Andrew Gabriel wrote:
> On 21/10/2024 13:21, Theo wrote:
> > Regular mechanical TRVs aren't able to call for heat, they can only control
> > flow from a boiler that's already running. If the boiler is off but one
> > room is cold, it can only get colder even with the valve wide open. If the
> > boiler is running it'll be wasteful if all the TRVs are throttling the flow.
> > If a TRV can call for heat, it can request the boiler run to actually change
> > the room temps, but only when needed.
>
> I've mulled this over in the past.
>
> One thought is to have a bypass loop with a pressure differential bypass
> valve, set to open only when all the rads are off. Also in this loop,
> include a flow switch. Any flow detected in the bypass loop can kill the
> demand for heat signal for, say, 10-15 mins. Obviously, you need the
> pump to run to detect if all the rads are closed.
>
> If I was implementing this, I would have the pump operate momentarily
> perhaps every 15 mins and checking the bypass flow switch, just to test
> if there's a radiator open and it's worth actually demanding heat from
> the boiler. This will depend how integrated the pump control is with the
> boiler.
>
> In practice, boilers are not great at driving just one rad at a time,
> and you really want some logic which says something more like, when
> there are enough zones calling for heat to absorb a significant
> proportion of the boiler's lower output power, then demand heat, unless
> there's a zone sufficiently far below its setpoint which can't really
> wait for more zones to call for heat before demanding heat from the boiler.
Sounds a bit complicated to me :-) This is strictly a wiring-only solution,
I don't want to touch the wet side.
Thinking is that the smaller rooms (bathroom, utility etc), and also the hall
where the thermostat is, will be non-controlled so there will always be some
flow. Then the call for heat will be in parallel with the hall thermostat
(which can be set low). Call for heat on any TRV will then fire the boiler
even if the hall is up to temps.
This boiler is dumb and doesn't modulate, it also doesn't weather compensate
- there's no OpenTherm or anything, the controls are just on/off. I don't
know what level of buffering there is in the system (there's no specific
buffer component), but how does a dumb boiler cope when most of the rads are
turned off? It'll short cycle but is that a problem when it's dumb?
> Having a setpoint range for each zone (which can also be used to
> implement hysteresis) gives more flexibility. When one zone drops to the
> lower acceptable setpoint, demand heat for that zone, and when the
> boiler has got that part of the heating circuit hot, bring in more zones
> which are below their upper setpoint and coming closest to their lower
> setpoint, such that the open zones can dissipate the boiler's lowest
> modulation power. I did hack this up in software controlling my system.
I came across this:
https://github.com/tyrken/heatmon
which is something that talks to the radio on the Radbot and extract
temperature stats and valve state - I think with that I should be able to
extract a call for heat.
Ordered a Radbot to play with and the new red version (Radbot 1) has a
different testpoint pinout from the previous orange version (Radbot 2R,
before the company was bought out by Secure Meters), so I'll have to take it
apart and see if they've changed any of the hardware and whether that still
works, or whether I need to scour ebay for 2Rs.
(Radbot is a lot preferable to systems with schedules IMHO, where you have
to programme them according to your routine - which assumes you do actually
have a routine, otherwise you have to endlessly fiddle with them, often
through complex and/or awful UIs[*])
Theo
[*] every programmer with a calculator screen, I'm looking at you...
|
|