On Wednesday, October 23, 2019 at 10:08:04 AM UTC-4, Clare Snyder wrote:
On Wed, 23 Oct 2019 06:44:19 -0700 (PDT), wrote:



Very interesting. On some of these they use a flame sensor that relies
on just a steel rod, whereby somehow the flame itself generates a
small voltage. Seems that would rely on sensing a voltage between that
rod and ground, ie EGC. And if the generator is not properly grounded to
the house ground, to the furnace, I can see how that would be a problem.
And don't ask me about the physics whereby that flame sensor works....




Trader, some electronic controls NEED the neutral power line (white wire) to be at or near the same voltage as the safety ground (green wire) or else they will throw an error code.

If the neutral and ground are bonded together in the generator as they should be, it should work fine.

Mark
Some engineers are lazy, and the accountants win out - so the flame
sensor only uses one wire, resorting to the "chassis ground" for
return. If the chassis ground and the neutral are at different
potential, the controller mis-reads the sensor foltage and determines
that the flame has gone out - so it shuts off the gas and the flame
DOES go out.

The flame sensor is with the electronics, on the DC side, and one would
expect that the DC side would be tied to ground via the metal frame
of the furnace. It has to be to complete a circuit for the flame
sensor to work. So, I'm not seeing how a difference in potential
between the neutral and ground would be an issue. For that matter,
there is typically a varying difference in voltage between them
all the time. It will depend on the current flow in the neutral,
which changes. So, if the furnace is sensitive to that, why doesn't
it trip frequently on line power too?