In a recent thread, I asked for info on the vague control of the wireless CM67 ch stat. It was recommended that I disable the optimisation. It is not clear from the installation or user manual how much this function does.
If I disable the optimisation, will the heating system actually come on at the time set on the stat rather than coming on earlier and trying to achieve the set temperature by the set-time? Or does the optimisation simply alter the degree of earliness?
Hope someone can understand that last sentence. Must be the most incomprehensible giberish I have ever posted to this group - sorry!
My understanding is that, if you turn optimisation off, the heating will come *at* the set time - not earlier.
So, if you want the house to be warm by (say) 07:30, programme it to come on at 06:30 (or other suitable time, depending on *your* estimate of how long it takes).
I've still got optimisation turned on on my CM67, but I have to admit that it does seem to come on ridiculously early on occasions.
I imagine it is supposed to "learn" over time, how fast your house heats up. That will allow it to predict when to start, in order to achieve the temperature you have set.
In practice, I think it would also have to monitor how fast the house was cooling down beforehand, as the rate of temperature increase is related to the outside temperature, which it cannot know directly (without an outside sensor) but which it could deduce from the temperature drop.
I'd be surprised if it was that smart, but if its model of temperature gain was sufficiently adaptive, it could cope with external temperatures changing over a few days.
It cannot know when the hot water is heating so If happens while the cm67 is tryint to heat the house then it will "learn" the house takes longer to heat and come on too early next time.
Honeywell Chronotherm literature for different models suggests 4 - 8 days for learning the characteristics of the house and the local climate. There is some suggestion that time/temp program changes may also need some re-learning over a period of days.
I suspect Honeywell have been developing the Chronotherm algorithm for some time. I would not be surprised if they have had to use all available information, including temperature fall rates, to get the most accurate predictions.
Also, I understand the Honeywell Chronotherms have a "wall temperature" sensor as well as an "air temperature" sensor. So they might well be able to monitor how the building fabric cools as well as the rate of cooling of the air.
I did notice that moving my CM67RF from its initial installation in free air to a wall mounted position did raise the achieved air temperature by about 0.5 C all round the house (for the same program) so it looks like it does sense and react to wall temperature.
I doubt that it can cope with all situations but so far mine seems to be fairly smart. However, my "settled" CM67RF will sometimes fail to reach the set temperature by the set time in the early morning after an "unexpected" very cold night. However it's usually just by 0.5 C, and not very often.
Also I went on a three-day business trip just before the start of the recent cold spell. I left it in holiday mode. On my return to a UK temperature of -1 C, the house was again about 0.5 C low of the target after being at 12 C for 3 days. Not perfect but pretty good IMHO.
Actually my CH system's combined radiator output is somewhat underpowered for the house's calculated heating requirements (circa 67%) so it may be that sometimes it needs more time than the 3 hour max advance. However I have not been up at that time to check.
I have to say that so far (for me) the optimization has not failed to produce either the desired temperature or something very close at the desired time. This has allowed me to reduce temperatures at other times.
Even if the CM67RF starts the heating before the due time I have no evidence that this is excessive for reaching the demanded set point.
But if the DHW recovery + CH recovery happens the same time regularly then the CM67 should learn from experience that it takes a bit longer than normal (for whatever reason). So it should then adapt anyway.
The problem with that argument in the case of our family is that DHW use is not consistent. So some nights the tank will be drained and require serious heat input during the morning ch period, other times it won't. The other obvious option is to keep hot water and heating to separate time periods. Trouble is that means the noise of running systems kicks off at maybe 4:00 am for DHW and we are reliant on tank insulation to maintain the temperature for those showering in the am.
Another option is to use the immersion heater..in conjunction with a timed boiler. And set the stats on each differently..
so e.g. eh immersion heater will take it up to say 40C overnight, and the boiler can kick in and take it to 60C at 7 a.m.
I found the same, and ended up running three periods for DHW, one at about 7 a.m. so hot water is always available ny about 8, one at midday, in case a particularly heavy session of washing up or something had drained it, and another at around 6pm to make sure evening baths are supplied.
In practice often these just result in 5 minute 'burns' to restore the temperature, no hot water having actually been used at all. DHW is not - in our house anyway - the biggest use of energy by a very long chalk.
OK, I guess the algorithm in the CM67 is not likely to cope with serious inconsistencies.
Well, noise from the system is a different problem that the CM67 can't be expected to solve. A quieter pump perhaps? Perhaps a Grundfos alpha to adapt the flow to the specific need?
Fortunately the noise from my CH system is negligible (some water flow noise but not much), so I don't mind the CM67 doing its job by coming on at the right time.
The only problem I've found with this - the otherwise excellent Grundfos Alpha, that is - is that it fires up 'flat out' and modulates down to a nice quiet speed after about 30 seconds operation. The relatively loud initial flow is enough to wake me up! Couple this with my still unresolved problem of the Danfoss TP7000 misreading the room temperature and causing 1 minute firings, means it can be quite a noisy start to the day. Hopefully, having now given up on the Danfoss controller and trying to source some CM67s (CM907s) which allow a minimum on time should help (Before you ask .... the difficulty is that I need remote room sensors which Honeywell don't yet provide in the UK and no-one seems to have any CM67's left!)
Now alternatives would be :
1) some sort of relay that causes an 'on' to run for a minimum of say
4 minutes. Anyone know how such a thing can be achieved?
2) some means of starting the Grundfos Alpha at low speed and getting it to move its speed upwards....can this be done?
Regards ian Please reply to group - email address is not monitored Ian
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.