|
Problem: SciMeasure Tip-Tilt Camera Malfunction
Manifestation:
-
SciMeasure Tip-Tilt camera malfunctioning (originally, at rates greater
than 125 Hz). Bias levels seem to readjust spontaneously.
Here is how one can talk to the TT camera via its serial connection
from the shaneao software:
tt=shaneao.tt
resp=tt.serial('@OAC?) # This will query the camera's bias offsets
print(resp) # This will print the response containing the bias offset information
resp=tt.serial('@OIC #0:$213') # This will set the bias offset for channel 0
These offsets are not saved if you change to another rate unless you do
an @SAV command, which should be done with care and caution.
Consult the following file:
user@shimmy > /local/u/user/observers/lgs/saocon_setups/ttcameraCommands.txt...
#TTCamera serial commands via shaneao
tt=shaneao.tt
#read bias offsets
resp=tt.serial('@OAC?')
print(resp)
#read temperatures
resp=tt.serial('@TMP?')
print(resp)
#1000 Hz
resp=tt.serial('@RCL 0')
#bad behavior offsets (Jan 19):
resp=tt.serial('@OIC #0:$3D7')
resp=tt.serial('@OIC #1:$3FF')
resp=tt.serial('@OIC #2:$373')
resp=tt.serial('@OIC #3:$398')
#better behavior offsets (Jan 25):
resp=tt.serial('@OIC #0:$1D6')
resp=tt.serial('@OIC #1:$1C2')
resp=tt.serial('@OIC #2:$1D6')
resp=tt.serial('@OIC #3:$1D6')
#save parameters:
resp=tt.serial('@SAV 0')
#500 Hz
resp=tt.serial('@RCL 1')
#bad behavior offsets (Jan 19):
resp=tt.serial('@OIC #0:$3D7')
resp=tt.serial('@OIC #1:$3FF')
resp=tt.serial('@OIC #2:$37D')
resp=tt.serial('@OIC #3:$384')
#better behavior offsets (orig):
resp=tt.serial('@OIC #0:$1D6')
resp=tt.serial('@OIC #1:$1C6')
resp=tt.serial('@OIC #2:$1D5')
resp=tt.serial('@OIC #3:$1D5')
#save parameters:
resp=tt.serial('@SAV 1')
#125 Hz
resp=tt.serial('@RCL 2')
#bad behavior offsets (Jan 19)
resp=tt.serial('@OIC #0:$2DO')
resp=tt.serial('@OIC #1:$2EC')
resp=tt.serial('@OIC #2:$2C6')
resp=tt.serial('@OIC #3:$2D0')
#bad behavior offsets (Jan 26)
resp=tt.serial('@OIC #0:$3DE')
resp=tt.serial('@OIC #1:$3DE')
resp=tt.serial('@OIC #2:$3D4')
resp=tt.serial('@OIC #3:$3D4')
#better behavior offsets (Jan 25)
resp=tt.serial('@OIC #0:$1EF')
resp=tt.serial('@OIC #1:$1E5')
resp=tt.serial('@OIC #2:$1EA')
resp=tt.serial('@OIC #3:$1EF')
#save parameters
resp=tt.serial('@SAV 2')
#40 Hz
resp=tt.serial(@RCL 3')
#bad behavior offsets (Jan 19):
resp=tt.serial('@OIC #0:$237')
resp=tt.serial('@OIC #1:$22B')
resp=tt.serial('@OIC #2:$234')
resp=tt.serial('@OIC #3:$233')
#better behavior offsets (Jan 26)
resp=tt.serial('@OIC #0:$1F4')
resp=tt.serial('@OIC #1:$1EA')
resp=tt.serial('@OIC #2:$1EF')
resp=tt.serial('@OIC #3:$1F4')
#save parameters
resp=tt.serial('@SAV 3')
- 1000 Hz
-
2016-11-22: Every pixel reads 0 DN, unless the
source is bright.
-
Originally (2016-10-27): Lower right
quadrant displays all values to be 16383 DN, but the
other 3 quadrants display high, but varying values
(13500--15900 DN).
- 500 Hz
-
2016-11-22: Every pixel reads between 500-700 DN.
-
Originally (2016-10-27): Upper right, lower
right, and lower left quadrants all have pixel values at
16383 DN, and the upper left display high values (circa
15900 DN), but appears to function reasonably normally.
- 250 Hz
-
2016-11-22: Every pixel reads between 500-700 DN.
-
Originally (2016-10-27): All quadrants show pixel
values varying from 7000--8000 DN at the top and
bottom of the detector and 16383 DN at the middle
of the detector.
- 125 Hz
-
2016-11-22: Every pixel reads between 450-550 DN.
-
Originally (2016-10-27): At 125 Hz all quadrants show
pixel values in the 5000--6000 DN range and varying from
image to image as would be normally expected.
- 40 Hz
-
2016-11-22: Every pixel reads between circa 550 DN.
-
Originally (2016-10-27): All quadrants show pixel
values circa 1500 DN and varying from image to image
as would be normally expected.
Workaround:
-
Set to 500 Hz for several cycles, allow to stabilize.
Try to re-acquire with desired rate.
If unsuccessful:
-
Use 125 Hz for field acquistion, centering of targets and monitoring.
125 Hz and 40 Hz are most often used for LGS operations with the
typical TT star magnitudes.
However, if we have brighter TT guide stars, it may be necessary to
use the TT filter ND filters to prevent saturating the detector.
Note the TT filter
wheel stops communicating after 256 or something
queries about its status. Hence, in order to move the
ND filter wheel it is likely necessary to cycle its
power, wait for communications to initialize with the
dispatcher, then attempt to apply the move.
Cause:
-
Inspection of cables and connections at the controller end (2017-11-10)
revealed at least one bad cable (green channel) and electronic noise
and odd behavior just by wiggling the cable. The quadrant in which these
effects manifest moves when the bad cable is connected to a different
channel.
There may be a problem with the red channel cable as well. Though that
may be in the camera as it ramps up with problems over time
(e.g. after about a minute that channel is unusable) and problem moves
with the camera/cable when channels at the controller end are swapped.
Problems are intermittent, but more consistent than not at this time.
Every once-in-a-while, we get lucky and all four channels work normally,
though the red channel (lower right quadrant on the TT Peeko
display) will soon not be usable at rates higher than 125 Hz.
We do not have a spare set of Little Joe BNC cables on Mount Hamilton.
Swapping-out the cables is impossible without disassembling the
tip-tilt camera stage, which poses its own risks for alignment.
-
Spontaneous improvement noted (22 November 2016). No known intervention.
-
It seems as if bias levels are not being set correctly for the
higher rate modes or there is some other problem.
(High) ambient temperature related?
-
Original manifestation (25 October 2016) followed significant software
intervention by A. Rudy on real.ucolick.org.
Associated Issues:
Change Log:
- 2018-03-29: Updated.
- 2018-01-27: Updated.
- 2017-12-04: Updated.
- 2017-03-14: Novel workaround discovered (operate at 500 Hz for a while).
- 2016-11-22: Spontaneous improvement noted.
- 2016-10-27: First version.
|