Issue with dispersion/rf response measurement in live control mode.

[gubaidulinvadim]

I'm trying to make an example for orbit correction using SOLEIL II configuration (#191). I think there's an issue with dispersion / rf response measurement in live mode. I remember I discussed this a bit with @JeanLucPons during the workshop.

I'm using the following config for the measurement:

from pyaml.tuning_tools.dispersion import ConfigModel as Disp_ConfigModel
from pyaml.tuning_tools.dispersion import Dispersion

dispersion = Dispersion(
    cfg=Disp_ConfigModel(
        name="DEFAULT_DISPERSION3",
        bpm_array_name="BPM",
        rf_plant_name="RF",
        frequency_delta=500,
    ),
)

control_mode.fill_device([dispersion])
dispersion = control_mode.get_dispersion_tuning("DEFAULT_DISPERSION3")

I'm able to get some rf response in design mode:

But there's no response in live mode.

@kparasch is it possible that some sleep is missing in pySC / pySC interface?

@gupichon I don't think it's our digital twin that creates the problem. I can measure other things (chromaticity, orbit) that use the same devices without an issue.

I've also tried to use rf_response from sr.design for correction on sr.live, then it works but we should be able to measure rf_response on sr.live, no?

Edit: it does seem to work with sr.design input for rf_response but our virtual-accelerator goes to ALARM state. So, probably it still fails to correct the orbit. I've had no issue with running orbit correction in sr.design mode.

[kparasch]

@kparasch is it possible that some sleep is missing in pySC / pySC interface?

It is certain!

I will make a PR but meanwhile you can try putting

time.sleep(self.set_wait_time)

between L53 and L54 in pyaml/external/pySC_interface.py

[kparasch]

Orbit corrections with rf_response is not yet implemented yet actually. It is one of the next things I am implementing.

[gubaidulinvadim]

If it's not implemented, then how did it work in design mode?

Edit: I think I was trying initially to put None for rf_response but this did not seem to work (not sure)

[JeanLucPons]

As @kparasch mentionned, RF correction is not yet (or partially) implemented.
You may observe steerer that try to compensate the dispersive orbit if you try RF shit.
We need for this steerer weights to avoid RF mode dilution.