2005-06-21 Dec. APE Engineer Data
Introduction
In June 2006, we installed the DEC APE electronic. This directory
contains data collect from the new ape while running the forth
TCS system so we can determine how well the ape is working and
get some calibration values for the new APE to do the
position->Declination calculation.
Note: 1st set of engineering data was collect on 06/15, but
their was a problem with the polarity of the coarse and fine
sensor -> the sin/cos cables on the transductor was switch
to fix the problem.
This page contains the data collected, and our results.
1. TCS1 and MCC APE.
TCS1 zenith position is: 00:00:00 19:49:35
Tcs3 initialize to: 00:00:00 19:49:34 pmac cnt = 1425305
Look at tcs1 and MCC apes.
tcs1 MCC apes tcs3 Pmac counts
---------- ---------- ----------------
40:00:00.00 39:59:51 2874777
+20:00:00.00 20:00:04 1437797
+10:00:00.00 10:00:11 719303
+00:00:00.00 +00:00:29 818
-20:00:00.00 -19:58:54 -1436076
-30:00:00.00 -29:58:29 -2154434
-40:00:00.00 -39:57:55 -2872648
2. Switch to new APE.
1.1 For the dec axis, the daycrew must slew over for access to the mirror cell
Make sure that the tcs1 system is kept running.
1.2 Switch over to tcs3 Dec apes electronics.
1.3 record tcs3 APE position:
coarse_data:
fine_data:
1.4 Make sure position is loaded in forth TCS before removing brakes.
Here is the position initialization command:
hhmmss.ss ddmmss.s !START
3. Collect coarse Data.
tcs1 lost position, so we move it to 1425305 pmac counts &
use the '000000.0 194934.0 !START' command to initialize the forth system.
tcs1 coarse: fine: position PMAC counts
19:49:35 11805 24198 2121350 1425303
Note: tcs position provided by the forth system using incremental encoders.
We move to various Declinations and recorded the coarse & fine data.
tcs1 coarse: fine: position PMAC counts
40:00:00 deg: 8069 18607 1460399 2874775
20:00:00 deg: 11773 18471 2115623 1437796
10:00:00 deg: 13616 18402 2443234 719303
00:00:00 deg: 15445 18261 2770773 816
-20:00:00 deg: 19072 17904 3425776 -1436078
-30:00:00 deg: 20865 17659 3753211 -2154435
-40:00:00 deg: 22645 17317 4080549 -2872650
4. Collect Fine Data
Take 4 deg. of data for 960 seconds at 15 as/s.
(started at 17:49:34 to > 21:49:34 at 15as/s).
Data is collected using the readdata command in the apes/tester.c program.
Here is the command to start it, for 10Hz and degrees format.
> readdata 0.1 1
Data files is: 050621-rawread_10hz.dat
Format of the data is:
timestamp esec r_raw t_raw r(dec) t(dec) pos(d) pos(rad) pos(deg or Tm) pos_rad_chg.
11:06:09.12 0.000 0x0000 0x0001 0 1 1 0.000001 00:00:0.01 Tm 0.000
timestamp - timestamp of sample.
esec - elapse second (of sample)
r_raw - resolver data (raw)
t_raw - transductor data (raw)
r(dec) - resolver data (decimal)
t(dec) - transductor data (decimal)
pos(d) - correlated position
pos(rad) - position as radians
pos(deg or Tm) - position as Deg or Tm string
pos_rad_chg - changed of position from last sample (radians).
5. Fine Data analysis
In step 4 we collect find data at 10 Hz with move at 15 as/s. This should
give us a constant velocity over a distance of 4 degees.
Graph of Coarse Data, Fine Data, and rad_chg over time:
- Coarse Data (resolver)
- Fine Data (transductor)
- Position Change
- Absolute Position (rad) .
Graphs produced using this gnuplot script: 5.gp
6. Coarse Data analysis
I combined data collect in section 1 & 3 in this table: 6.dat
Also play about with the data in an excel spreadsheet, The tcs1 vs ape data
looks good (except for the point at -40 deg). The encode ratio vs the APE
is very steady too.
Function is dec_rad = 1.47865405552700224234 + -5.3384416906172e-07 * ape_cnt
6_idl - directory contains the program, input data, and output.
6_fit.dat - This file shows how well the equation works.
Here it is in graph form:
MCC_APE vs. fit_error, Over -40 to 40 degrees.
MCC_APE vs. fit_error, Over 10 to 30 degrees.
Note it is about 32 arcsecond near zenith. (0.35 radians).
7. Comments