For COIL BALANCE for MC suspensions, I repeated her measrement all for the MC suspensions, and came to a strange conclusion.

Measurement conditions
- all Oplevs were centered before the measurement
- Oplev matrix are confirmed to be the same between the all three suspensions (which doesn't mean the all of them are correct).
- Oplev calibration factors [urad/cts] are off
- Applied offset on each coil +30000 cts

MCO

MCI

MCE

From the table, we can assume MCE's magnet sign of H1 or H3 is flipped. I cannot determine whether H1 or H3 by this measurement.
Presumably H1 is flipped, because so far -1 sign had been applied to H1 coil output filter.

Assuming MCE's H1 sign is flipped, PIT and YAW sign are opposite between MCE and MCO.
 
Assuming MCE's H1 sign is flipped, the actual coil configurations (coil assignments for the digital system) are all different.

MC coils signs doesn't make sense. Here are another tests. This entry is about the investigation for MCO.

I have checked oplev segment assignments are correct in the digital system w.r.t the actual quadrants of PD. We can believe oplev P/Y diginition.

As reported to this post, when each coil on MCO were actuated, motions in PIT vary depending on the coils. To check if any coil is touched to something, each coil was actuated again, with plus and minus signs.

The sign of coil output filters are (H1, H2, H3, H4) = (1, -1, 1, -1), according to Ohishi san's suggestion on this original post.

Result: Since the amplitures of the resulting motions are same for +20000 or -20000 actuation for all the coils. So they are not toughing to anything, but the balance between the coils is very bad for re-installed MCO. It's almost impossible to balance...

H1

H2

H3

H4

This entry is about the investigation for MCI.

First, oplev segment assignment in the digital system is confirmed to be correct compared with the physical segments of PD - which means we can believe oplve P and Y difinition.

We say,
1. Coil configuration is correct: from HR surface,
H1  H4
H2  H3

2. We believe the magnet polarizations are correct as:
N  S
S  N
from HR surgace.

3. We are not sure if the four coils directions are correctly installed. So there are uncirtaintity on the signs of coil output filters. For now, as a hypothisis, we use (-1, 1, -1, 1) to make the behaveir consistent with MCO reported in the earlier commet of this posts.

a. Responces to each coil with +20000cts
This agrees with MCO's behavior and consistent with conditions 1 and 2.

b. Responces to pair of coils with +20000cts

This is NOT consistent with the condition 1 because it's as if H1&4 are aligned vertically, and so are H2&3. It's either the coil configuration is different, as if the coils are
H1  H2
H4  H3
or the hypothesis of signs (which currently agreeing with MCO) is wrong. I don't think magnet configuration can be wrong (can it?).

Any ideas and suggestions?

Discussing with Nakanokun, as noted on http://klog.icrr.u-tokyo.ac.jp/osl/?r=484, MCI coil and MCO coil directions are different. For MCI, coil output filter signs of (1, 1, 1, 1) give the same direction of the force, where as (1, -1, 1, -1) for MCO. MCE is supposed to be the same as MCO. So the above measurement makes sense. Apparnetly, imitations of the coil wire lengths and so on when the manufacture made this inconsistent configurations of the coil directions. It is rather confusing, but it was the simplest solution at that time, according to him.

MCE is to be checked later today to make sure the connections and signs after the cable cut accident.

After the recabling of all MCE coils, their connections are tested as following. In summary, the cabling are correct and no setting change from the prior-accident configuration.

a. MCE's response for each coil actuations were:

so, coil out filter's sign should be (1, -1, -1,-1) to agree with MCO when each coil was actuated respectively. The configuration of coils are same as MCO, however, with the opposite magnet polarization of H3.

To push/pull all four, resulting coilout filt signs are (1, 1, -1, 1) to be applied on the digital system.

b. Just to make sure, here are responces to pair of coils with +20000cts (with coiloutF signs of 1, -1, -1, -1):

With (1, 1, -1, 1) (L actuation) on H1&H2&H3&H4 gave P=0, Y=0.2, which is not a very good balance of Y. But the behavior seems OK.