
Bill Shillue
bshillue at nrao.edu
Wed Jan 11 10:38:29 EST 2006
Jeff et al,
John Conway refers to a draft ICD. I think he means draft CRE. The
change request was a change to the System Requirments to allow for more
phase drift for "large" antenna motions. This was for supporting
baseline calibration, but recognizing that the large motions will have
some additional residual phase errors. My understanding is that this CRE
was dropped this past Summer and instead the issue was wrapped into a
complete rewrite of the Systems Requirements doc, which was being
pursued for many other reasons. I think this effort is still in
progress. (Dick Sramek is cc:ed to verify).
The thread on minimum required interval between line length corrector
reset has been very useful. I will adopt a design for something longer
than one hour. I'll be sending some more info on this soon.
Bill
Jeff Mangum wrote:
> Hi John,
>
> John Conway wrote:
>
>> Hi,
>>
>>
>> Please remember that 'normal' astronomical oberving is not the only mode
>> one has to design for - one important requirement is the ability to do do
>> observations to solve for antenna positions after antennas are moved.
>> This requires 'geodetic' observations involving rapidly switching
>> between
>> calibrators all over the sky, During this process it is vital that
>> the instrumental delay stays relatively stable.
>>
>> In memo 503 it was estimated that this antenna calibration process
>> would take 30minutes -60minutes using the four antennas that had been
>> moved plus a couple from the rest of the array. I think therefore
>> from the point of view of antenna position determination setting
>> the requirement for these 'stretcher' induced delays at tens of minutes
>> is therefore probably too loose and 60 minutes seems a safer target.
>>
>>
> Then the stretcher reset interval should be greater than 60 minutes.
>
>> John
>>
>> P.S On a more general front the information about antenna based
>> changes in
>> delays of tens of microns on timescales 3 -30 minutes is new to me, and
>> assuming this 'stretcher' reset is on large enough timescale not to be a
>> problem this may set the accuracy to which antenna positions can be
>> found.
>>
>>
> Note that the antenna-based pathlength variations over this period will
> be ~50 microns or so, but they are a mixture of repeating and
> non-repeating components, so this value is a worst-case for the antenna
> pathlength stability.
>
>> There was also some disussion about one year ago about the level of
>> Azimuth induced phase errors due to cable wrap (another critical area
>> for antenna position determination), a goal was determined and ICD
>> written. Has this goal been achieved, what is the expected level of Az
>> induced delay errors and what are the implications for antenna position
>> determination? Maybe it is time again to look again at the position
>> calibration issue. Unfortunatley I am no longer funded to work on
>> ALMA - so I guess someone else has to be assigned to look at this
>> problem.
>>
>>
> Can you send me the draft-ICD and other background material? Thanks.
>
> -- Jeff
>
>>
>> On Tue, 3 Jan 2006, Al Wootten wrote:
>>
>>
>>
>>> Recall that Jeff will be speaking this pm at 4 on the AEG prototype
>>> antenna
>>> tests. Jeff pointed out to me that the antenna is supposed to be
>>> stable to
>>> variations which could result in path length changes on timescales of
>>> 3-30 minutes or so. The AEG tests show that there are changes of tens
>>> of microns over these timescales. Thus, I think we would want to do an
>>> instrumental calibration on timescales of tens of minutes. That's in
>>> accord with our previous comments.
>>>
>>> For further details, tune in at 4pm.
>>>
>>> Clear skies,
>>> Al
>>> Al Wootten writes:
>>> > Hi Bill
>>> >
>>> > If it is more than one phase wrap I think this would not be a good
>>> thing
>>> > to do. I gather from what you write that the number of phase turns
>>> could
>>> > be deduced from the before and after stretcher voltage. Perhaps this
>>> > should only be done between schedule blocks--not a well defined
>>> interval
>>> > but probably about a half hour or so. It would be good if disaster
>>> didn't
>>> > strike if it had to be an hour.
>>> >
>>> > I see Mark and Darrel just sent a note with a similar comment, though
>>> > more detailed.
>>> >
>>> > Clear skies,
>>> > Al
>>> > Bill Shillue writes:
>>> > > Right. thanks for reply.
>>> > >
>>> > > So I think the question is, can I reset some or all of the
>>> stretchers at
>>> > > every instrument cal? In so doing, there is a *big* phase jump.
>>> The
>>> > > phase jump will be different for every antenna, between zero and
>>> maybe 5
>>> > > millimeters (many fringes). The exact size of this phase jump
>>> will not
>>> > > be known exactly but should correspond well with the before and
>>> after
>>> > > stretcher voltage, maybe to within ten microns.
>>> > >
>>> > > Bill
>>> > >
>>> > >
>>> > >
>>> > >
>>> > > Mark Holdaway wrote:
>>> > > > Bill Shillue wrote:
>>> > > >
>>> > > >> Mark, Darrrel,
>>> > > >>
>>> > > >> I have a question that nobody seems interested in, and which
>>> thought
>>> > > >> one of you might be able to help with.
>>> > > >>
>>> > > >> Short version:
>>> > > >>
>>> > > >> How long do the 1st LO reference line correctors need to stay
>>> active
>>> > > >> and in-range?
>>> > > >>
>>> > > >> Long version (sent to a list 3 weeks ago and no reply)
>>> > > >>
>>> > > >> I cannnot find reference to a requirement on the LO phase
>>> stability
>>> > > >> for time periods greater than 300 sec. And yet I seem to
>>> remember
>>> > > >> Larry D saying, "Oh no, you cannnot rely on calibration to
>>> take out
>>> > > >> all phase drifts and thus you need to make the LLC range
>>> cover much
>>> > > >> longer time period, ideally forver."
>>> > > >>
>>> > > >
>>> > > >
>>> > > > Bill,
>>> > > >
>>> > > > I cannot fully respond to this at this moment -- I am busy
>>> preparing for
>>> > > > URSI today.
>>> > > > I probably don't understand the issue sufficiently to issue this
>>> > > > statement -- but we
>>> > > > were counting on doing an instrumental calibration every 5-10
>>> minutes.
>>> > > > You should
>>> > > > probably ask the horse himself.
>>> > > >
>>> > > > -Mark
>>> > > >
>>> > > >> I need a NUMBER from 300 to infinite seconds!!! What is
>>> it!? This
>>> > > >> requirement will impact our final LLC stretcher design. We
>>> cannot
>>> > > >> make further progress without it. I am concerned that if we
>>> pursue
>>> > > >> this then someone will say, "yes, forever sounds good, please
>>> design
>>> > > >> for that." But that ignores that more stretch means more
>>> cost, greater
>>> > > >> size and complexity, and possibly worse performance. So I need
>>> > > >> someone at Science/Systems/Cal to really think this through and
>>> > > >> determine the frequency that we will be allowed to reset the
>>> LLCs.
>>> > > >>
>>> > > >> One more thing - if we decide that 30 minutes is suitable,
>>> then I need
>>> > > >> to design so that under expected worst case environments
>>> (sunrise on
>>> > > >> Chanantor) - that I have enough dynamic range that all EIGHTY
>>> line
>>> > > >> correctors will stay in range with some level of certainty.
>>> > > >> Otherwise, astronomers will get messages like this
>>> > > >> "Warning : line length corrector #31 just went out of range"
>>> > > >> The going-out-of-range can easily be accompanied by a RESET
>>> to the
>>> > > >> mid-range but there will be a phase jump at that moment in
>>> time and
>>> > > >> the size of the phase jump will probably have an uncertainty
>>> of ten
>>> > > >> microns.
>>> > > >>
>>> > > >> Bill
>>> > > >
>>> > > >
>>> > >
>>> > _______________________________________________
>>> > mmaimcal mailing list
>>> > mmaimcal at listmgr.cv.nrao.edu
>>> > http://listmgr.cv.nrao.edu/mailman/listinfo/mmaimcal
>>> _______________________________________________
>>> mmaimcal mailing list
>>> mmaimcal at listmgr.cv.nrao.edu
>>> http://listmgr.cv.nrao.edu/mailman/listinfo/mmaimcal
>>>
>>>
>>
>> _______________________________________________
>> mmaimcal mailing list
>> mmaimcal at listmgr.cv.nrao.edu
>> http://listmgr.cv.nrao.edu/mailman/listinfo/mmaimcal
>>
>>
More information about the mmaimcal
mailing list