Someone needs to make a relay module

I know that MyForest has tackled this topic in detail, but as an easy fix (easy improvement) for ASHPs doing what my graph shows, maybe a simple module with relay-off could help.

A simple bit of code… [ if power drops from say >400w to <50w… (its stopped) hold relay off for a time ] . The time could be related to outside temp or simply a ‘frugal-ness’ setting.
The graph shows that the 5kW Panasonic can only go down to 3kW… shame its not lower. Anyhow to reduce the average heat to system down to say 1kW would be difficult. However, by ‘holding off’ for say 1/2hr after stopping, then the radiators would cool, and on starting, it would look a little more like the session from 17:30… average flow temperature lower and better.


Linking to @MyForest’s thread on Improve efficiency: Turn the heat pump off, which proposes the same idea but through software.

A hardware relay would certainly be a useful option, especially when a software solution isn’t available.


I often look for simple low-cost solutions. An EmonTH with flow & return sensors could look at dt. If dt drops by say 3 degrees suddenly, then the unit has probably stopped. It could then hold a relay off for a while…

1 Like

What is the current draw? You could simply use a contactor and a Sonoff Mini (or something similar).

Most ASHPs now have a handy ‘smart grid’ input… a control switch input to turn off when the tariff is expensive. Not actually used one recently. (German GSHPs had this for 20 years)

1 Like

Great. Have you set anything up to control one?

My 11 kW EcoDan has a pattern of running two cycles of 15-20 minutes each…

The first cycle ends when the flow is above target temperature (28°C), then it pauses for a couple minutes before starting another cycle that well exceeds the target temperature. It will sometimes do third cycle, othertimes it will wait for the flow temps to drop.

The COP doesn’t seem to suffer massively, though each cycle tends to have slightly lower COP than the previous. I’m now wondering if I should turn the HP off for some more minutes between cycles to keep the flow from getting too high.

(I would really prefer one long cycle, but my HP just can’t modulate low enough in these mild temperatures).

That shocks me. The 8.5kW Ecodan often does 3.5kW. I forget what the book minimum is.

In which mode?

Heating mode. During the hot water cycle it will happily run for a solid hour until it reaches target, so I don’t think it’s a hardware issue. I was able to get longer heating runs when it was cold last winter, just not this season yet.

Is that a new R32 Ecodan? do you actually get steady-state at 3.5kW?
I’m guessing the 5kW Panasonic is really a 7kW unit. still, its not a very good turn-down ratio. The 6kW Ecodan is really an 8.5, and mine (6kw Ecodan) won’t stay at 3kW for more than 2 mins before it strays up to 6kw… then stops.

Auto adapt, compensation curve or set flow?

Yes and I think so but it is not called for for long yet in this mild weather so I doubt I have pretty graphs. I’ll check and post if I find it.

Its been on both fixed and curve. The performance seem the same…, as you might expect. I have tried auto adapt, but have never managed to get steady state. My old one does it quite nicely and often purrs away happily for hours.

Set flow temp, which is controlled externally via a script to implement weather compensation and room influence.

1 Like

Seeing quite different behaviour today, with 20 minute cycles separated by longer gaps, and the CH circulation pump staying on throughout. Is this better? Would turning off the heat pump in between cycles disrupt the controller’s own logic?

Interesting that the max, flow and return temperatures rise on each cycle.

This is mostly due to the target temperature increasing during this time too.

Why does the target temperature increase and not stay static?

Makes sense, perhaps add to graph.

Because (I think)

1 Like

I’ve discovered that my system tends to cycle when flow temperatures are under 35C. Above that it does better at doing longer runs.

  • The first run of 42 minutes to get radiators up to temp (target = 33C) COP over 4
  • Then three cycles of 20 minutes each. COP drops to 3 each time, gradually recovering.
  • I manually turned off the heat pump for ~25 minutes, letting system cool
  • The next cycle is 27 minutes, with a stable COP of 4+ for the entire period.
  • The following cycles then fall back into the minimum of 20 minutes with lower COP.

This does suggest that turning off the heat pump between cycles, as John proposes, might help with the performance of the system. I have implemented a “software relay” to force a pause between cycles. I’ll report back how well it works…


Good to see your investigation here. Yes, the benefit of the first section is the low average flow&return temperatures. during ‘off’ periods, radiators are cooling, and this is an advantage for the next period, BUT unfortunately the system can ‘revs up’ in a mild panic to get up to the target flow temperature, but you are showing here that this cold start ‘rev up’ isnt really a problem. I actually did exactly the same test on a Panasonic today… held it off for 20 mins. In my case, the difference between its natural 8min rest and my forced 20min rest was minimal. its such a shame that ther isnt an option to vary the ramp-up speed of the compressor… This could be varied due to outside temp. slow ramp-up in mild weather.
Well done for your hold-off function. Will be interesting to see how well it works.