I.Temperature stability in the MMF:

The figures 
2006-9-20_0-0-0_to_2006-10-4-_20-49-26-Undulator17Temps.jpg
2006-9-20_0-0-0_to_2006-10-4-_20-49-26-Undulator102Temps.jpg
show the total time history of all monitored temperatures 
for the last two weeks.
From 9/22 noon to 9/27 noon and 9/30 to 10/2 noon apparently
reasonable temperature regulation was achieved. 
This period is shown enmlarged in the two figures,
2006-9-23_0-0-0_to_2006-10-2-_23-58-7-Undulator17Temps.jpg
2006-9-23_0-0-0_to_2006-10-2-_23-58-7-Undulator102Temps.jpg
which
show the temperatures in the indicated 10-day period for the
Undulators 17 on the dover bench and 102 on the Kugler bench,
respectively

The temperatures shown are those of sensors at the
start(upstream)/middle/end(downstream) top half of undulator,
bottom half of undulator (in about the middle), as well as
(for undulator 17/Dover bench) on the granite of the bench.
Red shows tha air temperature near the undulator,
and yellow the air near the rack (sensor on post near south entrance).

A vertical bar on the bottom |---28 marks the beginning of a calendar
day ( the second | marks noon on that day)

The following can be noticed:
- All Dover bench temperatures are 20.1 to 20.25 deg C,
  that is 0.15 deg too warm. 
- On 9/23 to 25 morning, and 10/1 to 2 morning, the system
  regulates quite well, to +/- 0.03 deg C.

- 9/28-29 there is a dip, perhaps following some repair.
- The peak on 9/30 ~14:58 is due to single (mis-)measurement
  preceding an outage until 16:47 (network disturbance? problem
  accessing calibration files?)  

- There are various temperature differences:
-  For Undulator 17, Start middle bottom seems to be warmer by 0.1 deg
  than end, air near undulator(measured closer to end), and granite.
- All Kugler bench/Undulator 102 temperatures are about 0.1 deg colder
  than Dover bench temperatures.
  
To address the higher temperature of the Dover bench, Yuri had asked 
for a lowering of the setpoint. 
As shown in the temperature time history for the last 4 days,
2006-10-1_0-0-0_to_2006-10-4_21-19-27-Undulator17Temps.jpg
2006-10-1_0-0-0_to_2006-10-4_21-19-27-Undulator102Temps.jpg
the temperature is still dropping.
These figures also shows good regulation on Oct1 to Oct2 ~noon. 



II. Extraction of cooling times.

From the limited and noisy data it is difficult to extract a reliable
estimate for the time required until an undulator is in equilibrium
with the surrounding air.

The Figure 2006-9-28-drop17.jpg shows the temperature drop seen on
9/28, when the air temperature near undulator 17 fell from ~20.13 
to 20.08 deg C. (red curve; ignore 2 peaks at 11:30, 17:30, due to
people standing close), beginning at ~8 a.m, and reaching a new 
stationary value about 3 hours later.
The undulator 'end' temperature (green curve) began dropping 
~ 1 1/2 hours later, and reached the same value as the air 
temperature at ~18:00, that is about 10 hours later. 
  Considering that the air temperature drop occured over 3 hours,
one could estimate the time to equilibrium (on the surface, where 
measured) to be more than 7, and less than 10 hours.
  The other undulator temperatures also dropped, and came to equilibrium
in about the same time. Their equilibrium is however about 0.1 degree 
higher than that of the undulator 'end' temperature.
In Figure 2006-9-28-drop102.jpg the contemperaneous temperature drop
of undulator 102 on the Kugler bench is shown