Real-time Controller

Communications and Commands

The socket sends and recieves data streams (as opposed to data-grams). Although this has the benefit of assuring successful transfer of all packets, and in the order sent, it also means that there is no guarantee that data will be sent in the 'chuncks' that the program sends them in. In other words, the messages can be split into multiple packets and messages can be concatenated into single packets. For this reason, the packets have to be parsed. Start and end sequences will be includedin the messages, as necessary, to facilitate this.

All data, in both directions, are in ASCII format.

Host to Control System Communications

COMMAND	DESCRIPTION	NOTES
MISCELLANEOUS
gsmode x	selects natural guide star (ngs) mode or laser guide star (lgs). (x: integer; 0 selects ngs mode, 1 selectes lgs mode)
abort	aborts long commands (such as recent or cm)
quit	quits the control system program
restart	restarts the control system from scratch (i.e., reboots)
shutdown	shuts down Linus (in preparation for powering down)
parms	reads all parameter files into memory	same as cparms
CAMERA INTERFACE
rate x	sets the frame rate of the camera (x: integer; 55, 100, 200, 250, 350, 500, 750, or 1000 frames per second)	used by tcl
cgain x	set the camera gain (x: integer; 1 or 2)
crtgain x	sets the CRT gain (adjusts the video output only, has no effect on the images used by the control system) (x: integer; 1, 2, 3, or 4)
srate	reads the frame rate from the camera and reports it to the host.
scgain	reads the gain from the camera and reports it to the host
scrtgain	reads the CRT gain from the camera and reports it to the host
centbin	changes to 4x4 binned by 2 quad cell centroider algorithm (to make a quad cell with larger pixels)
centcm	changes to center-of-mass centroider algorithm
centquad	changes to quad cell centroider algorithm
thresh x	sets threshold used by the centroider on the images (new data) from the WFS (the threshold is subtracted from the raw data and them any resulting values less than zero are set to zero) (x: integer; 0 to 4095 ADC counts)
cflat	determines new centroid flat field values (these are actually backround, or dark data, values), makes them current, and stores them to disk; these values are subtracted from the raw image data in the centroiding algorith; the file that the data are stored in is yyddmm/cflat_xx
refcent	determines new reference centroids, makes them current, and stores them to disk; these values are subtracted from computed centroids in the centroiding algorithm; the file that the data are stored in is yymmdd/refcent?_xx, where ? is b, c, or q depending on the centroiding algorithm
centoffs x1 ... x80	makes 80 centroid offsets received from the host current (these values are computed by the host as a function of image-sharpening slider positions); these values are subtracted from computed centroids in the centroiding algorithm (x1 ... x80; floats; -1.0 to 1.0 pixel units; 40 X centroids followed by 40 Y centroids)
saveoffs	stores 80 current centroid offsets to disk; the file that the data is stored in is yymmdd/centoffs_xx
cparms	reads all parameter files into memory	same as parms
sparms	makes current reference centroids all zeroes (but does not store to disk>
RECONSTRUCTION
cm	creates a system matrix and saves it to disk (in preparation for computing a control matrix, cm, offline)
fillcm filename	loads the control matrix from the file specified by 'filename', which must be in the parms directory, from disk and makes that matrix current (filename: null-terminated string; can be any file name, including additional path information, e.g., cmNew, yymmdd/cmTest, etc.)
DM LOOP CONTROL
close	closes the AO loop	used by tcl
open	opens the AO loop, leaving the DM in its latest closed-loop state	used by tcl
estop	emergency stop - opens both the AO and TT loops
gain x	sets the AO loop gain to x; can be issued with loop open or closed (x: float; 0.0 to 1.0)	used by tcl
int x	sets the AO loop integrator value to x; can be issued with loop open or closed (x: float; 0.0 to 1.0)	used by tcl
cwt1 x	sets the AO loop compensator weight 1 value to x; can be issued with loop open or closed (x: float; -1.0 to 0.0)	lead/lag compensator
cwt2 x	sets the AO loop compensator weight 2 value to x; can be issued with loop open or closed (x: float; -1.0 to 0.0)	lead/lag compensator
mflat	averages actuator voltages and stores the result to disk as the 'flat' voltages; the file that the flat voltages are stored in is yyddmm/flat_xx
flatten	'flattens' the DM by loading the last flat voltages recorded and applying them to the DM
msharp	averages actuator voltages and stores the result to disk as the 'image-sharpened' voltages; the file that the flat voltages are stored in is yyddmm/sharp_xx
sharpen	'sharpens' the DM by loading the last image-sharpened voltages recorded and applying them to the DM
reg x	repeatedly cycles the specified actuator to test the registration of the DM with the WFS, the health of hte actuator, etc. (this can be executed multiple times to move multiple actuators) (x: integer; see Registration Codes for details)
regok	stops the cycling of all actuators started with reg
TIP/TILT CONTROLLER
ttclose	closes the TT loop	used by tcl
ttopen	opens the TT loop	used by tcl
ttgain x	sets the TT gain to x (x: float; 0.0 to 1.0)	used by tcl
ttxmov x	sets the TT X mirror voltage (x: float; -5.0 to +5.0 Volts)	this is a temporary setting
ttymov x	sets the TT Y mirror voltage (x: float; -5.0 to +5.0 Volts)	this is a temporary setting
lxmov x	sets the laser FSM X mirror voltage (x: float; -5.0 to +5.0 Volts)	this is a temporary setting
lymov x	sets the laser FSM Y mirror voltage (x: float; -5.0 to +5.0 Volts)	this is a temporary setting
apdclose	closes the APD TT loop
apdopen	opens the APD TT loop
apdrate x	sets the APD loop rate (x: integer; 50, 100, 200, 300, 400, 500 Hz)
apdgain x	sets the APD TT gain to x (x: float; 0.0 to 1.0)
apdback	determines background values for the APDs and stores them in the parms file
apdzback	zeroes the background values for the APDs and stores them in the parms file
DIAGNOSTIC DATA HANDLER
telem x	starts sending telemtery data to the host, depending on which bits in the 'x' are set: raw image = 0x01 centroids = 0x02 intensity = 0x04 DM voltages = 0x08 tt data = 0x10
trate x	rate at which telemetry data should be sent to the host (x: integer; 1 to 50 times per second)
diag x	gathers diagnostic data into automatically named files, depending on which bits in 'x' are set: raw image = 0x01 (100 frames) centroids = 0x02 (2048 frames) intensity = 0x04 (2048 frames) DM voltages = 0x08 (1024 frames) tt data = 0x10 (1024 frames)
images	same as diag 0x01
data	same as diag 0x0E	used by tcl
apddata	same as diag 0x10
pon	turns psf data accumulation on
poff x	turns psf data accumulation off, appending the integer value, x, to the filename (e.g., a value of 3 for x will give the filename "psfdata_0003.fits" stored in "/home/winderbean/AO/Data/psfdata/yymmdd")

Command notes:

1. The 'parms' directory, where all parameter files are stored, is /home/winkerbean/AO/Data/Parms/.
2. yymmdd, in filenames, refers to year, month, and day.
3. xx in filenames refers to a sequential number from 00 to 99.
4. The 'automatically named files' referred to in the diag command are stored in the /home/winkerbean/AO/Data/RunData/ directory and are:
   • yymmdd/rawImage_xx
   • yymmdd/cent_xx
   • yymmdd/inten_xx
   • yymmdd/mirror_xx
   • yymmdd/tt_xx
5. Ann integer parameter should have a minus sign if it is negative and no sign if it is positive, followed by 1 or more decimal digits.
6. A float parameter sould have a minus sign if it is negative and no sign if it is positive, should have at least 1 decimal digit, and may or may not have a decimal point.

Control System to Host Communications

bytes 0-2	message start sequence	"~S~"
byte 3	message identifier	'0' = text message '1' = raw image data '2' = centroid data '3' = intensity data '4' = DM voltage data '5' = tip/tilt data
bytes 4-x	message	see Notes
bytes x+1 - x+4	message end sequence	"~E~\n"

Message Descriptions

• Error: DMA buffer has been overwritten, counter = 2
• Error reading cflat data, previous values being used
• Warning: recon processing getting behind
• Notification: wrote file: 030706/centoffs_02
• Notification: mflat command complete

Error messages indicate that an event has occurred that puts the performance of the AO system into questions. These are usually not fatal in the sense that the system has not 'crashed'. But the user should be aware of the reported error and make a judgment, based on the performance of the system, as to whether the system can continue to be used without corrective action. Repeated errors indicate a more serious problem.

Warning messages indicate that an event has occurred that the user should be aware of. The cause of isolated warning messages should not adversely affect the performance of the system. If warning messages occur repeatedly, they should be treated in the same way that error messages are treated.

Centroid Data

DM Voltage Data

Tip/Tilt Data

1. X tip/tilt error, in pixel units
2. Y tip/tilt error, in pixel units
3. X tip/tilt mirror setting, in Volts
4. Y tip/tilt mirror setting, in Volts
5. X telco setting, in Volts
6. Y telco setting, in Volts
7. X laser steering setting, in Volts
8. Y laser steering setting, in Volts
9. not currently used
10. telco enable, in Volts (0V = enables, -5V = disabled)
11. APD A counts
12. APD B counts
13. APD C counts
14. APD D counts
15. X APD tip/tilt error, in APD units
16. Y APD tip/tilt error, in APD units
17. X tip/tilt mirror LVDT reading, in TBD
18. Y tip/tilt mirror LVDT reading, in TBD
19. APD over-limit flat (0.0 if all APDs are reading less than or equal to 1,000,000 counts per second, 1.0 if any APD is reading more than 1,000,000 counts per second)
20. total number of APD counts, for the four APDs, per second

Registration Codes

	Field	File Name or Default Value	Example: 2003 Dec 16
1	current cflat filename	yymmdd/cflat_xx	031215/cflat_06
2	current refcentb filename	yymmdd/refcentb_xx	031210/refcentb_00
3	current refcentc filename	yymmdd/refcentc_xx	031210/refcentc_00
4	current refcentq filename	yymmdd/refcentq_xx	031210/refcentq_00
5	current centoffs filename	yymmdd/centoffs_xx	t5
6	current cm filename	yymmdd/cm_xx	CMSep03_00wls3.sdt
7	current flat filename	yymmdd/flat_xx	031120/flat_01
8	current sharp filename	yymmdd/sharp_xx	031215/sharp_00
9	# of frames to averages when determining cflat and APD background values	100	200
10	# of frames to averages when determining reference centroids	100	2000
11	# of DM voltages to average when determining a system matrix	100	2000
12	#of DM voltages to average when determining flat and sharp mirror values	100	200
13	0 sets to ngs mode, 1 sets to lgs mode	0	0
14	camera frame rate, in frames per second, 55, 100, 200, 250, 350, 500, 750, or 1000	100	500
15	camera gain, 1 or 2	1	2
16	camera CRT gain, 0, 1, 2, 3, or 4	0	4
17	indicates which centroid algorithm is currently being used, 0 = 4x4 binned, 1 = center-of-mass, 2 = quad-cell	2	2
18	centroider raw-data threshold, in ADC counts	0	0
19	AO loop gain (proportional gain), 0.0 o 1.0	0.35	1
20	AO loop integrator gain, 0.0 to 1.0	0.95	0.998
21	AO loop compensator weight 1, -1.0 to 0.0	1.0	1
22	AO loop compensator weight 2, -1.0 to 0.0	1.0	1
23	tip/tilt loop gain, 0.0 to 1.0	1.0	1
24	APD loop rate (50, 100, 200, 300, 400, 500 Hz)	100	500
25	APD loop gain, 0.0 to 1.0	0.003	0.01
26	APD A background values (in APD counts)	0	0
27	APD B background values (in APD counts)	0	0
28	APD C background values (in APD counts)	0	0
29	APD D background values (in APD counts)	0	0
30	next sequential file number (for data files)	0	69
31	serial number for DAC board farthest to the left when looking at the front of the lgs7 chassis, that DM channels 0 to 31 are plugged into	-1	18362
32	serial number for DAC board farthest to the left when looking at the front of the lgs7 chassis, that DM channels 32 to 63 are plugged into	-1	18366
33	serial number for DAC board farthest to the left when looking at the front of the lgs7 chassis, that DM channels 64 to 95 are plugged into	-1	18364
34	serial number for DAC board farthest to the left when looking at the front of the lgs7 chassis, that DM channels 96 to 127 are plugged into	-1	18361
35	serial number for DAC board farthest to the left when looking at the front of the lgs7 chassis, that TT channels are plugged into	-1	18333

Important: If the format of the parameter file changes, the Tcl program parmsRead.tcl will have to also be updated to reflect this change (otherwise the LickAO and CentDiag software will be unable to properly read the current parameters).

If runnig the program as the lgs user, do the following:

1. Establish a remote session, logging on as lgs, and using the lgs password.
2. At the prompt, enter: sudo loffload
3. Data will star being displayed, as described below.

If the AO code is quit, loffload will be automatically terminated and will have to be restarted after hte AO code is restarted.

The data displayed by loffload will be updated once per second and will look similar to the following:

Avg TT Volts X: 0.113277 Avg TT Volts Y: -0.370412

The data represent the voltage being applied to the X and Y axes of the laser uplink mirror, averaged over 1 second (using a rolling average).

Note that, although loffload is intended to be used in lgs mode, it can also be used in ngs mode, in which case, the data will represent the voltage being applied to the X and Y axes of the tip/tilt mirror (on the optical table), averaged over 1 second (using a rolling average).

Note that CentDiag now uses this information and will create an audio alert ,via the soundserver software for IRCAL, when the laser fast steering mirror voltage is above a threshold value (currently 1.5 Volts) for over a certain time period (currently 60 seconds) when the uplink TT loop is closed. This signals to the laser operator that a manual adjustment of the laser pointing is needed.