Date: Mon, 6 Sep 1999 17:40:02 -0700 (PDT)
From: Ted Tarbell <tarbell@sag.lmsal.com>
To: Rock Bush <rock@quake.Stanford.EDU>
Cc: John Beck <beck@quake.Stanford.EDU>, mdiers@mdisas.nascom.nasa.gov
Subject: Re: Recent detunes (fwd)


I analyzed the 6-Aug and 1-Sep detunes, and they both look normal 
and give consistent results.  They show that Michelson #2 (the
broader one) continues to be stable while Michelson #1 continues
to drift in wavelength.  Now the tuning positions in the continuum
are set to (M1, M2) = (24, 131).  The detunes show that to put them
together in wavelength (cotuning) requires a setting of (~22.5, 131).
This is not defined precisely because the gradients are large and 
because there are small differences between CAL and OBS modes and 
between disk average and disk center values.  Even so we should use 
this occasion not only to take out the large orbital velocity but also
to cotune the Michelsons better.  If we set the tuning to (22, 131),
then the disk-averaged velocity on September 1 would become about 
V = -70 m/s.  The partial derivatives of this value are about 
dV/dM1 = -125 m/s and dV/dM2 = -160 m/s.  This slightly overcompensates
the drift of M1, so we would have good cotuning for the next 6 months.

On September 1 when the detunes were taken, the disk averaged velocity
<V> was about <V> = -320 m/s and the orbital velocity (judging from 
previous years) was about -400 m/s.  The observed velocity range 
around <V> is about +3500, -2500 m/s; average disk center and average 
CAL mode velocities are about <V> -250 and <V> - 290 m/s, respectively.
So it is reasonable to keep <V> around 0; it should never get more 
than a few hundred m/s positive, since the velocity measurement goes 
bad rapidly for |V| > 3500 m/s.

Orbital velocity will drop to about -550 m/s in late October, then go 
back to 0 in late December.  So the (22, 131) tuning would give us 
<V> > 300 m/s and rising rapidly in January, time for another tuning 
before dynamics.  On the other hand, we could set the tuning to 
(23, 131) now, which would give <V> ~ -300 m/s on September 1, -450 
in late October, and > +300 in February, postponing the need for 
another tuning by a month.  This doesn't seem better, and it also 
has slightly poorer cotuning.  Any other tuning has even worse 
cotuning and/or greater extremes in <V>.  I think the optimum times of 
year to retune would be about a month after perihelion and after
aphelion; so maybe we'll do it right in January.

So I recommend the new tuning of 

MTM1   FC = 22
MTM2   FC = 131  (formerly 41)

Ted