autolog wrote:The current status is that the MK 1 set of scripts is working. I am in the process of reworking this so that it is all contained in one script (the MK 2). Once I have got that working and tested, I will try to turn it into a plug-in which as I haven't done one before will be more of a challenge
That would be brilliant, are you programming in python or applescript?
autolog wrote:The system as currently implemented enables you to set the following target Heat Setpoints:
An 'off' temperature e.g. 6 deg C
A morning 'on' temperature e.g. 20 deg C
An evening 'on' temperature e.g. 21 deg C
I only allow for two "on" periods in a day as I don't need any more. This wouldn't be too difficult to extend if necessary.
My need has a further complication. I am hoping to have a 15 deg setback, when a room is ready for use but currently unused. The idea is that the room can be brought to full heat when my daughter's iPhone is spotted in the house.
autolog wrote:The system also has a "boost" feature. Basically this looks at the current temperature and sets a Heat Setpoint a user specified number of degrees higher (e.g. 1 deg C) than the existing temperature being recorded by the Stella-Z thereby forcing it on. The "boost" lasts for a user specified amount of time (e.g. 15 minutes). It handles the situation where a "boost" is done just before a scheduled Am or PM heating period.
I'm not sure I understand the logic behind the boost, is it the result of observing the system?
autolog wrote:This all links to a common "boiler connect" relay (Secure SSR303) to turn on the boiler when any Stella-Z is calling for heat. I have just replaced the master thermostat with this device as that was the easiest way to interface into the original heating system. The Secure SSR303 has to be kept alive - it goes to default 'off' mode if it isn't given an 'on' or 'off' command at least once an hour. The system handles this.
I am hoping to use Fibaro relays but I am not sure if they can stand the inductive load of the zone control valves, I have three zones plus water, so I need four switches.
autolog wrote:One aspect of the Stella-z that is interesting to me is how the temperature is read. When the radiator is 'off' it seems reasonable. Once the radiator starts heating up the temperature goes up higher than I think it actually is. This is probably because the temperature sensor is located in the radiator valve. I might have to put an "allowance factor" in for this. Another possibility is to associate the thermostat in some way with another temperature sensor (e.g. on the wall) and use this to control the Stella-Z. I will probably put this feature into the script.
I have only done radiator-off testing so far, but was worried about that. I notice the stella has a gap low down on the front face, if that is the temperature sensor we may get better readings by pointing it away from the rad. I am hoping that a second sensor is not required to keep the cost down. Although, I have been using a second z-wave temp sensor while debugging, to make sure the stella is accurate. As you say, it looks ok with the radiator off. I suspect you are correct and calibration will be required.
We may also find we need to apply some control theory to this problem to prevent overshoot. The Stella spec said that we can directly control the water flow through the valve so a simple PID would be possible.
http://en.wikipedia.org/wiki/PID_controllerThis is for an oven, but a good read:
http://newton.ex.ac.uk/teaching/CDHW/Feedback/ControlTypes.htmlautolog wrote:If the script was to fail there are a number of fall back scenarios. Operate the Stella-Zs from the Indigo Devices list or schedules list to set the heat setpoints and control the boiler switch the same way. As a last resort, remove the Stella-Zs, refit the TRVs and use the physical buttons on the boiler connect relay to turn the boiler on and off.
Yeah .. my fallback position will be openzwave. I am *VERY* happy with indigo, but not being able to open the lid and tweak (which I fully understand) is a bit frustrating.