Hacking an Amftra Hestia BT thermostat

Despite to the oddity of our central heating system, the temperature control was far better than the Honeywell could do. Still, the Hestia reacted too fast and too unpredictable in the other, uncontrolled, rooms of the house. So one option was to go back to the shop and admit my defeat, before buying a more expensive one, probably to find out that it would be as inadequate as this one.
So I decided to hack the Hestia BT thermostat. See what it does. See what it should do. Try to understand the factors, and then decide on a modification.

There are some factors that I can change in order to fully optimize the thermostat by controlled hacking (aka product development):

Change hysteresis to 1 °C
Improve the airflow through the Hestia
Distance between Hestia and brick wall
Wooden panel between Hestia and brick wall
Chimney inside the case of the Hestia
Lower the maximum water temperature of the furnace

I installed the Hestia in the very same spot as the old Honeywell since that was the default place (the furnace control cable is available there). After setting the system (time, hysteresis, °C or °F, etc) it was mounted to the wall. But the control was rather clunky. So I started doing measurements.

My conclusion was, that the NTC of the Hestia didn't have enough interaction with the air. So I hacked the case of the Hestia. With a side cutter, a square file and a hobbyknife I took out two of the vent-slot separators which are closest to the NTC inside the Hestia. Plus the matching pieces above the NTC.

Now there is a gaping square hole right below and above the sensor and control got better immediately. It got better, but not perfect yet.
As a result of the fast heating (see above) at startup of the furnace, the warm air reached the Hestia NTC rather fast and the Hestia kills the furnace prematurely. So there is no residual heat in the system when it switches off. So the air temperature rapidly drops. So the furnace is restarted again. This produces a rather nervous system.

The next hack I did was raising the hysteresis from 0.5 °C to 1 °C. This means that the furnace is fired up when the temperature drops below (set point - 1) °C and commences heating until the temperature is at (set point + 1) °C. The general idea for this was that now the furnace has more time to fill the entire system with water so that all convectors get the chance to cooperate in the heating process and not just the one closest to the thermostat.

This was a disaster. It took hours for the temperature to drop 2 °C and when the system engaged, it got way too hot in the livingroom. So I reset the hysteresis to 0.5 °C.

Last night I changed the maximum kettle water temperature to an intermediate value (70 instead of 85 °C). Today, the system was much better in control. The heater kicks in with acceptable intervals. The residual heat in the system is comfortable. The change in temperature during one full heat/cool cycle was hardly noticeable.
Due to the hacked case, the sensor gets enough airflow and the Hestia is now able to generate high comfort in the house. Much better than the Honeywell did.

For the time being, no further experiments will be carried out. The measuring equipment will remain in place. Measurements are knowledge.

Pictures of the Amfra Hestia : http://www.mijnalbum.nl/Album=VOX3E4N8

Page created on 12 January 2007 and

Page equipped with FroogleBuster technology