[SOLVED] AC signal processing in PIC

[betwixt]

I'm siding with Susan on this topic. Simply waiting for zero crossings to coincide then closing the breaker is asking for trouble. It will stop the instantaneous current as the switch closes which is good but one of two things will happen immediately afterwards: 1. The hydro will be damaged, 2. the breaker will be damaged, depending on which is weaker. What you are trying to do is essentially the same as when power stations are brought on line, you use a synchroscope to compare the phases and crucially, you adjust the frequency (generator RPM in a power station) until the two source frequencies are the same. If you switch two different frequencies together, even if their crossing points are the same, you will drive power from one to the other instead of both working in harmony.

What you need is to sense the frequency phase of the line power then adjust the hydro inverter frequency and phase to be exactly the same before closing the breaker. I use a similar arrangement with PV grid-tied systems and they take three minutes to synchronize frequency and verify it stays constant, you would be expecting yours to do it in just a few cycles!

Brian.

 

[denujith]

Simply waiting for zero crossings to coincide then closing the breaker is asking for trouble.

Yes V, f and Slip have to be matched. From the start, I wanted to know how to measure slip (phase angle) since V and f are adjusted to match the running parameters.

 

[betwixt]

Use an LPF to eliminate the high frequency pulses so all you have is the underlying 50Hz. Then you can compare the phase using a PLL, which will give you an analog error voltage or simply an XOR logic gate fed from 'squared up' waveform. The average output of the XOR gate will be proportional to the difference in phases.

Brian.

 

[KlausST]

Hi,

be careful. The LPF will introduce phase shift.

I always find it critical to use an LPF with synchronizing circuits.
If you want to use it, then be sure to have exactely the same phase shift with both input signals.

Klaus

 

[DanyR]

Hi, I do struggle with the same type of questions regarding grid tied convertors, see this thread: https://https://www.edaboard.com/threads/346709/

 

[denujith]

Thank you all for your valuable inputs. I will start building it and will post here with the results. My plan was shaped by your inputs and they helped a lot indeed

 

[betwixt]

Klaus is right about the filter introducing a phase shift but the amount will be constant so you can either factor it out when calculating the switching time or use an identical filter in the other signal as well. The danger in doing that is although the two filtered phases to the detector may coincide, the main power may not be at exact zero voltage switching point. The delay will be minimal and the two voltages should be equal even if not zero so it should still work OK.

Brian.

 

[denujith]

Klaus is right about the filter introducing a phase shift but the amount will be constant so you can either factor it out when calculating the switching time or use an identical filter in the other signal as well. The danger in doing that is although the two filtered phases to the detector may coincide, the main power may not be at exact zero voltage switching point. The delay will be minimal and the two voltages should be equal even if not zero so it should still work OK.

Yes I can calculate when it will be zero in advance. I am so confident now. Thank you all!