On the PA28 carb heat should be checked on the downwind and then selected off for the approach as selection to hot for the approach can lead to icing on this aircraft
I believe the logic is that as the heat is supplied by the exhaust manifold, an idle engine would not only produce much less heat, but as stated above actually take the air temp into the icing zone. But I would have thought that one must have visible moisture as warmed air will be less humid than cold, regardless.
I certainly don't recall ever seeing any official words suggesting that use of carb heat can cause icing.
As above and to follow on from above, is that properly functioning Carb Heat will heat the air to circa 150 deg C, which should zap any ice. However, the warning was to not use partial carb heat for that very reason - unless you have a carb thermometer (I believe piston helos do). It is also the reason why a less than optimal carb heat system (often a corroded box) is a no go.
Ice crystals in the air would be visible as cirrus cloud. That would indeed be "somewhat unusual" on the approach!
Good point. I used to think the same thing, but I recently received instruction on the use of WX radar to the effect that they are less effective when aimed at altitude due to the CBs at that level being predominantly ice. What happened to supercooled water droplets?