@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
        @                                               @
        @    MTA Monitoring Report 3/24/00 - 3/30/00    @
        @                                               @
        @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@



             ************ CTI trend *************

No prominant changes are observed. All of them are steady.
CCD 7 Al Ka may have larger standard deviation.

>>>>>>>> Please read the memo from Scott Wolk below <<<<<<<<<<<<<<


     ************** MTA Problem Tracking. *****************

No new problems are reported this week; only same preoblems are 
keep coming:

EB1DI, EPOVSPR, TSAMYT, TSAPYT, EOESAC2C, EOESAC1C, AOSAILLM, EB2DI,
EB3DI, ESAMYI,AOPSAMUL, 1DPDBBON, ES1P5CV, ES2P5CV, ASPEB5DV, AGWS2V,
EB2CI, EB1CI,EB3CI, AIOBP5CV, C2SVMONB, CUSOB28V, CIUB5V, CIUB15V,
EB3


                ******** Trending  ***********


Although there are no new trend is observed, here are some trends
whcih we are seeing for a while.

* Acis Thermal: some msids show larger scattering for the last few weeks.
* Epin temp/electronics:  continue to show strong upwared trend.
* HRMA struts/cone/tfte: show mildly strong upward trend.
* ACA thermal: many shows upward trend
* IRU temp/electronics: AIRU1G1L show sudden drop.
* Spacecraft temp/electronics: larger scatter


          *********** ACIS Focal Temperature ********

Focal palne temperature is shooting upto around 108 C about every 2 days.
see http://asc.harvard.edu/mta_days/mta_fp/fp_week_temp.html

The peak and width are expected to be larger as seaon progresses.

             *************  Telemetry ***********

Battery Usage: EOEB1DIC showed a big jump from the end of 3/25 to the
early part of 3/26. The peak reached 25.63 (from usually 0).

One large serge is observed at the end of 3/27 to the first part of 3/28:
	* Spacecraft OOBAPWR
	* Spacecraft electronics Side B: ELBI
	* HRMA temperature related msid (esp. HRMA Cone)

             *********** Photon Page ************

Although there are about 40 new entries, following are obsid are
analyzed this week. 

OBSID	DETECTOR	GRATING	TARGET		ANALYSIS
---------------------------------------------------------
611     ACIS-I  	NONE    W3B             OK
872     ACIS-S  	NONE    NGC 6503        OK
539     ACIS-I  	NONE    Z5247           OK
859     ACIS-S  	LETG    NGC 4051        NO SKY
71      HRC-I   	NONE    REJ1236+47      WRONG INSTRUMENT
69      HRC-I   	NONE    REJ1236+47      WRONG INSTRUMENT
47      ACIS-I  	NONE    STAR C,CYG OB2  OK
62      HRC-I   	LETG    HZ43,2          WRONG INSTRUMENT
70      HRC-I   	NONE    REJ1236+47      WRONG INSTRUMENT
614     ACIS-S  	HETIG   FKCOM/HD117555  NO SKY
934     ACIS-S  	NONE    M101            NO SKY
715     ACIS-S  	HETG    GX 249+2        NO DATA POSTED
640     ACIS-S  	HETG    ZET PUP         NO SKY



        ************ Some Comments from Exparts ********


------------------------------------------------
From Scott Wolk (3/27/00) ----- Comments on CTI
------------------------------------------------
Hi Roger and Harvey,
I took an action last week at the CTI telecon essentially to answer the above 
question.  I will be unable to attend that telecon (Travel) and I wanted to 
give you guys a heads up on the situation.  It is not too good because we
probably will run into trouble with the focal plane temperature.


Here is the status:
Our current automated measurements are good to about 3x10^-6 per measurement
(i.e 1 line,  one node).
The limitation is photon statistics, not the algorithm.  No human effort 
can improve the formal error.  (We are working on tweaks to the code which
impact the absolute value)

In the simplest analysis it takes about 9 measurements to determine a change 
of 1x10^-6.  This is 5 orbits or 2 weeks.

One could imagine trying to correlate trends in the 24 nodes and 3 lines 
measured at any given time.  We might be able to attain 1x10-6 sensitivity 
per orbit with such an approach, unless the noise is correlated.
The ACIS team will doubtless point out many reasons why monitoring the 
mean CTI measurement is meaningless, but if they don't
We could start to monitor a "mean CTI" within 1 week.


complications:
Background - the ACIS team demonstrated that high background (measured by
 dropped exposures) gives the effect of improving CTI by up to 1x10-5 
since the noise is filling the traps.  We have speced a code to monitor this 
background and expect it to be in place in a month.  But, the relationship 
between background rate and observed CTI is not well understood.
At this point (and for the next 2-3 months) the best plan is to drop 
measurements where the CTI seems to improve.

Focal plane temperature - THIS IS THE CRITICAL ONE
Rino and I have been independently putting together data on the temperature 
spikes. Short result: Due to seasonal effects the spikes will get higher and 
wider for the next 3 months.  Peaking about July 1.  Soon,( <1 month ) 
the focal plane will be warm during the CTI measurements.  
We won't really be monitoring anything if we make CTI measurements when the FP 
is say -117.  We may have to
change to an operating procedure where HRC observations start and end each 
orbit and CTI measurements are made during science time.
The alternative is a good calibration of the effect of temperature on CTI.
  
 
Wish the news was better,
Scott

---------------------------------------------------------------------------
From MAxim Markevitch (3/23/00) -----  Frequency of ACIS background flares
---------------------------------------------------------------------------

I have just finished looking into some recent ACIS data to estimate the
flare frequency. It looks like they are becoming somewhat less frequent
compared to August (as some already suggested). Take a look at the figure on
the internal cal page:

http://icxc.harvard.edu/cal/acisbg/bg_newhist.html
username: CXCCal
password: ChaXAF93

The figure shows frequency for three periods -- August, September-November
and December-January. I think the note being sent to observers should now
say "the S3 background is more than twice the quiescent rate about 20% of
the time" (was 30%). The probability to have flares above a factor of 1.2 is
30-40% for S3 (this is when the data become useless for most galaxy cluster
analysis). This probability for S2 is now 5-15%.

Maxim