I am not familiar with that machine, but I assume this is AC picked up on the rotor, rectified to create a DC, to be the adjustable 'permanent magnet' that rotates inside the main generating stator. That would seem about right for a machine of that size.
So, if that is right...In normal operation, the varistor does not conduct, but
Consider a violent back EMF on the generator output (a load fault or a nearly lightning strike) - at a snapshot in time some of the stator coils are in the position to be transformer coupled into the coil that is this rotating magnet, and depending on the impulse polarity, may be pushing back against the rectifiers but with nowhere to go. Unchecked the voltage at this point may keep rising until the diodes break down, shorting their 3 phase excitation winding.
Far better then to clip the voltage by allowing the varistor to break down at a slightly lower voltage than the diodes - the varistor has a characteristic as per the example curves below and is the basis of the modern surge protector. Once the voltage surge has passed, the varistor stops conducting and the rectifier carries on un-damaged, the AC output recovers within a fraction of a cycle; the lights may flicker a moment but everyone is much happier if they stay on afterwards ?
If the alternator could be guaranteed to only ever drive slowly changing resistive loads, and be nowhere near switching transients or thunderstorms the varistor would not be needed, but in reality it is.
The one in this genset will be probably a very large area one, rated to dissipate some thousands of amps surge for some tens of microseconds, and will be more like the red curve, but with the current calibration in hundreds of amps not A.....
It protects the rectifiers, but also limits the peak voltage that can be induced in the rotating coils so protects them too, when just using more or higher rated diodes would not.
I am not familiar with that machine, but I assume this is AC picked up on the rotor, rectified to create a DC, to be the adjustable 'permanent magnet' that rotates inside the main generating stator. That would seem about right for a machine of that size.
So, if that is right...In normal operation, the varistor does not conduct, but
Consider a violent back EMF on the generator output (a load fault or a nearly lightning strike) - at a snapshot in time some of the stator coils are in the position to be transformer coupled into the coil that is this rotating magnet, and depending on the impulse polarity, may be pushing back against the rectifiers but with nowhere to go. Unchecked the voltage at this point may keep rising until the diodes break down, shorting their 3 phase excitation winding.
Far better then to clip the voltage by allowing the varistor to break down at a slightly lower voltage than the diodes - the varistor has a characteristic as per the example curves below and is the basis of the modern surge protector. Once the voltage surge has passed, the varistor stops conducting and the rectifier carries on un-damaged, the AC output recovers within a fraction of a cycle; the lights may flicker a moment but everyone is much happier if they stay on afterwards ?
If the alternator could be guaranteed to only ever drive slowly changing resistive loads, and be nowhere near switching transients or thunderstorms the varistor would not be needed, but in reality it is.
The one in this genset will be probably a very large area one, rated to dissipate some thousands of amps surge for some tens of microseconds, and will be more like the red curve, but with the current calibration in hundreds of amps not A.....
It protects the rectifiers, but also limits the peak voltage that can be induced in the rotating coils so protects them too, when just using more or higher rated diodes would not.
