Evohome firmware 02.00.19.31 Beta Trial - Exclusive for Automated Home Members

Printable View

Show 100 post(s) from this thread on one page

7th April 2020, 09:09 AM
Andy the Minion

Quote:

Originally Posted by DBMandrake

I've had this happen to me on only one of my zones this morning during the optimal start period. (Kitchen showing 30C)

I'd like to point out though that this is not just a display error on the controller as you're suggesting here - I checked the HR92 and it was also showing a set point of 30C at the time, a few minutes before the end of optimal start, and a 100% heat demand despite the room temperature being 19.5C for what should be a 20C set point. (Normally heat demand would be a lot less for a 0.5C error)

So the set point of the HR92 is definitely going temporarily to the zone configured maximum during the optimal start period. As soon as the optimal start period was over the set point did go back to 20C, however the zone overshot a lot due to the HR92 going flat out set to 30C right until the end of the optimal start period instead of throttling back as it approaches the set point. That zone is prone to overshoots anyway but it overshot more than normal because of this temporary 30C set point.

@DBMandrake This is correct but the bug dosnt use to incorrect set temperature to calculate the heat-up time, so it wont have been given enough time to do anything too nasty to the room temperature. I am not exactly sure but it could be that reaching the correct scheduled setpoint may also reset the bug.
7th April 2020, 09:27 AM
SteveP

Quote:

Originally Posted by DBMandrake

Another thing I'm interested to see is whether the beta firmware fixes the issue where the controller lets the batterries run down and/or does not charge them completely when it is always left on the stand like mine is. This issue has been there since day one for me, (late 2015) and a few others have reported it too.

I actually replaced the rechargeable batteries in my controller about 6 weeks ago as they were over 4 years old and didn't seem to be holding a charge very well - at the time I used the "power on initially with batteries removed" trick to force the controller to charge them properly. (or so I thought)

Yesterday while trying to use the controller off it's stand for testing purposes I only got about 20 minutes of use before the red low battery warning appeared forcing me to put it back on the stand, (it should last at least 3 hours) so clearly the controller had let the batteries go flat again over the last 6 weeks or had never charged them properly in the first place. :(

I put it back on the stand now running the beta firmware and this time it charged the batteries for a full 9 hours! The replacement cells are a bit bigger in capacity than the originals but that is still far longer than I have ever seen it charge before. I was actually beginning to wonder if would ever stop.... So I wonder if the charging algorithm has been tweaked.

It seems kind of silly to have battery backup that should last for many hours that when the time of need comes runs out in 20 minutes because it has failed to keep the batteries charged properly. :)

Hm that's interesting. JUst taken my unit off the wall charger and yes for the first time ever on this evohome version (first gen didn't have this issue) the batteries are showing as full rather than half full. Has always been a bit of an irritation and Ive tried the various suggestions for fully charging the batteries to no avail :( Will leave it of the charger for a bit and see how long it lasts.
As has been said, would be nice to have some for of release doc rather than us "falling" over these potential (or coincidental?) changes !!
Ah even as I type this my batteries have now dropped to half :(
7th April 2020, 10:01 AM
DBMandrake

The beta firmware gave it a good long charging session but it will be a month or two before I see whether it keeps them topped up or not.

The issue I used to see on all previous firmwares is that if the controller was ALWAYS on the charging stand the controller would never charge the batteries to make up for the natural self discharge in the cells, which is quite high for NiMH. (To be fair, trickle charging NiMH is not easy to do as it's hard to detect when they're full at a very low charge rate as they don't show the normal dip in voltage as they near full charge that you get at a higher charge rate, so there is a risk of overcharging)

So eventually after maybe a couple of months the cells would be nearly flat and not run the unit for long despite it having been on the stand the whole time. If on the other hand it spent significant time off the stand to run the batteries down then the controller would register that they were down and charge them.
7th April 2020, 10:04 AM
roydonaldson

I just ordered a new set of 4 AA rechargeables to see if starting with some fresh ones will help.

Roy.
7th April 2020, 10:50 AM
DBMandrake

Quote:

Originally Posted by SteveP

Yep agree with all your observations and that there is a high dependency as to where the local weather station is plus each evohome owners physical location and surrounding. We are protected by hills from northerly winds but this morning we have southerlies so nicely warming up the house :) the issue really is that when you hit the boundary of the cold boost in terms of the actual outside temperature (14.5c) vs the evohome reported outside temp (9C) the difference is enough to be boosting the heating rather than setting it back. That will always be an issue with using internet weather data. Would be useful to know what weather data is being used by evohome as may help identify the accuracy? What may be an option that doesn't involve physically changing the evohome electronics would be to link a DT92 as an outside temp stat and use a zone to record the outside temp until a water proof outside temp stat is available? The DT92 needs to be mounted in a waterproof box but I would gladly give up a zone for better outside temperature compensation as my boiler is a trusty Valliant Combi that I have been told has many more years life and so I wont be able to justify to "the boss" changing the boiler for some time just so I can enjoy opentherm benefits :)

This morning the lag of the internet reported temperature is hillariously bad...

I have cold weather boost enabled on two zones (0.5C boost) to test it out, it's now 10:45am, my own weather station is currently reporting 12.5C outside, which should be warm enough to disable cold weather boost for the zones but the Evotouch is still reporting 5C as the outside temperature! :D (To be fair, WeatherPro on my phone which uses Meteogroup data still says 7C)

Granted, the temperature has risen very quickly this morning due to a clear sky overnight and this morning and sunshine, looking back at my weather station graph we passed 5C at about 8am this morning and passed 11C about an hour ago.

I wonder how useful these outdoor temperature related features really are if there is too much lag in the reported outdoor temperature coming from the weather service... ? A local sensor is definitely going to give better results.
7th April 2020, 11:53 AM
PJB

I've had the Beta firmware installed since 3Apr2020. Yesterday morning & evening and again this morning I noticed the 35degC CH demand in one of my zones. Other zones are unaffected. I have optimum start enabled. The zone in question has just an HR92UK rad controller. I'm afraid that I didn't investigate too much - I just manually decreased the zone setpoint via my HomeAssistant setup.

I put a call into Honeywell/Resideo support. They were aware of the firmware Beta trial. Their suggestion is

"At any hour and 5 minutes (f.ex. 12:05, 1:05) please remove the batteries from the HR92UK for couple of seconds and plug them back.
Make sure to follow that procedure exactly at some hour and 5 minutes, to be able to reset the unit properly."

Which sounds like hocus-pocus to me. However, I'll try it & let you know.
7th April 2020, 12:42 PM
DBMandrake

Quote:

Originally Posted by PJB

I've had the Beta firmware installed since 3Apr2020. Yesterday morning & evening and again this morning I noticed the 35degC CH demand in one of my zones. Other zones are unaffected. I have optimum start enabled. The zone in question has just an HR92UK rad controller. I'm afraid that I didn't investigate too much - I just manually decreased the zone setpoint via my HomeAssistant setup.

I put a call into Honeywell/Resideo support. They were aware of the firmware Beta trial. Their suggestion is

"At any hour and 5 minutes (f.ex. 12:05, 1:05) please remove the batteries from the HR92UK for couple of seconds and plug them back.
Make sure to follow that procedure exactly at some hour and 5 minutes, to be able to reset the unit properly."

Which sounds like hocus-pocus to me. However, I'll try it & let you know.

Believe it or not that isn't hocus pocus, but a workaround to a different issue (phantom override) that has been there for a long time. (Not related to the current beta firmware) Quite a bit of discussion about that particular issue on old threads here. It's an issue I've had myself and yes, randomising the boot up time of the HR92's does work around it. I don't think it will help this particular issue though.
7th April 2020, 01:42 PM
frankmalia

1 Attachment(s)

OK I am wondering why some of my rooms are boosting by 0.5C when temp difference to SP is only 9C. One room is not boosting when difference is 11C.

Attachment 1529
7th April 2020, 01:58 PM
DBMandrake

Quote:

Originally Posted by frankmalia

OK I am wondering why some of my rooms are boosting by 0.5C when temp difference to SP is only 9C. One room is not boosting when difference is 11C.

Attachment 1529

I've noticed some anomalies too, I think there may be a delay on re-evaluating whether cold weather boost applies as conditions change.

I had two rooms scheduled to 21C with cold weather boost enabled boosting them by 0.5C. Outside temperature was reported as 12C (had recently increased from a reported 9C) and despite the difference only being 9C cold weather boost was still applying.

However in the time it took me to start writing a post and go back to take a picture of the controller screen the cold weather boost had turned off for those two zones.

So either it's slow to react to changes or there is something else at play than just the temperature reading in the top left.

Also do we know whether the outside temperature is being compared to the normal setpoint of the zone or the boosted set point ?
7th April 2020, 02:04 PM
SteveP

Quote:

Originally Posted by DBMandrake

This morning the lag of the internet reported temperature is hillariously bad...

I have cold weather boost enabled on two zones (0.5C boost) to test it out, it's now 10:45am, my own weather station is currently reporting 12.5C outside, which should be warm enough to disable cold weather boost for the zones but the Evotouch is still reporting 5C as the outside temperature! :D (To be fair, WeatherPro on my phone which uses Meteogroup data still says 7C)

Granted, the temperature has risen very quickly this morning due to a clear sky overnight and this morning and sunshine, looking back at my weather station graph we passed 5C at about 8am this morning and passed 11C about an hour ago.

I wonder how useful these outdoor temperature related features really are if there is too much lag in the reported outdoor temperature coming from the weather service... ? A local sensor is definitely going to give better results.

I noticed this as my weather station was showing 10.5C and yet the evohome temp was showing 3C !! luckily (or more because I was aware of this from my experience last week) I had not set the Cold Weather Boost on and I was glad otherwise the house would have been in melt down. Do agree that unless we can get more reliable data for the outside temp then this feature will only be useful in the depths of winter :(

Show 100 post(s) from this thread on one page