I made a followup measurement for an old coil-balance test done on the MCo suspension (3295).
The situation seems the same as before --- the measurement doesn't make sense and something is wrong with the MCo suspension.
(measurement)
When Telada san and Uchiyama san were not using the main beam in this afternoon, I took this opportunity and repeated the measurement as described in 3295.
Here is the result:
P(urad) Y(urad)
F1: +1.3 +28
F2: -5.5 +30
F3: -2.2 -19.6
F4: +5.1 -20.8
The qualitative behavior seems unchanged from the past. That is: while the yaw responses look OK-ish, the pitch responses vary a lot by a factor of about 4 between the coils.
(Some imagination)
The equation of motion (in the frequency domain) for an angular degree of freedom θ of a rigid body under external torque τ can be expressed as
- I ω2 θ + i ω γ θ + k θ = τ
where I, γ and k are the moment of intertia, damping factor and the restoring coefficients, respectively. Since ω=0 in this measurement, it leaves the equation as
k θ = F L
with F and L being the applied force and the characteristic lever length, meaning τ = F L.
My impression is that: the applied force F on each coil may be almost identical within a factor of two to others given the yaw responses with the assumption that Ls are well adjusted to the designed values. Looking at the pitch responses, I am guessing that something might be wrong with k.
With that said, I am thinking of measuring similar responses at a sufficiently high frequency in which the equation should turn into
- I ω2 θ = τ
which should allow us to determin whether k is the culprit of the strangeness or not.