Originally Posted by
Smurf
Agreed they should be tested.
One problem with some inverters though is that the thermostats are outright dodgy.
Say you set it to 22 and it's in heating mode. Obviously if the temp drops to 21 then you want the unit to do whatever it takes to bring it back up to 22. If that's running at 100% constantly then so be it, you want 22 not 21.
But some cheat. They'll never run at 100% unless there's a big gap between the setting and actual room temperature. And so the only way for the consumer to get the unit to work properly is to set it warmer than they really want - with some you literally need to set it at 28 just to get full output in a 21 degree room and warm it to 22.
All split systems ought to be limited to a normal operation temperature variation of +/- 1.5 degrees otherwise they fail MEPS. That is, if it's set to 22 then at the very worst it ought to be at full output at anything below 20.5 and at full cooling output at or above 23.5
Ideally, a narrower range is desirable as this will reduce energy consumption - most people set thermostats to maintain comfort at the minimum point in the cycle (in heating mode) so the less accurate it is, the more over-heating takes place. Given that a simple bi-metal thermostat has a range of less than 1 degree for a quality unit it's not difficult to do. For that matter, a simple Dick Smith kit keeps my fish tank within 0.1 degrees. So a $2000 air-conditioner ought not have temperature variation of 6 or even 2 degrees from the set point.
3 degrees from hottest to coldest (that is, hottest point in cooling mode to coldest point in heating mode including the changeover) is pretty well accepted as being about the limit for comfort. Hence +/- 1.5 is the practical limit from the set point for a reverse cycle air-conditioner. And since most measure the temperature quite close (or in) the machine itself, that reduces the margin even further.
