Room air temperature/quality sensors

[dnb]

I am starting to look at ways to monitor air temperature and quality within rooms and can't find exactly what I am looking for.

 

My ideal specification goes something like this:

* Cheap - using simple commonly available parts, these thing shouldn't cost much!

* POE supply for ease of installation in a new build. (A battery and wireless version would be OK too, but batteries are usually a pest in these things)

* Simple TCP/IP or UDP comms with sensible (and preferably open) message format. (The sort of thing you can talk to with 20 minutes of playing with Perl or Python)

* Measures temperature, CO2 level, humidity and possibly light intensity as a minimum set. Pollutants would be nice.

* Discreet size and shape

 

My idea is to run a set of these and other house sensors on on a dedicated and air gapped network to collect data to assess how well the A/C and MVHR is functioning, monitor DHW tank capacity (naturally with other sensors!) etc in a common building management system. I am looking at various of the open schemes for this at the moment.

 

I can find plenty of home brew examples of nearly what I am after, but there must be something I can just go and buy! Any thoughts? Is there a demand other than me for such a thing in the first place?

[Jeremy Harris]

I bought a box of NDIR CO₂ sensors from the US and use these for measuring CO₂ concentration around the house, together with DS18B20 1 wire temperature sensors and HH10D RH sensors.  Each measurement station has a small microcontroller that takes the raw data from the sensors and sends it via a serial wired connection back to a central controller that both logs the data to USB, displays it locally (with GPS derived date and time, automatically corrected for BST) and sends it to a display on the wall in the hall.  The whole system runs from a single plug-in power supply, with internal back up batteries that can run it for a day or so if the power goes down, as the whole system only needs a few tens of mA to run.

 

I also bought an air particulate sensor a while ago, and keep meaning to incorporate that into an outdoor unit.  This can bin particulates into three size categories, and sends the data out via a low voltage serial port.

 

Some of the temperature measurements are probably the most useful, I've found.  I measure floor temperature (sensor embedded in the slab), ASHP flow temperature, ASHP return temperature, buffer tank temperature, room temperature, outside air temperature and hot water flow temperature.  Out of those, outside air temperature is the one we tend to check the most, as is the GPS-derived date and time (really just to check other clocks in the house).  It's useful to know how cold it is outside, so as to know whether or not to remotely precondition the car if I'm going out.  After being fascinated by all the other data initially, the novelty wore off a while ago, once I'd got a feel for the way that the house responds to changes.  I don't think I've bothered to look at the logged data for a long time now, as it will only show stuff I've already found out.

[joe90]

I too wanted to monitor CO2 levels to control the MVHR (a bit like central heating running off a room stat) no point in running it if it’s not needed (windows opened!!!). Quick search found this. https://www.flamefast-gas-safety.co.uk/products/co2-air-quality-monitoring-equipment.html?gclid=Cj0KCQjw753rBRCVARIsANe3o47B7IrhbuWZ3Ab3ofsokp1-YwCXh9S5u39Sm5CWx95IxOWM3bDMwcQaAv9ZEALw_wcB.   I too would need plug and play as I don’t understand tech ?

[SteamyTea]

Is it really worth monitoring CO2? There is such a good correlation between RH and CO2 that it is not really worth the expense and extra complication.

It is fairly easy to fit some DHT22s into MVHR ducts, log data and do control from that. A few of use do it already.

Like @JSHarris I have collected data for a few years, but the novelty soon wears off.

There are things that I would like to automate, basically how long my night storage heaters are on for. But that is more for a technical challenges than for any real and meaningful savings.

That trouble with buying stuff that is ready to use. Is that it is expensive. You are paying people to design and test stuff, and offer a warrantee. Then it probably only does half of what you want, and in a format that does not work well with other kit.

Take just controlling a motor, it can be done with a voltage or frequency input, any control system bought in needs to be able to handle both of these.

10 quid raspberry pi and a few sensors, a couple if days coding and you are away.

[joe90]

13 minutes ago, SteamyTea said:

10 quid raspberry pi and a few sensors, a couple if days coding and you are away.

 

Thats ok if you know what your doin!, I’ll stick to driving my tractor ?

[Jeremy Harris]

23 minutes ago, SteamyTea said:

Is it really worth monitoring CO2? There is such a good correlation between RH and CO2 that it is not really worth the expense and extra complication.

 

 

Probably not, but I felt initially that it would be useful to know how much CO₂ there was in the house.   There's also an issue with the correlation between internal RH and CO₂ not holding up well when the outside RH changes.  For example, right now the indoor CO₂ concentration is pretty normal, 517ppm, but the  indoor RH is quite high, 51%.  The reason for the high indoor RH may well be that last night was cool outside (about 11°C, it's currently 13.8°C and rising) and it rained heavily yesterday evening.

 

Now there seems little merit, other than curiosity, in continuously monitoring CO₂ concentration, as it tends to stay within a fairly tight range, from around 420ppm (if the house has been empty for some time) to maybe 700ppm or thereabouts if we've had visitors.  Most of the time it sits around where it is at the moment, between 500 and 600ppm.

[joe90]

The relative humidity in my new build has always been a bit high, but I am in damp Devon!. The MVHR is off as er indoors always opens windows so no point in it being on. Currently it is 67%. It is a cheap Chinese gadget tho! How accurate are these things? Can someone recommend an inexpensive one that is fairly accurate?

[Jeremy Harris]

20 hours ago, joe90 said:

The relative humidity in my new build has always been a bit high, but I am in damp Devon!. The MVHR is off as er indoors always opens windows so no point in it being on. Currently it is 67%. It is a cheap Chinese gadget tho! How accurate are these things? Can someone recommend an inexpensive one that is fairly accurate?

 

I calibrated the sensors I bought, using a saturated salt slurry in a sealed plastic food container.  You could probably check a unit in much the same way.  I used a couple of different compounds, so that I had two calibration points.  One check point would probably be OK as a test for accuracy, though.

 

I found that the pre-calibrated HH10C sensors I bought were pretty accurate, better than 3%, which is pretty good for an RH sensor.  The DHT11 sensors I bought were way out, though, so much so that I didn't bother to use them.

 

To calibrate a sensor at a single point, just get some sodium chloride (salt) and mix it with a small amount of water to form a fairly thick slurry or paste.  Put this in a sealed container along with the sensor, and place it somewhere where the temperature will be relatively stable for a few hours.  After about 10 to 12 hours the RH will have stabilised inside the container to about 75%, so that's what the sensor should indicate.  If you want to check it at the lower end of the scale, then instead of sodium chloride you can use magnesium chloride, using the same method.  This should give an RH of 32.8%.

 

When I checked the calibration on the RH sensors I use, I used this method, but with magnesium chloride for the lower calibration point and potassium chloride for the upper calibration point (this gives 84.34% RH). 

 

 

[Ed Davies]

2 hours ago, SteamyTea said:

Is it really worth monitoring CO2? There is such a good correlation between RH and CO2 that it is not really worth the expense and extra complication.

 

To back up @JSHarris's response, here are my bedroom RH and CO₂ for the last 24 hours (since 12:00Z yesterday) and since the beginning of the month. As you can see, in the short term the CO₂ and RH track pretty well [¹] but longer term  the change in outside conditions messes this up.

 

[¹] But not brilliantly, see the largish rise in humidity around 02:00Z compared with the steady fall in CO₂ over that time.

[SteamyTea]

Outside temperature and humidity will make a difference, but should be able to compensate for that easily enough I would have thought.

Those DHT11s are rubbish, even at a quid each.

I still think that CO2 levels are not really worth monitoring, you could just ventilate a bedroom, or living room, a bit more.  I suspect that we react to humidity and temperature faster than we do CO2 levels (any know how to put a subscript in on an Android phone).

As an example, yesterday was 15°C down here. When it was raining, I was shivering, sun out, I felt fine.

If CO2 levels are consistently high, there may be other reasons for that, what I am not sure though.

Edited August 29, 2019 by SteamyTea

[SteamyTea]

1 hour ago, Ed Davies said:

As you can see, in the short term the CO₂ and RH track pretty well [¹] but longer term  the change in outside conditions messes this up.

Can you do some plots of RH against CO2, maybe of short, medium and long-term time scales.

[joth]

On 29/08/2019 at 09:33, SteamyTea said:

It is fairly easy to fit some DHT22s into MVHR ducts, log data and do control from that

Just a note that controlling MVHR via sensors of any type in the ducts seems fraught, as once you've reached setpoint and turn the MVHR off, there's basically no air flow so it'd be extremely slow to respond to subsequent drop in quality in any of the rooms. 

[joe90]

If I do use a RH sensor to control MVHR on/off I would put it on the landing, a central point within the house rather than in the duct work. I do plan to have one in the bathroom duct to control boost after shower/bath, this may have to also over ride landing sensor if the MVHR is off.

Edited August 30, 2019 by joe90

[MikeGrahamT21]

The MVHR should never be switched off surely? the sensors generally activate a boost condition to remedy the issue with air quality, and then when its resolved, it goes back to normal speed.

[Jeremy Harris]

40 minutes ago, joth said:

Just a note that controlling MVHR via sensors of any type in the ducts seems fraught, as once you've reached setpoint and turn the MVHR off, there's basically no air flow so it'd be extremely slow to respond to subsequent drop in quality in any of the rooms. 

 

 

Our MVHR is controlled via a humidity sensor in the extract duct.  Works very well, as it switches the MVHR to boost when the humidity rises in that duct, so when the shower's running, or we're cooking.

[redtop]

6 minutes ago, JSHarris said:

 

 

Our MVHR is controlled via a humidity sensor in the extract duct.  Works very well, as it switches the MVHR to boost when the humidity rises in that duct, so when the shower's running, or we're cooking.

 that sounds like a good use of tech.  I only want to add things that end up being useful (i.e. saving effort / money) over the longer term.  Perhaps we need a thread of what tech has proved useful and what hasn't (there probably already is one lol).  As our ground floor is raised off the ground on stilts i was thinking about temp sensors low down in the ground floor make-up so i can detect fire very early

[joth]

16 minutes ago, JSHarris said:

 

 

Our MVHR is controlled via a humidity sensor in the extract duct.  Works very well, as it switches the MVHR to boost when the humidity rises in that duct, so when the shower's running, or we're cooking.

Sure, I can understand a duct sensor working OK to control boost mode, but not to turn the MVHR off and on completely which I thought was being discussed in the prior posts.

 

[Jeremy Harris]

For those interested, the humidistat I used was the one in this post:

 

[dpmiller]

built-in humidity sensor in our Vent Axia works remarkably well

[dnb]

Thanks for the insights so far. It's much appreciated.

A friend told me about Loxone light switches earlier in the week. They seem to do most of what I need, albeit at a price. Any thoughts about Loxone as a home automation system?

It seems I can use it to do 80 to 90% of what I want to do:

3 zone controls via air dampers on the A/C ducting

Temperature and humidity monitoring via light switches in all key rooms that can be used as control inputs

Electricity use monitoring and thus immersion heater control to use "spare" PV

Music server etc - although it seems they are selling an expensive rack PC with software to do this... (could I just buy the software? I have lots of hardware... ;) )

 

I wonder how much the system can be customised to cover the last 10 to 20%.  Is there practical experience here?

 

[joth]

@jack self installed a Loxone system and has good experience with it by all accounts. You can search the forum for his various helpful insights into it, but may have more thoughts to add specifically on heating & climate control.

 

It's also my currently preferred option, for an install not totally dissimilar to yours. We like the aesthetics of their switches over any other smart switch options and the built in temperature and humidity sensors are a bonus. To manage costs I'll only use them in living and bedrooms, at least initially, and put cheap retractive switches everywhere else. Radially wiring means I can upgrade then in future if needed.

I'll use it for lighting, heating, cooling and shading control. I looked into their music, power monitoring add weather forecasting, but think I might control those in home assistant instead. These are not critical services, I want to keep Loxone for the critical BMS only, and remain understable e.g. to a new owner if we eventually sold the house with it still in (reversionary rights, as I think you put it?)

Their music server is so lo-fi it beggers belief the price tag they put on it. This is symptomatic of the main reservation I have with the to be honest: in recent years their position has moved towards locked in maintenance through certified installers and new products have spread them over more and more market segments with more lock in through proprietary protocols and integrations rather than focus on the core competency of making their controller work perflectly with the market leaders for each other device type (Sonos streaming, Philips smart bulbs, KNX or WiFi weather stations, etc..)

At first I genuinely thought their smarthome connected BBQ thermometer press release was an April fools joke.

 

Hence I'm going into this cognisant  that they may not be around forever and there's a chance it all might have to be replaced at some future point, again radially wiring the lighting is key to my comfort in selecting them. If I wasn't using them for lighting comtrol, the dependency would be constrained to the plant room.

[jack]

9 hours ago, dnb said:

Thanks for the insights so far. It's much appreciated.

A friend told me about Loxone light switches earlier in the week. They seem to do most of what I need, albeit at a price. Any thoughts about Loxone as a home automation system?

It seems I can use it to do 80 to 90% of what I want to do:

3 zone controls via air dampers on the A/C ducting

Temperature and humidity monitoring via light switches in all key rooms that can be used as control inputs

Electricity use monitoring and thus immersion heater control to use "spare" PV

Music server etc - although it seems they are selling an expensive rack PC with software to do this... (could I just buy the software? I have lots of hardware... )

 

I wonder how much the system can be customised to cover the last 10 to 20%.  Is there practical experience here?

 

I have a Loxone system running lighting, exterior blinds, heating, alarm, and some temperature sensing and display. I've wired for (but don't yet use) garage door and electric gate control. There might be a couple of things I'm forgetting.

 

I don't have the music system set up. The Loxone branded amp is just a rebadged Dayton (I think it is) and seems a wildly expensive way of doing things. By all accounts there are cheaper ways of integrating sound into the system, although even then I'd only integrate if you wanted to actually do things using the sound system through Loxone.

 

Their proprietary touch switches are good, by all accounts, but a very expensive way of implementing physical controls. We just have retractive switches for lighting and blinds, and that's fine. Like @joth, I'd consider them for areas where you want multiple inputs in a neat format (eg, in the kitchen, where you might want to control lights and music without using the app).

 

If you stick to radial wiring of lighting (in particular), you shouldn't have any problems with moving to another system in the future. You do need to think carefully about dimming - there are many ways of doing this, each with their own advantages and disadvantages.

 

Their software is probably the best thing about what they do. It's insanely powerful. Basically, you can take just about anything as an input, do just about anything with it (eg, apply logic and/or maths functions), and then control just about anything with the resultant output. What sort of "last 10 to 20%" were you thinking about @dnb? If you can explain the sorts of things you have in mind I'll let you know whether I think they can be done (bearing in mind I'm not an expert, but I did do all the programming for my system).

 

As @joth says, Loxone is increasingly moving to closed systems. They used to be very self-installer friendly, but I understand that's less and less the case. That said, there are mailing lists and forums with some very helpful people on them. if you have any sort of technical bent at all, you shouldn't have any problems.

[dnb]

Sorry for not replying sooner. It's hard to get around to everything at the moment!

 

My plan was to wire for many things (blinds, garage door, window indications etc) with the view that I would have to cut funding for the expensive parts of these and implement them later. So the first items are lighting and heating/cooling control.

 

I was considering DMX for lighting on the grounds it's an industry standard and will be around for a while yet. I think I need 23 lighting zones (approx 11 dimming and the rest not) so it will be expensive to use Loxone, it is a potential obsolescence issue

 

Having read the Loxone website and watched the videos I think it can do just about everything I want (at a price!) The first 80% of desires were things like dimmed lights in the hallways in a night mode, one touch control to turn on lights from each bedroom to the kitchen. I was also looking at controlling the DHW immersion heater by reading the electricity meter, perhaps via modbus comms. The last 20% were about fine control over the heating & cooling. Such as using a weighted average temperature for each floor (HVAC is to be zoned by floor) with the weights based on likely location of people and time of day etc to control the air dampers. Their heating control algorithms may of course already do this without me having to interfere, but the documentation is not exactly thorough!