According to my measurements I conclude that LSCsignal is retarded from OMCsignal with the constant retarded time of 92usec.
It means there are no timing jitter between them. Only a constant timedelay exists.
(Timing jitter)
Let's begin with basics.
If you measure the same signal at OMCside and LSCside, they can have some time delay between them. It can be described as followers.
where tau_0 is the time delay. If the tau_0 is not constant, it causes a noise of the timing jitter.
(method)
I have injected the sinewave with 200.03Hz into the OMCLSC_DRIVE_EXC. Then by using get_data, I measured the signal at 'OMCLSC_DRIVE_OUT' and 'LSCDARM_ERR' where the exciting signal comes out.
In the ideal case the two signals are completely identical.
In order to find the delay, I calculated the difference in these signals based on the method described by Waldman. The method uses the following expression.
Here the tau_s is the artificial delay, which can be adjusted in the off line data.
By shifting tau_s we can easily find the minimal point of the RMS, and at this point we can get tau_0=tau_s.
This is the principle of the method to measure the delay. In my measurement I put T=1sec. and make the calculation every 1sec. in 1 min.
(results)
Attachment is the obtained results. The above shows the minimum RMS sampled every 1sec. and the below shows the delay in terms of number of shifts.
1 shift corresponds to Ts (=1/32kHz). All of the data are matched with 3 times shift, and all of the minimum RMS are completely zero.
Therefore I can conclude that LSCsignal is retarded from OMCsignal with constant retarded times of 3*Ts exactly, and no timing jitter has been found.
