User story #2561
Review datasheets of PS of RF system and check code of managers
Related issues
History
#1 Updated by Angela Salom 9 days ago
- copied from User story #2553: Provide datasheets of PS of RF system added
#2 Updated by Angela Salom 8 days ago
- % Done changed from 100 to 0
#3 Updated by Iago Dapena 7 days ago
- % Done changed from 0 to 20
The PLCs SW controls the power supply by means of a manager. This manager does the following functionalities:
1.Monitors V and I to be within alarm ranges.
2.Monitors V and I to be within warning ranges
3.Calculates the voltage slope and applies the value to the power supply.
The control has 4 modes which they are, probably, set from EPICs.
- OFF: The PS is powered off. The PLC monitors
- ON: The PS is on. The PLC only monitors.
- NOMINAL: The nomimal output voltage has been achived. The PLC only monitors
- RAMP: The ramp is calculated and applied to the PS.
Possible malfunction causes
- During RAMP mode, if there is a warning or alarm the ramp stops. Is the state machine capturing this?
- When in NOMINAL the PLC is not updating the target voltage to its output. In the event of ramp failure, the PLC could be writing a wrong value to the target voltage.
Further questions
- In which state is the state machine when the problem occurs on the filament PS, for instance?
In order to gather more information it would be valuable to connect to the system while the problem is occuring. Would that be possible?
#4 Updated by Ivan Moya 7 days ago
Possible malfunction causes
- During RAMP mode, if there is a warning or alarm the ramp stops. Is the state machine capturing this?
Re: Uhmmm... not sure. Typically, it should do it, but I cannot find an example now.
- When in NOMINAL the PLC is not updating the target voltage to its output. In the event of ramp failure, the PLC could be writing a wrong value to the target voltage.
Re: I remember the problem very little times on filaments and more in Screen grid.
Further questions
- In which state is the state machine when the problem occurs on the filament PS, for instance?
Re: the system should be in "setting filaments". In the case of the screen grid it is the "powering up tetrodes"
In order to gather more information it would be valuable to connect to the system while the problem is occurring. Would that be possible?
Re: If it happens again we can try to do it. But it is totally random for us.