|
Hi Chris,
As a rule of thumb with regards Tpoint CH error, which is what you have,
you want to ensure that it is mechanically about under one degree to
be modelled.
Depending on what your pointing goal is, you would be starting to
push your luck if your CH error is greater than about 5 degrees.
During a pointing run, it is important you take samples across
a whole range of zenith distances and on both sides of the meridian.
It will really pop out on the graphs the first time you do a meridian flip.
It goes without saying that the source of the error can't be random.
If those couplings shift wrt each other, all bets are off.
However if the two are rigidly affixed with respect each, you have
an excellent chance of nulling CH out.
|