New load cells were installed in td12-1, td4-2 and td8-1 on 28nov17. We calibrated all 6 load cells on 01dec17.
To check the calibration:

• the ch was put at stow and the dome at 8 degrees za. This is close to balance for the azimuth arm.
• Each tiedown should see similar tensions (but they could vary slightly).
  
• the tiedowns were driven (in tracking mode) 21" -> 2" -> 21" at .05 inches per second.
• all 3 tiedowns move together, so the change in tension at each tiedown should be the same.

The plots show the td tensions after the calibration (.ps) (.pdf):

• We finished the calibration around 14:00
• Page  1:
• Top tiedown tensions for 01dec17 (after the calibration)
• Each color is a different loadcell tension.
• the ramps around 14-15 hours are when the tiedowns were cycled between 2 and 21 inches.
• Bottom: Tension difference between cables on a tiedown
• Each color is a different td.
• Black: td12-1 - td12-2
• red: td4-1 - td4-2
• green: td8-1 - td8-2
• The tension in each cable should be the same (since the yokes can pivot)
• The tensions are not the same, and they change with the tension in a tiedown.
  
• A linear fit was done to the difference vs the average tension in the two cables.
  

• td	fit offset Kips	fit slope kipDif/kipTension
  12	-.27	-.26
  4	-.22	-.3
  8	-.16	-.16

  
  
• So the difference in td4 cable tension changes .3 kips for each kip in a cable
• At 30 kips cable tension, the difference is 10 kips..
• Page 2: look at ramping tiedown position 2 to 21 inches.
• We moved the tiedowns 21" to 2" and back .. twice
• the tiedowns moved at .05" per second  in tracking mode.
• So the change in tension should have been the same for all of the tiedowns
• Top:over plot the tensions in the 6 ldcells during the scanning.
• Each color is a different load cell
• At 14.5 we were at 2 inches jack pull down. This was the minimum tension.
  
• the tensions were not the same because
• The dome/ch positioning did not guarantee that all the td  tensions started the same. There could have been an offset.
• 2nd-bottom frame. plot tensions from each tiedown separately
• black is load cell 1, red is load cell 2.
  
• For each load cell I removed the average tension at 14.5 hours  (when it was at a minimum)
• The plot variation should then show problems with the gain calibration.
• The slope of the tension is definitely different for ldcell1,2.
  
• Page 3: over plot ldcell1, ldcell2 for all tds
• The black curves are the loadcell 1 tensions for td 12,4,8
• the red curves are the loadcell 2 tensions for td 12,4,8
• all of the curves should lie on top of each other since the change in tension should be the same.
• The black curves (ldcell1) show lots variation.
• The red curves lay on top of one another.. The gain calibration must be very good.

Summary:

• 3 load cells were replaced on 28nov17
• All 6 load cells were recalibrated on 01dec17
• The difference in the tensions for the 2 cables at one tiedown got up to 10 Kips ( when 30 kips was inn each cable)
• When scanning the tiedowns 2 to 21 inches:
• The calibration for ldcell2 on on 3 tiedowns gave exactly the same results..
• The tensions in ldcell1 for each tiedown varied a lot.
• Something is going on causing the gain in loadcell 1 of each td to be incorrect.
• The problem is not noise, since the value does not jump around.
  
• load cell 1 is always plugged into the leftmost slot of each amplifier baseboard.
• loadcell 1 is read before loadcell 2 on the a/d converters.
• I thinks there is a single a/d and a mux that is used to read the different inputs.