This is a page to track lenslet mask performance issues and the steps we take to troubleshoot. So far the lenslet mask performance has become more unreliable. Due to the lack of encoders, there is no feedback to track which lenslet mask position is really in use. The unreliable behavior has made it more difficult and time consuming to make rectification matrices.
Default lenslet mechanism setting:
speed = 300 steps/s
acceleration = 200 steps/s/s
homing speed = 400 steps/s
A few things to try to troubleshoot:
- Try reduce speed to 200 steps/s
– scans looked bad, spotty. - Try reduce acceleration
Try changing om4s homing speed- om4s homing speed is 400 steps/s, try increasing speed up to 400 steps/s
– scans looked bad, spotty. - Try moving between narrowband and broadband and see if it is reliable
– stage arrived to narrow band correctly but appeared lost in OSGUI
– stage arrived to broad band correctly - Move between Broad and Narrow in steps. Go to an intermediate location (stage will be lost), drive in steps the known amount and see if the switch is triggered (stage will no longer be lost)
– stage remained lost the whole time. arrive to some locations 1 step short or 1 step too far, arrived to some locations correctly.
– this is seen in both directions - Check whether scan_number locations are defined correctly
- Step the mask across all the lenslets in even steps.
– mask stage was suck at Scan 2, no light. - Reverse scan direction
– Saw no light at all. ofls is running.
Perhaps the flat lamp has died, saw no light at all in scans taken after 6/6. Flux in scans taken on 6/6 was orders of magnitude fainter than what it should be.
If the flat lamp was failing, it could explain spotty illumination in previous scans. The lamp brightness would have become unstable.
There were RPC errors, lenslet stepping errors, and lenslet stage mortar update errors.
Problematic recmat scan symptoms:
After about 100 scans (in narrowbands) lenslet fluxes drop significantly and become uneven
Uneven lenslet fluxes in every scan
