About the LVDT calibration in Type-B.

[MiyoKouseki]

Parent Issue #9

---

(From Fabian via email)

Comment and question to "https://github.com/gw-vis/kagra-vis/blob/master/calibraiton/LVDT/README.md"

Comment

With respect to calibration factors:
* BS F0:  I will check the situation once I can connect to the site from home; 
   I wan to check when the value of such channel changed.
* SRM F0: The value in klog is 0.1855 µm/count. The medm screen might have 
   round it off to 0.185 um/cnt.

Question

I checked some of the offsets and many are reported in the klog or can be 
calculated from calibration data. Still, you write in your list that they should be 
zero or other values. Where do those values come from?

[MiyoKouseki]

Answer

The absolute value of SRM-F0 is ok, but the sign is flipped.
I can check the time BS-F0 and SRM-F0 were changed because I have an
 operator's shift today.

The offset value is defined in "Calculate the calibration factor" in this document.
1) If linear sensor range is symmetric; e.g. from -2e4 count to +2e4 count, the 
offset value should be 0 count.
2) If the range is not symmetric; e.g. from -1e4 count to +2e4 count, the offset 
value can be represented by - (+1e4-2e4)/2 = - 5000 count in case that the middle
position of the primary coil in the measurement is truly the nominal position for 
operation.

[MiyoKouseki]

From Fabian

About the signs
In Type-B all the calibration factors in medm screens should be positive. We defined the polarity in hardware: when the keystone moves up the readout must increase. We modified the pin assignment in LVDT wires when necessary to make this possible. In IPs signal increase happens for positive yaw.

In two cases (SRM F0 and SR3 F0) the calibration factors are negative in the calibration plot because of the reference point selected to measure the displacement. Depending on mechanical access and ability of the person doing the job, the reference point was either above the keystone of below. When the reference point was below the keystone the slope is positive. When above, the slope is negative. Sometimes we corrected the data to remove such affect but sometimes we didn't. In either case the calibration factors in medm screens should be positive. (I have to check but there might be an exception and a hidden sign in BS IP-H2.)

About offsets
The offsets are selected to produce a readout of zero at the nominal position of the keystone according to the 3D CAD (or some other smart criteria). It's very hard to make the centre of the LVDT linear range coincide with that point.

TO DO:
Check the history of calibration factor in BS F0 (Miyo-kun who is at the site today.)
Check medm screens for BS IP-H2 which may have a hidden sign some place. (Fabian next week when remote access is allowed.)

[MiyoKouseki]

The calibration work of BS-F0-GAS is posted in 21 November 2018 (7072).

The gain value jumped from +0.886 to +1.116 around 17 Jan 2019 11:00.
But I could not find any strange activity on klog.

↑From 21 Nov 2018 00:00 JST

↑From 31 Dec 2018 00:00 JST [Zoom In!]

↑From 17 Jan 2019 00:00 JST [Zoom In!!]

[MiyoKouseki]

I found klog about the new calibration measurement of BS_F0_GAS
http://klog.icrr.u-tokyo.ac.jp/osl/?r=7694

Finally, the calibration factors of all type-B suspension are consistent with klog!

[MiyoKouseki]

Close