Strange test results, generator supplied installation

If there is a CPC from the point of test to the genset, the earth rod is not affecting this result. (You need one, so good you have,  but it is not in the loop for these tests)


You have not said anything about the genset, but the output impedance of a genset is a function of how fast you measure it, as the controller will wind up the throttle to compensate for falling volts if it can. I suspect that the two testers are being fooled in different ways, the MFT is presumably a 'no trip' design, and as such draws a very small current L-E, but modulated with a very unique pattern and looks for changes in the LE voltage that match that pattern. The genset controller, especially on an inverter genset,  may well be unable to partly keep up with that, and gives you one answer, as it is carefuly winding the voltage up and down and filling the gaps in the meter waveforms. A large genset with a regualtion time of seconds won't do this, and looks more like the mains.

The other meter will presumably make a similar test, but with a different value of test current and over a different period, which may or may not be tripping the genset controller to respond.

All readings will also probably depend on the current setting of the controller, so will vary a lot with what other loads are connected at that instant.

The mains at home by comparison is essentially a zero ohm source, or at least a very nearly constant voltage, and the reading is dominated by the cables and joints en route which are not non-linear.

Note that a real fault on a genset of any size takes the controller out of the place it can compensate for,  with a long overload the regulator winds up to max and then sulks, with a fast fault it cannot, and what happens depends on the load of the day - and it is even possible to stall some gensets,  indeed 'a friend' once had to climb into a large container alongside a big diesel engine of lorry size, with a hammer and a replacement drive pin after such an event - the pin is a deliberate weak link, so that if you do stall the windings, it spaps off and you do not break the cranksaft. Moral, something about taking care when paralleling to repect the phase sequence.

Also note that for the same reason you cannot step a large fraction of a gensets rated load on or off suddenly without problems - the load needs to ramp slowly or come on in stages. Lighting for stages and sound systems can be especially stressfull or a genset for one big motor.

So fault currents, and predicted PSCC from loop tests are often a long way different on small gensets. The variation gets less when the genset is well loaded, and the machines are bigger and control loops are  slower.

If there is a CPC from the point of test to the genset, the earth rod is not affecting this result. (You need one, so good you have,  but it is not in the loop for these tests)


You have not said anything about the genset, but the output impedance of a genset is a function of how fast you measure it, as the controller will wind up the throttle to compensate for falling volts if it can. I suspect that the two testers are being fooled in different ways, the MFT is presumably a 'no trip' design, and as such draws a very small current L-E, but modulated with a very unique pattern and looks for changes in the LE voltage that match that pattern. The genset controller, especially on an inverter genset,  may well be unable to partly keep up with that, and gives you one answer, as it is carefuly winding the voltage up and down and filling the gaps in the meter waveforms. A large genset with a regualtion time of seconds won't do this, and looks more like the mains.

The other meter will presumably make a similar test, but with a different value of test current and over a different period, which may or may not be tripping the genset controller to respond.

All readings will also probably depend on the current setting of the controller, so will vary a lot with what other loads are connected at that instant.

The mains at home by comparison is essentially a zero ohm source, or at least a very nearly constant voltage, and the reading is dominated by the cables and joints en route which are not non-linear.

Note that a real fault on a genset of any size takes the controller out of the place it can compensate for,  with a long overload the regulator winds up to max and then sulks, with a fast fault it cannot, and what happens depends on the load of the day - and it is even possible to stall some gensets,  indeed 'a friend' once had to climb into a large container alongside a big diesel engine of lorry size, with a hammer and a replacement drive pin after such an event - the pin is a deliberate weak link, so that if you do stall the windings, it spaps off and you do not break the cranksaft. Moral, something about taking care when paralleling to repect the phase sequence.

Also note that for the same reason you cannot step a large fraction of a gensets rated load on or off suddenly without problems - the load needs to ramp slowly or come on in stages. Lighting for stages and sound systems can be especially stressfull or a genset for one big motor.

So fault currents, and predicted PSCC from loop tests are often a long way different on small gensets. The variation gets less when the genset is well loaded, and the machines are bigger and control loops are  slower.