[evlatests] L352 RTP meters update.

[Vivek Dhawan]

Summary and question:

We have many weeks of field tests, and a few days of lab tests. So I
think we should go back to lab testing of the newest meters before we
buy in bulk. Opinions?

The pros and cons are:

Evaluating the meters in the field is difficult, because the antenna
environment is not well controlled. The interferometer phase variation
contains effects not measured by the RTP meters, and so cannot be used
to verify them.

However, we want the whole system to work in the field. So finding the
limits, set by components inside or outside the meters, is important
in the long run.

------------------------------------

On Aug 10th there was a baseline run that I examined in detail. I am
trying to write clearly, believe it or not.

1. As previously noted, the L352's register daily changes of about
   50ps from the temperature response of the fiber. In addition, there
   are shorter variations of a few ps in meter reading, that may or may
   not be real changes in fiber delay.

2. The interferometer phases show large variations with elevation, (at
   some antennas), separate from the diurnal fiber change, and over 10
   times larger than explained by the RTP meter readings.

3. It turns out the airhandlers were cycling the temperature by
   several degrees, in a ~30min cycle, and then stopped doing it just as
   the baseline run began. So I could see the separate effects of
   temperature and elevation. The RTP meters change by +_1 ps, as the
   rack temp varies (+-0.5C to +-2C, depending on location within rack),
   OR the elevation changes (8 to 120deg).

4. The temperature of some modules (monitored by the MIBs) changes as
   the antenna is tipped in elevation, probably from changes in airflow.
   However, the many temperature monitors installed in the EA13 rack by
   Doug Gerrard show little change with elevation.

5. The 2 RTP meters both monitoring EA16 agree, i.e., their difference
   is well within spec, for many hours at a time, BUT they diverge from
   each other a few times a day, rapidly over ~10min, sometimes by
   several ps.  This is out of spec. The cause is not understood, but 2
   possible causes can be eliminated:

   The temperature of the L352 units themselves is probably NOT a
   cause, since the other 2 units for EA13, 14, are in the same rack
   in the control building, but have a different pattern from 16a,b.

   The readings of 16a,b change with elevation but agree with each
   other at all elevations in the baseline run.

6. To check if the RTP meter difference follows the rack temperature,
   we could move the temperature logging to EA16 from 13.

----------------------------------------------------------------------
Monitor point    Variation    Varies with or     Comments
or Variable          +-       is caused by
----------------------------------------------------------------------
Interferometer   +-50 psec    Elevation/Temp     EA13: +_50ps
phase                             etc, ?         EA14,24: > +_50ps
                                                  EA18,16: < +_5.ps


Rack air temp    +-0.5 C to   Does NOT change    Various log points.
EA13:   Doug     +_2 C        with elevation
Gerrard's data                ( < 0.2 C)


F320 module      +_2 C        Changes w. ELEV.
temperature                   NOT with rack temp.


L305 Analog      +_1 C        Changes w. ELEV.
board temp.                   NOT with rack temp.


L352 RTP         +_1 ps       Changes with ELEV.   (After large diurnal
reading                       ALSO w. rack temp.   trends from the fiber
                                                    itself are removed)
--------------------------------------------------------------------------

Vivek.