Yea, unfortunately I do not have monitoring on the secondary circuit. I have been checking temps with an infrared sensor, so those readings are only comparable between themselves.
I am considering buying the PAC-TH011E thermistors and connecting them to THW6/7 inputs and attaching them to the secondary pipes in/out of the buffer tank, possibly with a hose clamp and some thermal paste. At least the internal flow/return thermistors on the Ecodan unit seem to be done in similar way. Then I guess I would have them in HA, as Z1 Feed/Return Temp.
The next question is then what should the feed/return of both circuits be like? I read a lot about the dT5 out there, but I am yet to get the understanding, other than “heat pumps like it”. I have both seen people saying that the flow speed should be low to allow water to stay longer in the rads and allow them to dissipate heat, and others saying that it should be high. I have tried measuring, but there are too many variables to draw a conclusion.
About the modulation: It clearly can, it just doesn’t want to. At 16Hz/2kW, it runs for hours, but the issue here (as you confirmed previously), is that it only does that above a certain outdoor temp. In any case, I don’t think “my unit cycles a dozen times per day” will be enough to convince the installer to do anything about it, let alone replace the unit.
Regarding the 17th: it happened overnight, so I am 100% certain there were no open windows. The buffer temp cannot be the reason either, because it did 2 long cycles. Also, the tank also gets up to temperature in about 30m or less (100l)
It’s handy to know the delta T, but don’t spend too much money on it. The reason why I asked for delta T for the secondary circuit is to get an estimation of the capacity at that moment. And verify if its indeed not taking enough heat.
Regarding flow and delta T in general. Its different for radiators and UFH.
UFH:
high flow
low delta T
Radiators
lower flow
higher delta T
In general a heatpump is happiest when the output is ‘low’. So if we recall the formula
output = flow / 60 * (delta T) * 4.2, we can see that we have 2 parameters to make the heatpump happy (flow and feed temp (implicit via delta T)).
But since you have radiators, you need to maintain a higher delta T. I’m guessing that the radiators have a smaller surface to exchange the heat (higher delta T), and we would like the heat to be kept longer in the radiators (lower flow).
For UFH this is the opposite. So flow and feed temp are always 2 things that should be taken into account together. You just need to find the right balance for your system.
Mitsi probably has a valid reason to only unlock those frequencies at certain temps. But I do know folks running fake NTC sensors to unlock lower frequencies.
Yea, as I mentioned in another thread, I got the thermistors hooked to the buffer secondary outlet/inlet, so now I can properly compare the numbers.
Long story short, with the secondary pump on PP3 and and primary on level 2, the delta t are as follows:
~4kW/26Hz: primary and secondary average around 4.5 and 5C, respectively
~2.5kW/16Hz: primary and secondary average around 3.5 and 3C, respectively
(+/-0.5C variation)
Based on these numbers, maybe I could do with lowering both pumps a notch. I’ll have to try and see the result.
In any case, I came up with something that aliviates the cycling, based on your suggestion of heating in blocks. Basically I have automations in HA that increases the room setpoint to 23C when the heat pump starts running, and lowers it to 22C, when it turns off by itself, after reaching setpoint.
This is not 100% ideal, as it obviously allows for bigger indoor temp fluctuation, but it is in line with my original goal. It is also a solution that has a natural failsafe, in case any part of the smart home stuff stops working.
I’ve had this for a few weeks, and it seems like it averages at 1.5 cycles per day, so I am satisfied. For completeness, this is in AA, room temp normal mode 60m, thermo diff +5/-9, flow temp min/max 30/50C.
Now I would need to see how it behaves in negative temps, as the minimim output/frequency will be higher, but that will have to wait for next winter.
Depends on both flows, but you now have the data to tune to get it in balance.
Since it’s working in blocks, its an indication of oversized heatpump or undersized capacity (receiving end)
I only recently made this connection between the min frequency and the outdoor temperature.
It seems very arbitrary (in a similar way to the fan rpms being stepped based on outdoor temperature) rather than basing it on what is going on in the refrigerant cycle.
@johncantor do you know of any reason to have a minimum compressor frequency based directly on outdoor temperature rather than what is needed to get the output of the heat pump to the right level? It strikes me that if at mild temperatures it starts off to high it is much more likely to overshoot.
@gekkekoe those tweaker forum threads are very long and with the added translation needed I’m struggling to see if anyone came up with a refined way of wiring up the outdoor sensor so the temperature could be set to different values?
Do we know if other heat pump brands have these hard limits on the lowest frequency they will run the compressor at?
Yes, it makes sense to vary compressor speed on ouside temp. If you look at an old fixed-speed ASHP.. say it outputs 5kW at -5c, then it might naturally output twice that (kw) at +10 outside. This is an opposite ‘slope’ to what is needed. Its simply to do the source refrigerant pressure. The lower the refrigerant pressure (less density), the less energy transfer. The old ecodan seemed to ‘cap’ the output. it was a flat line at 5kW. Of course, you might be heating swimming pool in summer, so that cap could be a hinderance.
It would nice if the effect of outside temperure was adjustable. Annoying that my 6kW revs up high at start up even in summer
So much has changed since the original discussion, and yet, the behaviour maintains.
Yes, I am still using Ecodan’s dumb AA (until I can find the time and willpower to try gekkekoe’s flow temp AA). I don’t recall exactly what settings I have atm, but it’s probably flow temp min/max 30/45, room temp mode normal 60 minutes, thermo diff +5/-5.
And then I have home assistant lowering the room setpoint by 0.5C when HP turns off and setting it back to normal when it turns on.
This combination has worked up to -10 outdoor temp to reduce daily cycling to under 2 times on average. However, it would be so much more efficient and comfortable, if the unit didn’t have this minimum frequency on lower temps restriction, and if it also didn’t just start every negative temp cycle at max power…
And commenting on the finding on minimum frequencies per temp, I’d say they depend on the unit and probably some combination of settings. The lowest stable I’ve seen on mine: outdoor temp: >5 16Hz, >0 26Hz, >-3X 42Hz
I just wanted to add what I’ve found lately in order to improve run times and reduce cycling in my system - it’s something I haven’t seen you make reference to. I turned off my thermo diff altogether in AA. It’s lead to my system running continuously until it hits the room temps. I have a fully open loop system so I don’t know what this would do with a buffer as you have. I imagine you would need to turn the secondary flow rate down to ensure max dT to give it the best chance.
For ref I have time set to 10 mins and max flow temp set to 36. It’s still over shoots this but it seems to help limit the system. I’ve also turned on quiet mode level 2 but not clear yet if this has much benefit.
This trial has been 2 weeks now with the average outdoor temp of about 5 so it remains to be seen. What happens at higher or lower temps.
That is interesting, because it is my understanding that the HP thermo diff’s purpose is precisely to allow over/under shoot, and therefore reduce cycling. Though with all that the weird behaviours I’ve seen so far with this unit, I don’t doubt it.
I’ll give that a go once the outdoor temp reaches the problematic range, which is between 0 and -10 in my case. Warmer than that, it does very long cycles at low frequencies and happily keeps the house warm, and lower than that, it slowly starts struggling to keep up with the heat demand of the house, so it basically runs constantly as well. This because I set the max flow temp to 45, which was the best way so far I’ve found to keep it in check.
And while I’m here, I actually disabled the HA “hack” that I had in place, a few weeks ago, because the HP started to ignore it at some point, for whatever reason. It just stayed in “heating house” mode constantly (even when HP was not running), instead of going to idle.
I have also considered using the quiet levels to limit it, but just like the max flow temp, it does not seem possible to find a combination that is able to keep the house comfortable across the 40C seasonal delta(-22<>18), and does not require changing HP settings as outdoor temps change, while keeping compressor cycles in check.
Thermo diff has been off with a temperature limitation in place since 15th Jan if you want to take a look at the behaviour. Today is the first cold day since so we’ll see how it behaves and ‘adapts’ when the next temperature drops comes around.
The controller is set to a schedule of mostly 21 with a setback to 20 in the late evening and late afternoon.
I’ve been preparing a comparison document. Ideally I wanted more data but I may as well share now. Unfortunately there’s probably too many variables to make clean conclusions - certainly about efficiency. But one thing that is clear just from looking at the images is that it’s produced longer run times (bottom left is the present operating arrangement with thermo diff off).
It is not clear from those charts, but I suspect that the behaviour you are experiencing is actually due to the temp limit and/or quiet mode. Both should limit the compressor frequency (to different extents), which will effectively reduce over shooting, therefore reducing cycling.
If you can monitor the compressor frequency/output power, you should be able to see a difference there.
It may well be the temperature limit which is doing it. I’m less convinced that it’s quiet mode as I’ve implemented this before and there was no significant difference.
I’ll remove the temperature limit soon and see what difference that has.
I only have the Melpump app to monitor the compressor frequency which I started with last week so not enough useful data from that yet.
A lot has changed and I use a new method: