Hot water temperature overshoots

[DBMandrake]

Hi,

I've had Evohome (self installed) for about a year running nicely and recently purchased the hot water kit to go along with a gravity hot water to S-Plan conversion I've just done.

Everything is working well except for one issue - in the 10 days that the hot water kit has been in place I've had large overshoots of the hot water temperature occur four times now.

I have the hot water set to 50 degrees, 5 degree differential and 0 minute overrun. 99% of the time the hot water heats to 50 degrees and shuts off as it should. But on four occasions now it has kept heating a further 20 minutes past the point where it should have shut off with the Evotouch still showing a <50 degree temperature. By the time a temperature update was displayed the hot water temperature is now around 62 degrees.

HW and CH zone valves are Honeywell V4043, and I am using a 3x BDR91 S-Plan configuration with separate boiler relay. There is no problem with the zone valve or relays controlling them, the issue seems to be a lack of temperature updates from the CS92 to the Evotouch.

I also graph my temperatures with Domoticz with 5 minute granularity and the following is typical of what happens when a reheat overshoots:

Code: Select all

DateTime	Set Point	DateTime	Temperature
02/11/2016 11:45	60	02/11/2016 11:45	45.53
02/11/2016 11:50	60	02/11/2016 11:50	45.53
02/11/2016 11:55	60	02/11/2016 11:55	45.53
02/11/2016 12:00	60	02/11/2016 12:00	44.44
02/11/2016 12:05	60	02/11/2016 12:05	49.45
02/11/2016 12:10	60	02/11/2016 12:10	49.45
02/11/2016 12:15	60	02/11/2016 12:15	49.45
02/11/2016 12:20	60	02/11/2016 12:20	49.45
02/11/2016 12:25	60	02/11/2016 12:25	62.67
02/11/2016 12:30	60	02/11/2016 12:30	62.67
02/11/2016 12:35	60	02/11/2016 12:35	62.67
02/11/2016 12:40	60	02/11/2016 12:40	62.67

(Ignore the 60 degree set point - Domoticz doesn't know that the set point is 50 degrees and just fills in 60 as a place holder value)

At 12:00 the temperature dropped below 45 so hot water reheat started. At 12:05 it had reached 49.45 already so should have reached 50 just a minute or two later. However the reported temperature remains stuck at the same 49.45 for a full 20 minutes. By the time a new reading is received at 12:25 the temperature has overshot by nearly 13 degrees. Here is a previous example:

Code: Select all

DateTime	Set Point	DateTime	Temperature
31/10/2016 04:20	0	31/10/2016 04:20	39.48
31/10/2016 04:25	0	31/10/2016 04:25	39.48
31/10/2016 04:30	0	31/10/2016 04:30	39.48
31/10/2016 04:35	60	31/10/2016 04:35	39.48
31/10/2016 04:40	60	31/10/2016 04:40	45.34
31/10/2016 04:45	60	31/10/2016 04:45	48.83
31/10/2016 04:50	60	31/10/2016 04:50	48.83
31/10/2016 04:55	60	31/10/2016 04:55	48.83
31/10/2016 05:00	60	31/10/2016 05:00	48.83
31/10/2016 05:05	60	31/10/2016 05:05	48.83
31/10/2016 05:10	60	31/10/2016 05:10	62.29
31/10/2016 05:15	60	31/10/2016 05:15	62.29
31/10/2016 05:20	60	31/10/2016 05:20	62.29
31/10/2016 05:25	60	31/10/2016 05:25	62.29
31/10/2016 05:30	60	31/10/2016 05:30	62.29

On this occasion the re-heat was triggered by the hot water schedule switching to on at 4:30 (set point changed from 0 to "60" in domoticz, in reality 50 degrees) and the situation is similar - the reading rises quickly to 48.83 then there are no further updates to the reading for 20 minutes followed by a reading of 62.29, about 12 degrees over the set point.

Here is an example where everything is working normally, which is the vast majority of the time:

Code: Select all

02/11/2016 04:10	0	02/11/2016 04:10	41.68
02/11/2016 04:15	0	02/11/2016 04:15	41.68
02/11/2016 04:20	0	02/11/2016 04:20	41.68
02/11/2016 04:25	0	02/11/2016 04:25	41.68
02/11/2016 04:30	0	02/11/2016 04:30	41.68
02/11/2016 04:35	60	02/11/2016 04:35	41.68
02/11/2016 04:40	60	02/11/2016 04:40	46.14
02/11/2016 04:45	60	02/11/2016 04:45	51.53
02/11/2016 04:50	60	02/11/2016 04:50	51.53
02/11/2016 04:55	60	02/11/2016 04:55	51.53
02/11/2016 05:00	60	02/11/2016 05:00	51.53
02/11/2016 05:05	60	02/11/2016 05:05	51.53

Hot water was scheduled on at 4:30 and was reported as being up to temperature at 4:45 with the temperature stopping at 51.53.

As far as I know I have met the RF requirements for positioning etc as well as I possibly could - the boiler is inside a closet built in the porch beside the kitchen, along with a small hot water cylinder so the choices for mounting locations for BDR91's and CS92 are fairly restricted. I ended up with them vertically spaced from relatively low down to high up with the CS92 at the top, and it is about 300mm from the cylinder and nearest BDR91, no further improvement in positioning is physically possible.

Line of site from the CS92 is through a wall which is low density fire brick with plaster on the other side, then through a wooden framed glass door to the Evotouch on a wall mount in the hallway total distance no more than 7 metres. All internal walls of the house aside from the boiler room/kitchen partition are brick and plaster. All devices around the house including CS92 give a full 5/Excellent signal test at all times.

After a lot of testing and some discussion over at the automated home forum my feeling is this isn't a signal strength or interference issue, but a problem with the way the CS92 decides when to send temperature updates, (which are deliberately minimised for battery life) and that for some reason it sometimes doesn't send a timely temperature update just after the set point threshold is crossed.

I would appreciate any assistance in figuring out why this is happening as at the moment I'm contemplating having to wire up a non-wireless safety thermostat to prevent such large overshoots. While the cylinder is protected as it is a vented cylinder and can only heat to the flow temperature of the boiler anyway my concern is over scalding due to random unexpected much higher than normal hot water temperatures catching people unawares.

[Richard]

What are you using? The strap on sensor or the insertion probe? These are both 10K sensors and the CS92A is reading is in ohms. Can I have a picture of the installation to give the best of my advice?

I have moved this post to the customer only area, as some of the input will obviously be personal and some of the advice I will give will be for my customers only.

Thanks,

Richard

[DBMandrake]

Hi Richard,

I'm using the strap on sensor on a vented cylinder.

Here are a couple of pictures I took after finishing the S-Plan conversion before the covers went back on the boiler:





Some clarification of things that may not be clear from the pictures:

The boiler is in a room off the back porch, the kitchen is immediately through the wall the wiring boxes are mounted on, and the direction of the Evotouch is diagonally away from the camera towards the left. On the other side of the right hand half of the wall from about the cylinder centre line onwards is a fridge - this is why I have not mounted any wireless boxes on the right hand side even though the picture might suggest that to be a better location at the top. Where they are to the left they have at least 30cm clearance from the corner of the fridge and the signal does not need to travel in the direction of the fridge, its about 90 degrees away from the direction of the fridge.

The boiler relay which I had before the Hot Water kit is mounted on the other side of the wall in the kitchen where the old timer used to be - I decided to leave it there when I added the two extra BDR91's so I could still see boiler activation status. You can see the cable for it disappearing through the wall near the top left of the first picture.

I've spaced the BDR91's vertically as well as I can and allowed for spacing from the one on the other side of the wall. The picture may make it appear that the hot water BDR91 and CS92 are close to a copper pipe but that is the hot water header tank feed and it is well away from the back wall by at least 30cm. The CS92 is right at the top and is probably the best situated of all the boxes as far as being away from sources of interference or metal objects is concerned. As far as I can tell I am not having any problems with BDR91 comms - I have never seen any comms errors to them, nor have I observed any problems during testing, it's only the CS92 that seems to be giving me trouble.

As you can see I have a very tight space to work in without much choice of where to put things. On the left hand wall there are a lot of copper pipes that make mounting there infeasible.

The rectangular box at the bottom is a standard drayton wiring centre and beside it is an MRT16-REM timer used as a pump overrun timer - those and the BDR91 on the other side of the wall have been there for over a year. New with the hot water kit are the two round terminal boxes for the zone valves and of course the BDR91's and CS92.

Edit: Forgot to mention that the power adaptor brick for the Evotouch wall mount is actually inside the left hand side of the Drayton wiring centre, although it's probably not relevant.

Hopefully that gives you somewhere to start. By the way I saw an overshoot again this morning with hot water scheduled to come on at 4:30:

Code: Select all

03/11/2016 04:10	0	03/11/2016 04:10	35.61
03/11/2016 04:15	0	03/11/2016 04:15	35.61
03/11/2016 04:20	0	03/11/2016 04:20	35.61
03/11/2016 04:25	0	03/11/2016 04:25	35.61
03/11/2016 04:30	0	03/11/2016 04:30	35.61
03/11/2016 04:35	60	03/11/2016 04:35	35.61
03/11/2016 04:40	60	03/11/2016 04:40	35.61
03/11/2016 04:45	60	03/11/2016 04:45	48.22
03/11/2016 04:50	60	03/11/2016 04:50	49.45
03/11/2016 04:55	60	03/11/2016 04:55	49.45
03/11/2016 05:00	60	03/11/2016 05:00	49.45
03/11/2016 05:05	60	03/11/2016 05:05	49.45
03/11/2016 05:10	60	03/11/2016 05:10	49.45
03/11/2016 05:15	60	03/11/2016 05:15	62.67
03/11/2016 05:20	60	03/11/2016 05:20	62.67
03/11/2016 05:25	60	03/11/2016 05:25	62.67
03/11/2016 05:30	60	03/11/2016 05:30	62.67
03/11/2016 05:35	60	03/11/2016 05:35	62.67
03/11/2016 05:40	60	03/11/2016 05:40	62.67
03/11/2016 05:45	60	03/11/2016 05:45	62.67

Again I see the reading getting stuck at around 49 for over 20 minutes while the cylinder is still heating. When an updated temperature reading does come through it has reached 62 degrees and then shuts off.

Regards,
Simon

[Richard]

Have you tried the insertion sensor Simon? You can wedge this against the tank behind the strap on sensor to test it.

The reason I ask is that we have had a couple of faulty strap on sensors returned to us. The ohm values of the 10K sensors in relation to the temperature are here (page 3 - https://products.ecc.emea.honeywell.com ... 1r0215.pdf) and at 25 degrees C it should be reading 10000 ohms.

Is there not a perceived issue with the RF hitting the metallic items and bouncing off? I personally refrain from going anywhere near metallic objects and other sources of electrical interference with the RF devices, as my experience has shown they just don't work.

There is also a little trick I do with the CS92A, which involves taking the PCB out and bending the battery terminals up slightly to ensure full contact with the PCB. This has worked on numerous occasions.

I hope some of that helps?

Thanks,

Richard

[DBMandrake]

Have you tried the insertion sensor Simon? You can wedge this against the tank behind the strap on sensor to test it.

The reason I ask is that we have had a couple of faulty strap on sensors returned to us. The ohm values of the 10K sensors in relation to the temperature are here (page 3 - https://products.ecc.emea.honeywell.com ... 1r0215.pdf) and at 25 degrees C it should be reading 10000 ohms.

Thanks, I'll keep that in mind and give it a try if other ideas don't help. If the sensor is disconnected (due to an intermittent open circuit) would that be registered as a fault on the Evotouch, or would it just maintain the last known reading for a while ? If the latter I could potentially see that as the sensor heats up and expands a faulty connection inside the sensor could make and break the connection, causing a temporary loss of temperature updates.

Is there not a perceived issue with the RF hitting the metallic items and bouncing off? I personally refrain from going anywhere near metallic objects and other sources of electrical interference with the RF devices, as my experience has shown they just don't work.

It's not ideal no but as long as the metal is 30cm away and not in a line of sight of the signal path it shouldn't really cause an issue unless the signal level was marginal to begin with, which isn't the case in my situation. Unfortunately my options are fairly limited in placement.

There is also a little trick I do with the CS92A, which involves taking the PCB out and bending the battery terminals up slightly to ensure full contact with the PCB. This has worked on numerous occasions.

Interesting suggestion - I've taken it apart and increased the tension on the battery contacts on the rear of the board as you suggest, I'll see how this goes first before making any other changes. It might take a few days to know if it has helped as it may take a few days for the problem to exhibit again. Of course in the process of adjusting the contacts I was forced to remove the batteries, so if it does help I won't know for sure if it was the "reboot" of the device or the contact tension. (Although if the problem is solved I won't care!)

I did a quick test afterwards where I ran the hot water long enough to trigger a reheat then watched the temperature reading on the evotouch - this time it increased from 44 to 50 degrees in one degree steps (and then shut off when it should) whereas previously I'm sure I only got 2 or 3 different readings between 44 and 50, but only time will tell.

By the way I don't remember if it was the CS92, the BDR91 that came with it or the separate BDR91 I ordered but on one of them the circuit board was loose inside the case and wasn't clipped into its mounting clips at all when I received it - I had to carefully prise and click the board back into place during initial installation.

[Richard]

How is this going Simon, any joy?

[DBMandrake]

How is this going Simon, any joy?

No overshoots yet since I adjusted the battery contacts behind the board, so far it has done 12 reheats without exceeding 51 degres but its a bit too early to tell as the problem wasn't happening every day. Here's a graph of the last week or so showing four previous overshoots circled in red. As you can see in the last 24 hours it has stuck very close to the set point: (pink line)



What I also noticed is the sensor does seem to be reporting in more often when a reheat is occurring than before. For example during one test I ran the hot tap until it dropped to 44 degrees and fired the boiler, writing down the time every time the temperature reading changed, then did the same during a reheat and got the following results:

Code: Select all

Temp - minutes:seconds
51 - 0:00
49 - 1:13
47 - 1:46
45 - 2:19
44 - 2:28 (hot water relay turned on)

44 - 0:00
45 - 1:57
46 - 2:30
47 - 3:02
49 - 3:25
50 - 3:52 (hot water relay turned off)
51 - 5:37

Previously it would only update once or twice during a reheat for example it might go from 38 to 44 to 49 and then not update again until over 60 - now it seems to update almost every 1-2 degrees with quite short intervals in between.

I'll keep an eye on it over the next week before concluding whether it's fixed or not.

Thanks for the help so far!

[Richard]

Excellent stuff, fingers crossed then. Certainly performing now how I would expect it to!

[DBMandrake]

Hi Richard,

I was going to wait a full week but it's been 5 days now without any problems so I think the battery contact tension behind the PCB was indeed the issue, glad it could be fixed so easily.

Here is a graph of the last 5 days and as you can see the temperature only just peaks slightly above the 50 degree set point (pink line) and stops consistently between 50-52 degrees:



Nothing else has been touched on the system besides adjusting the battery contact, even the flow temperatures have been left alone.

If the problem does for some reason return I'll let you know but I'm happy at the moment that it seems to be solved!

Regards,
Simon

[DBMandrake]

Hi Richard,

You asked elsewhere if I had seen any comms errors from the hot water sensor - at the time I had not however yesterday morning I received a comms error for the hot water sensor for the first time.

The hot water was scheduled to come on at 7:30, by chance I happened to be up and in the vicinity as this was happening, the hot water started heating as normal, but at 7:40 I walked past the controller again to be greeted with a big red comms error on the screen. There were double dashes through the temperature reading and the hot water relay had been shut off. The fault logbook reported the fault occurred at 7:37 - just a few minutes after the hot water was scheduled on.

I immediately went and pressed the button on the CS92 - 5 flashes as per normal, and also checked from the controller end - "excellent" signal, and yet it had lost communications... I took the battery out of the CS92 and put it straight back in - a temperature reading appeared on the controller immediately and hot water heating re-started, with no further problems for the day.

This morning I had a 12 degree overshoot again then for the rest of the day it has been OK. I decided to try moving the CS92 to a different wall in the boiler closet - this wall is at right angles to the one it was on and puts the CS92 quite a bit further away from the BDR91's as well as a bit further from the hot water cylinder. It is however a worse position as far as line of sight to the controller is concerned, and the signal test has indeed dropped to 4 flashes.

However it's worth a try. If this doesn't work I don't really have any other locations the sensor could be installed and at that point I would be out of ideas or perhaps suspecting a faulty unit.