From: Bobby Bus
to: Anthony Denault
date: Fri, Feb 27, 2015 at 7:24 AM
subject: MORIS initial guidebox positions
Hi Tony —
I finally fitted the MORIS data I have that measures how the MORIS guidebox
shifts when the SpeX rotator is moved. The first step in improving the way we set up
MORIS for guiding is to include these functions, and provide an “Auto GuideBox Setup”-like
button on MORIS that queries the SpeX rotator motor step position (which ranges
from 0 to 1439999), and calculates where the best start position for the MORIS
guidebox should be.
Here are the function parts:
mpoff_x = 311.5
mpoff_y = 278.13
rs = rotator_step_pos / 100000;
mpx = 2.3026e-08*rs^11 - 1.8861e-06*rs^10 + 6.1438e-05*rs^9 - 0.00099381*rs^8
+ 0.0078916*rs^7 - 0.022528*rs^6 - 0.017011*rs^5 - 0.27057*rs^4
+ 2.3158*rs^3 + 4.091*rs^2 - 19.098*rs - 20.483 + mpoff_x
mpy = -2.8783e-09*rs^11 - 2.065e-07*rs^10 + 2.3275e-05*rs^9 - 0.0007359*rs^8
+ 0.010819*rs^7 - 0.075941*rs^6 + 0.1874*rs^5 + 0.16245*rs^4
+ 0.91093*rs^3 - 7.6503*rs^2 - 12.024*rs + 20.256 + mpoff_y
Yes, these are 11th order polynomials… (sorry about that, but they fit the data nicely.)
mpoff_x and mpoff_y are offsets that could change if MORIS is taken off of SpeX and them
reinstalled. These are numbers that can be remeasured as needed, or tweaked
if they seem to be off. Therefore, if these values could be put in the “Engineering” part of
the GUI, so that they could be easily changed if needed, that would be good. The rest of the
coefficents are stable, I think, unless we change the MORIS camera itself (get a new camera
for example).
To do the fit, I needed to shrink the magnitude of the motor step position values, hence for the
calculations, the true step value needs to be divided by 100000. mpx and mpy are the pixel positions
on the MORIS array that we want to place the guidebox center. I think you already changed the
software to allow fractional pixel positions - if not, then just round the mpx and mpy values to
the nearest integer.
This is the first step. After this is implemented, I want to think about ways to improve
how we fine-tune the MORIS guidebox positioning to put the target down the slit as
viewed in GuideDog. I think this can be done with a GuideDog macro, but if we can get
this first step (the functions above) implemented and tested, then we can look into how to
improve efficiency on the rest.
I’ll check in with you later on this, but if you have any questions up front, please let me know.
Thanks — Bobby