Posted by: @mmiuse

↑

@derek-m I have a bungalow. 4 bed. around 100m2. fully refurbed in 2020/2021 with UFH throughout, each room is a separate zone. yes there is a low loss header.

A heat loss of 7.4kW sounds about correct for a bungalow of that size.

Looking at the last graph in the series it would appear that the cycling is probably due to defrosting, which is normal this time of year.

There would appear to be two distinct problems, the first being that the heat pump is not achieving the specified LWT, and the second that adequate heat energy does not appear to be transferred into your home.

As a quick test, try increasing the LWT setpoint to 50C and see if it produces an increase in the actual LWT. Report back once completed.

Posted by: @mmiuse

↑

@derek-m I have a bungalow. 4 bed. around 100m2. fully refurbed in 2020/2021 with UFH throughout, each room is a separate zone. yes there is a low loss header.

The next thing that I suggest is logging in under 'Serviceman' and going through all the various screens available and record the values/settings that you find.

Posted by: @derek-m

↑

The next thing that I suggest is logging in under 'Serviceman' and going through all the various screens available and record the values/settings that you find.

@derek-m - about half of the values/settings behind the tradesman's entrance are either blank or fathomless numbers meaningless to man, at least to ordinary Joes. Even with the Midea manuals they are still often impossible make sense of, as we recently found out when trying to understand the IBH/TBH mystery. TBH, I think the main thing to focus on is the reported parameters, and try to make sense of them.

Posted by: @cathoderay

↑

Posted by: @derek-m

↑

The next thing that I suggest is logging in under 'Serviceman' and going through all the various screens available and record the values/settings that you find.

@derek-m - about half of the values/settings behind the tradesman's entrance are either blank or fathomless numbers meaningless to man, at least to ordinary Joes. Even with the Midea manuals they are still often impossible make sense of, as we recently found out when trying to understand the IBH/TBH mystery. TBH, I think the main thing to focus on is the reported parameters, and try to make sense of them.

@mmiuse - I can see kW in/kW out in your second image. Not sure why it says kW (power) rather than kWh (energy) which it would more normally show, but can you do a plot over time of those two variables? That way you will be able to see how COP varies over time. Even better if you can put them in a spreadsheet and calculate the COP at each interval, to end up with a plot like mine above. I have noticed sometimes Midea aggregated COPs over a time period are rather lower than expected.  

I would prefer to see the data from the Midea controller rather than their add on box of tricks. Whilst I appreciate that you are a 'big fan' of 'Freedom' 😜 , it is the data on the Midea controller that is actually performing the operation and control. I don't know if the 'add on' is getting data from the Midea controller, separate sensors, or a combination of both. 

Since I don't have a Midea to play with, any useful data you can suggest, and how to obtain it, would be much appreciated.

Posted by: @derek-m

↑

I don't know if the 'add on' is getting data from the Midea controller, separate sensors, or a combination of both. 

From @mmiuse's description it sounds very like the setup I have been investigating, right down to the two wires into the wired controller. If so, the 'incredibly simple to install' box of tricks is using modbus/RS-485 to hoover data off the wired controller. If it has RS-485 written on it somewhere, I suggest that pretty much confirms the setup is as I suggest.

The problem with monitoring the wired controller itself is twofold. Firstly, many of the key parameters (on the read only Operational Parameter pages, which are not behind a password, unlike the 'For Serviceman' pages, which are are, and are where you can set more 'sensitive' settings) very volatile, eg LWT and RWT, current draw and capacity (output). Secondly, it is almost impossible to keep flicking through the various page accumulating data, and if you do manage to do it, you lose hours of your life that you will never get back again.

Nonetheless, it would certainly makes sense to see if the online data matches the data visible in the wired controller, perhaps by taking LWT and RWT readings from the wired controller once a minute for ten minutes (with times of readings), and then later see whether the same numbers appear in the online 'app' which it seems really is a black box from the images.

I think I went through the 'essential' or core data set in the Beginner's Guide thread, but here are my suggestions for that data set:

Current draw, and volts as they are there (to calculate energy in) 

Primary LWT/RWT/flow rate (to calculate heat/energy output)

The above two mean you can calculate COP; having LWT/RWT/flow rate are also interesting in their own right  

Ambient temp because it is such a central parameter to everything else

If you have a secondary circuit for the emitters, then LWT/RWT/flow rate for that is a useful addition, but needs extra sensors because there is no way that data can be got from the wired controller. All the others are visible in the wired controller, if the box'o'trix is pulling them out correctly, and they are reliable and valid, then 99% of the job is done.

@cathoderay perhaps we can get the code off the pi/Arduino?

It will have some kind of operating system which we can probably get into easily enough. I imagine it's an on SD and not encrypted 

Posted by: @batalto

↑

perhaps we can get the code off the pi/Arduino?

Now that is an interesting idea. Not sure about the legality - my view (and I am defo not a lawyer) is that in 2023 it is absurd to suggest anything other than our data is ours. But Headroom, who are probably already assembling lawyers in the Headroom boardroom, may take a view that the software is theirs, and not ours for the hacking. I also would not wish to have @mmiuse left feeling awkward/compromised.

Of course they (Freedom/Midea) could solve all this by making the code/protocols open source. I can't see there is any commercial sensitivity in letting householders get access to their own data. Unless, of course, there is a whiff of dieselgate in the air...  

If their modbus-> Wi-Fi gateway is built using a rPI then it will be running Linux and some sort of daemon. Provided this daemon isn’t compiled (written in C or Rust) then the code will be there for all to see.

If their modbus -> Wi-Fi gateway is built on an arduino or other, then it will be flashed with a binary which will need to be extracted and de-assembled.

The protocol is open (modbus rtu) and the register map will be available from midea - it’s only the glue that’s coming from freedom. You could just as easily write your own.

Posted by: @hughf

↑

The protocol is open (modbus rtu) and the register map will be available from midea

I agree, the protocol is open and the register map is available from Midea, it is even in the Operation Manual for the wired controller, along with a note about where to make the connections (H1/H2, these are the RS-485 A/B terminals, at least that is how I read the manual). @batalto's point is in effect no need to reinvent the glue if someone has already invented it, and I agree, my concern is the inventor/manufacturer of the glue might go all proprietary on us, so best to tread carefully.    

@cathoderay I don’t see the issue here then? Freedom have taken the register map and built a product (FHPCloud) out of it and some off the shelf hardware.

Perhaps I’m missing the point of the argument against FHPCloud… I mean, what’s the alternative for average Jo?

@hughf - from what I have read, it appears Freedom charge a lot - around £100? - to provide this. I don't see them taking kindly to us depriving them of that income by doing it ourselves using their blueprint (if we do it independently, as I am trying to do in the Beginner's guide to heat pump monitoring thread, that's another matter, none of Freedom's business). Furthermore, I don't want or need to use anyone's cloud, for much the same reason I don't like 'loyalty' cards, as it can be done locally, I want to do it locally. But I agree, for the average Joe, the Freedom cloud way provides an out of the box, and in my view over-priced, solution.

By the way, I think @mmiuse has been given access to the 'installer dashboard'. Freedom appear to have two dashboards, one for installers, with a lot on info, and a much more basic one, on a par I think with the Midea app version, for householders. Most if not all ordinary Joes will end up with this version.

There is another important caveat here. We are in effect allowing the manufacturer to tell us how good their product is, on a par with relying on a car manufacturer to tell us how good their car's mpg is, or how low their emissions are. As dieselgate showed us very plainly, their is nothing to stop even supposedly reputable European manufacturers from cooking the books behind the scenes when it suits them. Don't even start me on covid vaccines, perhaps one of the biggest book cooking exercises ever seen in human history by the likes of Pfizer, who also, as it happens, have form on this sort of thing. Why wouldn't Midea want to do the same? That's why getting the data in its rawest state matters, and the wired controller is about as close to the sensors as we can get without adding third party hardware.    

@mmiuse can you copy the SD?