Functional Checkout of the Secondary Mirror

This is a draft plan for the SMU FCO of July 2003 following its complete overhaul. It assumes passage of an electro-mechanical checkout on 22 July.

Participation of many JCMT staff is anticipated and appreciated. This document already contains contributions from JW and CW.

The document was written in advance of the completion of the repairs and has been updated in blue with the reality that followed.

Preliminaries

1. Settings prior to the overhaul are as described in smu.ifl (in mt_smudir), and in the SMU focus document in this tree.
2. It may be anticipated that nominal settings (zero-points) and limits will all change, but that relative positions may remain unchanged. For example :
   • the default Z focus position for SCUBA will change, but the relative offsets between SCUBA and other receivers will remain the same,
   • the values assigned to the soft limits will change, but (given the same screw pitch) the difference between the upper and lower limits should remain the same.
   Nonetheless, and to repeat, the FCO plan must anticipate 'all change'.
3. It is anticipated that Chopper functionality has not been changed - i.e. that the subreflector and Chopping hardware have not been separated - although a cautious FCO would include one perfunctory check.

1. Determine hard limits on X,Y and Z tables (TCC, NTM), and set soft limits.
   • A May 2000 document ( /JACdocs/JCMT/a/003/17) describes the limits, ranges and hard stops of the X, Y and Z axes as follows :

     Axis	limits (mm)	range (mm) encoder units	hard stops (mm)	range (mm) encoder units	current soft limits range
     X	+23.5	47.0 12866	+24.0	48.0 13140	12900
     Y	+23.5	47.0 12080	+24.0	48.0 12337	12975
     Z	+22.8	45.6 12486	+23.3	46.6 12760	12600

   • Set X soft limits such that range again = 12900 units.
   • Set Y soft limits such that range again = 12975 units.
   • Set Z soft limits such that range again = 12600 units.

     Set zero-points using the similarity arguments above :

   • Set new X-zero-point = XLSL + 7262
   • Set new Y-zero-point = YLSL + 5585
   • Set new Z-zero-point = ZLSL + 6570
     
     (XLSL = X axis lower soft limit, etc.).

   The reality :

   • On Tuesday, 22July, the positions of the XYZ tables at each end of their travels, as limited by their respective limit switches rather than the hard limits anticipated above, were recorded.

     In two of the three cases the encoder readings roll through zero. This caused momentary concern, but after a little consideration it became clear that in the general case where the positioning of the encoder w.r.t. the tables is unconstrained (eg if a replacement is ever necessary and the location of the zero is essentially unknown) then the micro software would have to account for such an occurrence.

     This is stated here verbosely so that it might pass into corporate memory and cause no similar concern next time.

     Switch limit positions were recorded as follows :

       
                                                Notes
        X : 1512 to                 14620     encoder units
        Y : 8377 to  5255 + 2**14 = 21739     encoders are 14-bit
        Z : 7906 to  4363 + 2**14 = 20747
     

     The errors in the switch positions above are of order 100 encoder units, as determined from a small number of repeats.

     It also became clear that the lower limit was being subtracted by the SMU micro code before being passed to higher-level software and before being displayed on, eg, the s3 screen at the OPERATOR's station. Since the limit switches were not repositioned these lower limits should not have changed (within the errors).

     The intervals between these limits may also be compared with the old configuration. The table below shows this new interval and the old upper & lower soft(ware) limits, LSL & USL :

     1	2	3	4	5	6
     axis	limit switch interval	LSL	USL	4-3	2-5
     X	13108	400	13300	12900	~200
     Y	13362	275	13250	12975	~375
     Z	12841	300	12900	12600	~250

     The soft limits are a couple of hundred encoder units 'closer in' than the switch limits - as expected. Crude (plastic ruler) measures of the switch intervals for the X & Y tables yielded 51 and 50mm resp., giving X_SCALE = 257 enc.units/mm and Y_SCALE = 267. Accuracies of +10 units/mm should be expected. These should be compared with current values from SMU.ifl of 273.756 and 257.02, resp. These values do not need changing : they're just not as accurate as the quoted 2 or 3 decimal places suggest, or were originally determined in a more precise manner.

     The soft limits are clearly set to be roughly 300 from each end. Since the re-positioning accuracy of all parts (eg limit switched w.r.t. the tables) is no worse than 1mm then changing any of the current values for the soft limits and the zero-points seems unjustified.

     On 24 July, the SMU.ifl parameters were restored to their pre-failure values.

     With these (old) default operational settings sources should be acquired within the hardware restoration accuracies of a mm or two; this translates to about 15-20" in the focal plane : not enough to move off the SCUBA FOV.

     Now - the weather . . . .