• tiedown tensions
• daily check of positions, commands, motor torques.
  

Schematics:

Little star Software:

• Third party documentation:
  
• Pk2200 users manual (newer little stars) (.pdf)
• Dynamic C 32 technical reference (from rabbit.com) (.pdf)
• Dynamic C Applications framework (.pdf)
  
• Dynamic C function reference (.pdf)
• changes to serial eeprom boards.
  
• Z180 user manual (.pdf)
• Our documenation:
  
• Interrupt usage
• How the little star timing works (1 sec tick,  control law,,...)
  

Hagen's tiedown  memos:

• BSD4 amplifier (.pdf)  : description of  Kollmorgan amplifier.
  
• kolmorgan (.pdf) : testing a kollmorgan 606 amplifier to replace the old tiedown motors.
  
• lightniing (.pdf)  Lightning protection references.
  
• loadcells (.pdf)  load cell calibration
  
• motor1 (.pdf)  Brushless DC motor servo motor operation.
  
• motor2 (.pdf)  Brushless DC motor servo motor operation.
• motor3 (.pdf)  Switching converter for motor control.
  
• motor4 (.pdf)   Switching converter for motor control.
• torque. (.pdf)   Motor torque required to lift tiedown block.
  

hardware

• Load cell amplifiers:
  
• scm5b. data sheet  pinouts. 
  
• board schematic 
• Fiber optic xmit/rcvr for 1 sec tick (.pdf)
  
• xmiter (min,typ,max) -20,-160,-14 dbm 62.5/125 um fiber
• page 14
  
• rcvr: (2412 page 19)
  
• Logic High : -40 dbm max
• Logic low   : -25.4 dbm min, -9.2 dbm max 100/140 um fiber (page)
• Optical power levels.1 sec tick  measured 25feb12 using 3 separate fibers driving ticks
• Prior to 25feb12 the 1 sec tick was sent down on a single fiber to box 12-7. It was then split  and sent to the individual tiedowns. The splitter was lossy (about 13 db). The  1 sec tick rcvr were receiving power levels near -25dbm. This was lower than the required power level (-24dbm) for the FO rcvr chip.
• On 25feb12 a fiber optic buffer was installed in the back of the pnt vme crate. It takes a single fiber input and generates 3 separate fiber outputs. These are then sent to tiedowns 12,4,8 on separate fibers (so the splitter is no longer needed). The table below shows the optical power levels measured on 25feb12 after the new 1 sec tick fibers were installed.
  

• Optical power levels for  1 sec ticks using separate fibers (25feb12).

  	td12	td 4	td8
  power @box b1 cable 5 in control room
  power @ box 12-7 before patch to each huffman box (n) is fiber number in 12-7	-17.8 (3)	-17.3 (5)	-18.5 (4)
  power @ 1 sec tick rcvr	-18.6	-18.1	-19.8

• 
  
• Motor personality cards
• td12 (.jpg)
• td04 (.jpg)
• td08 (.jpg)
  

Description of specific events

Replacement hardware for little star based system:

Limit switch positions/encoder reference:

    There is a pre and final hardware limit switch for the upper limit and the lower limit on each jack. The actions of the limts are:

• The final limit is a relay in the safety chain. When it opens, the safety relay opens causing the power to the motors to be cut off. To drive the jack out of this limit you must use the manual control built into each tiedown.
• When the prelimit relay is tripped (or the encoder value is within the pre limit) the program will automatically issue a stop commad. After it is stopped, the program will allow motion requests that move the jack out of the limit. It will limit the dac voltage to the sign that drives the jack in the correct direction. This assumes that the zero offset value for the dac is correct. When new dacs are installed, this value (parameter 19) needs to be updated. To check this, just watch the value of the Ki constant while the jack tracks the current position (eg montie 8). The integrated value needed to keep the jack fixed is the offset that needs to be added to the current value. The prelimit will stop the motion of the jacks if the dacs are controlling the motor amplifier. If the dac is blownout, the tiedown is in autorelease, or the manual control (hagens box) is enabled, then the prelimit will not stop the motion of the jacks (but the final limit will).
• There is also a software limit that is .2 inches inside the final limit. The VME software will never request a position outside of this software limit. When moving in rate mode (constant velocity) this limit has no affect.

1. move to a limit position (when the limit switch triggers) and then adjust the encoder to match the last encoder measurement at this position.
2. Measure the jack position with a tape measure and then interpolate what the encoder position should by using the encoder reading/tape measuremets and the two limits
3. Move the jack until the tape measure agrees with a miscLocation measurement and then set the encoder position accordingly.

    Whenever we survey from AO9, we use the current encoder values to determine how to move the tiedowns. So by definition, after an ao9 survey, the tiedown positions are now correct. This is not quiet true since we have been using the focus curves to determine the focus rather than the ao9 survey value.

The table below has the measured values of the position using encoder and tape measure. Encoder values 24 inches is when the jack is fully retracted (high tension). 0 encoder inches is when the jack if fully extended (low tension).The cols in the table below are:

• Limits encoder: Encoder readings at the limits (when the switch trips driving into the switch very slowly).
• Limits tape: The value measured (see illustration log section of td manual) with the tape measure when the td was at the final limit. The measured stop position for the final limit is a hardware stop (interlock opens). The measured stop position on the prelimit is the position where the software stopped it after seeing the digital bit change (and allowing for the slowdown).
• MiscLocations enc/tape . The encoder value, tape measurement at any other position.

• 17mar05. Changed the jack td12 for maintenance. Measured positions from bottom of jack to yoke.
• before:   after:
• measured the prelim encoder values with the vme. the final limits and the end of the jack travel were not checked. The final limit was computed from the prelim offsetkkj
• 28feb05. jack td4 broke. changed jack. Measured positions from bottom jack to yoke. difference before after was?? Looked at encoder values for prelims. Moved lower by about .08 inches. Did not measure the final limits or the end of jack encoder position.

Bar used to set encoder position:

    The software limits used to limit the requested positions are .2 inches inside the final limits. They are in the file /home/phil/h/tieProg.h
 

software request position limits.

start date	td12	td8	td4
08feb00	(22.29,1.43)	(22.87,1.93)	(22.20,1.19)
xxx	(21.79,1.2)	(22.14,1.52)	(23.21,1.95)

Tiedown power supply startup sequence:   (top)

1. the safety is closed.
2. the master contactor is closed.
3. there are no power supply faults.
4. you have requested the tiedowns to move.

1. POWER1 to energize the master contactor relay to startup
2. POWER2 to drive the master contactor relay during normal operation inline with the power supply fault contactor and the safety.

1. reset will close the safety if there are no safety problems (eg.blk liftoff, emg stop, etc.).
2. You must command the tiedowns to move (it can be to the current position) or use the pcu to turn the drives on.
3. The little star will driver 24 volts through POWER1  dio2 pin 2,3 for about 300 milliseconds. This provides power to the master contacter relay PS_MC_R if the saftey is closed.
4. PS_MC_R closes the master contactor and the 3 phase power is supplied to the power supply. The power supply will then close the power supply fault contactor (PSR_4_FAULT* goes high). 24 volts is wired through the power supply fault contactor, throught the POWER2 contactor in the little star, the safety, and then the PS_MC_R.
5. The little star loops up to 20 times checking to see if PSR4_FAULT* goes high (about 100ms/loop i think). This is pin DIO1 J3 pin4 --> lsPin03.  If after 20 times through the loop the contactor hasn't closed, then the little star will drive POWER1 to 0, open POWER2 and log fault 12 (PSR_FAULT psr4F).

Load cell resistances measured at input to amp.  (top)

In the table below:
        amp 1 is the top    amplifier
        amp 2 is the bottom amplifier
        cable 1 normally goes to the top amp
        cable 2 normally goes to the bottom amp
see the drawing in the td manual for where the load cell is located on the yoke.

To do the measurement:

1. turn off amplifier
2. remove the 4 wires that go to the amplifier.
3. remove tension on the load cell by raising the  jack. Point the  dome point at the tiedown.
4. measure the resistance at the 4 wires that connect to the amp.
5. The sparcarrestor,and db25 filter should be in place.

The values with ** had bad filters in the db25 connector.
++ removed the filter because it was bad (and had no spares)

• 1. 09nov06. ld cell 1 td12 failed. replaced with spare loadcell
• 2. 13jan16.. load cell had been damaged by hf back in nov15.
  
• could not get the balance to go to 0. tried switching cables at load cell, problem stayed with the load cell.
• the load cell cable had been replaced back in nov15.
  

Load cell Info:locations,history,calib    (top)

• At some point td 12 ldc1,ldc2 were switched.
  
• It used to be ldc1 left, ldc2 right (on this picture)
• 20aug10 it is ldc2 left ldc1 right.. we need to verify the correct number of the load cells.
• 09jun11
• checked that picture td12 is correct. ldc2 closest to cabinet (removed cable at load cell and watched which one went away at little star.
• Note: the loadcells at td12 are marked: one closest to box , and 2 farther away. this is reversed from what is really there. Some day we may want to  switch the cables at the input to the amps.

• (1)09jun11 went to td 8,12, checked loadcells:
• td12:
  
• load cell 2 is closest to box, load cell 1 is farthest from box.
• checked this by removing cable at loadcell and watching which ldcell reading changed at little star.
• The loadcell closest to box is marked loadcell 1(on top). This marking is incorrect.
• they have labeled  td12-1 (90408) td12-2 (90409). I have td12-1 replacing 90408 back on 14nov06 with 147697. I'm going with the 147697.
• (2)xxjan16  td8-2 load cell was damaged by the hf transmitter in nov15. 
• The load cell cable was replaced back in nov15.
  

• 147697
• 207629
• 253850
• 263010
• 263011
• 263012
  

Tiedown reference position:

 Pictures of tiedown system:

• td 8 jack
  
• td 12:
• td limit switches
• Bar to set encoder value (looking down from top)
• Bar to set encoder value (looking at bottom).
• transmission, jack coupling 
• transmission, jack separated
  
• td 12 boom,mast from td4 corner of triangle
• td 12 boom, backstay cables from center of triangle
• td 12? top of mast
• td backstay connection point at  u2-u2l3
• Grounding
• Jorge covering grounding trench at td12
• Jorge and eddie (1) at td12
• Jorge and eddie (2) at td12
  
• luis At td 12
  

  ---

debugging:

• If td does not stop when you send stop command, or off.. the it is probably being controlled by the watchdog timer.
• If the velocity feedback sits at a constant negative value (about -300) then this is probably the watchdog timer circuitry driving it.. the amp speed monitor should be around -400 counts.
  

<- page up
home_~phil