2014_07_18_SetD < Subaru/PFS

---++ Switching to Set D
   * Same as Set 1, with the following exceptions:
      * pId3 is SN 145 instead of 152 (de-bonded shaft coupler)
      * pId8 is SN 116 instead of 103 (flex interference)

   * Ran frequency sweeps and saved files to Dropbox (per the procedure added to Twiki)
   * New On-Time tuning mSim script
      * Moves 7000 steps in 20 tries (350 steps each) instead of using one big move of 7000 steps
      * Still takes one exposure so it will appear as a streak with a bunch of bright spots

---++++PID6 Investigation

During ontime streaks noticed that wasn't seeing any change on PID6. Confirmed theta stage works, and found that phi was not responding. Tried commanding with ontimes up to .3ms and still nothing but could hear the motor running when phi was commanded.
Started pathway and also could not jog phi (motor 12) in fwd or reverse but heard it running.
Looked at axis attributes and noted that current frequency is 92.937kHz
Called up the frequency sweeper and ran fwd/rev sweeps. They look bad but the step interval is wrong somehow at 0.29952. Changed it to 2.49984 per NST documentation for C0104 and centered sweep at 88.339 per that document as well. The sweep looks good now with frequency peak at 88.65. 

By hand moved phi to middle of its ROM and discovered that both FWD and REV commands move it in reverse.
Swept each phase independently and discovered that one of them is flatlined. 
Reseated the both ends of the flex extension cable which connects to the motor flex and driver.
Redid the individual phase sweeps and still found that phase 1 is flatlined.

     <img src="%ATTACHURLPATH%/pid6_phase1dead.PNG" alt="pid6_phase1dead.PNG" width='643' height='418' />

---++++PID5 Investigation
PID5 was getting stuck a lot in the ontime tests on theta fwd. Examination by hand and eye leads to suspicion that SN149 is hitting SN101. SN101 theta stage was rotated to what seemed like a better orientation but there may still be interference.

---++++ONTIME Calibration
Calibrated ontimes.  pId6 (S/N 104) did not move as expected (see above).  After on-times were selected pId6 was switched over to S/N 108 (pId5 from set 3).  Re-ran chooseOnTimesManually with the previously measured values for S/N 108.  Saved resulting .xml files as *_2.xml to make sure nothing else changed.  Re-ran frequency sweeps for S/N 108 with its new driver.

Verified with diff in D:\PfsTests\07_18_14_13_22_30_S2rv_onTimeTuning\Images:
<verbatim>
$ diff PhiRevOnTimes_071814.xml PhiRevOnTimes_071814_2.xml
384c384
<          <Link1_fwd_Duration>0.05</Link1_fwd_Duration>
---
>          <Link1_fwd_Duration>0.08</Link1_fwd_Duration>
386c386
<          <Link1_rev_Duration>0.05</Link1_rev_Duration>
---
>          <Link1_rev_Duration>0.08</Link1_rev_Duration>
388c388
<          <Link2_fwd_Duration>0.05</Link2_fwd_Duration>
---
>          <Link2_fwd_Duration>0.1</Link2_fwd_Duration>
390c390
<          <Link2_rev_Duration>0.05</Link2_rev_Duration>
---
>          <Link2_rev_Duration>0.1</Link2_rev_Duration>
392c392
<          <Link1_fwd_Duration_Slow>0.05</Link1_fwd_Duration_Slow>
---
>          <Link1_fwd_Duration_Slow>0.08</Link1_fwd_Duration_Slow>
394c394
<          <Link1_rev_Duration_Slow>0.05</Link1_rev_Duration_Slow>
---
>          <Link1_rev_Duration_Slow>0.08</Link1_rev_Duration_Slow>
396c396
<          <Link2_fwd_Duration_Slow>0.05</Link2_fwd_Duration_Slow>
---
>          <Link2_fwd_Duration_Slow>0.1</Link2_fwd_Duration_Slow>
398c398
<          <Link2_rev_Duration_Slow>0.05</Link2_rev_Duration_Slow>
---
>          <Link2_rev_Duration_Slow>0.1</Link2_rev_Duration_Slow>
</verbatim>

---++++Center and Link Length Metrology 
Created new centers script with correct ontimes called 2014_07_18_Centers01. Results are in 07_18_14_14_35_29_centers

Created new motor map script with correct ontimes 2014_07_18_msimMotMap

Generated centers/link lengths with the following notes:
   * By running chooseOnTimesManually rough centers are not calculated, which resulted in an error in the centers analysis
      * Corrected by picking rough centers from image using JImage and hard coding them into the xml
      * A new technique of determining centers will be needed.  One possibility is to put all positioners into their home position and go from there.
   * pId6 had a large sigma error from the S2 calculation.  There were a cluster of 4 centers and 6 centers.  Initial thought is that it is bumping into a neighbor, which pushes it around a bit. (need to think about this more...might not make sense).
     
<img src="%ATTACHURLPATH%/pId6linkerror.PNG" alt="pId6linkerror.PNG" width='400'/>

   * Continuing anyway with this error.  All other results look good.

Ran motor mapping script at ~5pm before leaving (Fisher)
   


-- Main.ChazMorantz - 18 Jul 2014
Attachments
Topic attachments
I	Attachment	History	Action	Size	Date	Who	Comment
PNG	pId6linkerror.PNG	r1	manage	5.4 K	2014-07-19 - 00:04	UnknownUser
PNG	pid6_phase1dead.PNG	r1	manage	47.7 K	2014-07-18 - 21:29	UnknownUser
This topic: Subaru/PFS > WebHome > POSitionerEM > POSEM_AL > 2014_07_18_SetD
Topic revision: r2 - 2014-07-19 - ChazMorantz