SonOff Low Cost Wifi Switches for Home Automation

Johnboy1951 wrote: »

Thanks that esphome is on my 'to read' list ..... was reading a bit about it yesterday.
Do you use it?

Not as of yet I do not, and the main reason for this was I had all of my established devices set up for HA autodiscovery via MQTT. But I did run in to an annoying issue with MQTT of late, and though about swapping over.

It may happen in the future.

Johnboy1951 wrote: »

I was looking at the 4 ch units rather than the Pro version.
Is there a reason I should consider the Pro?

Inching maybe ? I dont recall the main differences off hand .

Johnboy1951 wrote: »

The UFH will be done mainly on time .... mine is the slow response and lasting heat type, using about 4" thick concrete around the pipes.
Other parameters to the controller should be easy to add after it is all functional .... such as outside temp and wind which would impact on the duration required to heat the floors.

Now from experience I can tell you that heating based on time alone especially UFH may not give you the best return.

The overarching premise of UFH is, heat the floor, when the in room thermo says it's X degrees close the actuator.

If you open the actuator based on time frame and the room is say 18 degrees, run it for 4 hours you could upend with a room thats 22 degrees, couple that with solar gain ( even in winter) and you may have a sticky situation.

Johnboy1951 wrote: »

The other thing I wish to have is security camera control ..... maybe 3 cameras in use. I know I want to have these, but have not quite decided how best to make them easy manage and access etc etc.
It is for this use that I might need to buy a phone or tablet or such, and maybe incorporate a warning bell somewhere.
Living in a dead zone for mobile has meant that we never bothered with mobile phones much. There is one that is almost never used but taken out of house on trips ...... but it is not a smart phone, just an emergency calling device.

I have a similar concept in mind here, but may complete some of it with one wireless camera..
POE camera's Hikvision, backed with Blueiris should provide what you require, and now HA has streaming capabilities and Blue iris has alert zone ability.
Maybe include a google home or an alexa for the door bell( alert) portion.

Johnboy1951 wrote: »

All this is a long term project over the next 6 or 8 months I hope.
I intend to get the hardware and maybe do a temporary software set up, so I become familiar with things before I deploy.
Of course that time scale is depending on outside factors .... such as builders/renovators appearing and working to proposed schedule and price.
I live in hope!

6-8 months is a good time frame here, Maybe start with a Pi3b+ and see how you get on with HA, one word of warning as i've recently discovered, SDcards suffer in pi's and i've had one fail on me already, 6 months in to my deployment. always backup, and always keep a spare sd card

Dr4gul4 wrote: »

Not as of yet I do not, and the main reason for this was I had all of my established devices set up for HA autodiscovery via MQTT. But I did run in to an annoying issue with MQTT of late, and though about swapping over.

It may happen in the future.

Maybe you could elaborate on that difficulty you encountered?
MQTT is one of the options I was contemplating .... at least for some functions ... with esphome used for others

Inching maybe ? I dont recall the main differences off hand .

OK, thanks. Not sure what I would use it for in my proposed set up.

Now from experience I can tell you that heating based on time alone especially UFH may not give you the best return.

The overarching premise of UFH is, heat the floor, when the in room thermo says it's X degrees close the actuator.

If you open the actuator based on time frame and the room is say 18 degrees, run it for 4 hours you could upend with a room thats 22 degrees, couple that with solar gain ( even in winter) and you may have a sticky situation.

Several things affect this way of looking at it.
The temperature of the return flow of each loop (or group of loops) is easily monitored. As the return temp approaches a desired temp for the floor the actuator can be closed.
Desired temperature is dependent on room use and individual preferences so some time is spent achieving the best compromise, after which it is set.

Monitoring outside temp and wind direction and speed can be used to tweak settings.
I find that high wind will tend to strip heat from a room more quickly than extremely low outside temperature and no wind.

Yes solar heating into the floor during Winter is not something that can be predicted. But that applies regardless the setup.

I have a similar concept in mind here, but may complete some of it with one wireless camera..
POE camera's Hikvision, backed with Blueiris should provide what you require, and now HA has streaming capabilities and Blue iris has alert zone ability.
Maybe include a google home or an alexa for the door bell( alert) portion.

I have been testing one camera (inside window looking out) with Xeoma and have been quite pleased with the results.
No doubt it needs a little more control but the movement detection is good and reliable.
I also need to test out Shinobi when I get around to it, as I think this might suit me.

I have (I think) decided against POE ... will do a one time wiring job for the cameras and then be able to use whatever device I wish to attach it to. The wiring will be terminated in the media press from where I can distribute power as well as the ethernet cable.

6-8 months is a good time frame here, Maybe start with a Pi3b+ and see how you get on with HA, one word of warning as i've recently discovered, SDcards suffer in pi's and i've had one fail on me already, 6 months in to my deployment. always backup, and always keep a spare sd card

I think I might be being a bit optimistic with my time-frame I am not known for getting things done on time!
I definitely won't buy a R-Pi3 for the purpose. I do not consider them great value TBH when all necessary parts are added.

I had intended to use my i5 NUC for the purpose, but unfortunately recent lightening strike put paid to that idea!

I would be more comfortable with a SFF used Intel x86_64 PC, or I might consider this
https://www.hardkernel.com/blog-2/odroid-n2/
In the meantime I have an unused PC here with an i3 .... big and bulky but capable.
I also have an old single core Celeron laptop but doubt that would suffice in the long term.

As is obvious I have not made any final decision on that yet ..... am open to further suggestions

Thanks for your feedback!

Dr4gul4 wrote: »

As i've been runing HA for a few months it's also evolved a bit over this time, going from a hardcoded entry in a yaml file, to being fully integrated into the Hass.io OS.
Part of this was not every happy when they updated MQTT to 4.1 for me and several others at least.

The issue was all devices could not talk to the MQTT broker, even with correct creds provided, I repeatedly received socket disconnect error's

Once clean install later, and some choice restoring of settings /addons etc and I was up and running with a clean install.

Ah!
An update

If you close the actuator when the return temp hits say, 35 degrees you're closing off the loop fully and no longer heating the room. you have now stationary water sitting at 35 degrees for a small period of time.

What's you're heat source for the UFH ?

Not so ....... but no longer heating the floor ..... which is essentially one very large storage heater. A heated 4" slab of concrete the size of the room. It emits warmth for a few days if not heated and the room temp will gradually fall away.

Oil burner is the heat source.
I had considered earth loops etc. but could not justify the cost without guarantee of efficiency or indeed that it would even be fitted correctly.
At least with oil boilers most plumbers know their stuff

At the time the air to water systems were not efficient, although I believe they have improved in the last few years.

I've heard this mentioned, but as of yet it's not something i've looked at, could be good

I had been thinking of this for cameras
https://shinobi.video/docs/

It seems there is a community plugin for Hassio which I was unaware of
https://github.com/hassio-addons/addon-shinobi

Will all camera's come with power connections in the future for upgrade? It's ahrd to say of course, but I'd perfer Cat5e/ 6 and a POE switch. changing camera's becomes less of an issue.

If not then a simple adapter will inject the power into the ethernet cable from a power source in the media press for whichever camera might require it.

Looked at these recently, but it doesn't really fill a gap i have right now, Might consolidate some hardware soon in to a new mini server and put HA on there as a VM. but ! those nuc's are nice .

Expect to chop and change, be frustrated and elated. I fear it's all part of a custom centralized solution.

NUCs are costly and I won't be going back, while at the same time near impossible to work on because of form factor.
I guess the high cost is mostly related to reduction in size and as that won't bother me in the media press then a USFF used PC should do what I want.

I don't change easily, but I spend an inordinate amount of time researching and testing before I commit, so it is rare for me to need to change (SO FAR ) That is my excuse for my tardiness!

Dr4gul4 wrote: »

If there's one piece of advice I can offer here, it's that i have the same set up, and yes the slab holds heat for a day or so, but of course that depends on insulation,and air tightness etc.

If i had better of both, i'd swap to a Heat pump tbh. but thats outside the scope of this discussion i guess.

I'll be interested to hear how it goes, im sure with enough time and diligence it's possible to factor in weather, winds etc, and adjust the length of time a loop is open. but with out a room thermostate you may find you're heating a slab blind.

Yes we have really gone off topic
The heat pump is one thing .... but the heat source is what I was not convinced about.
At the time the air to water was not very efficient, especially in Winter when needed most.
If I had a good body of water close by (lake or large stream) I would have been tempted by that.
Underground was considered and dismissed for reasons I not do not recall.
Drilling for a 'well' of water was considered ... seriously ... but ongoing maintenance was awkward and problems with grit coming up were difficult to deal with.
As piped gas is not available to me, and I did not fancy storing it on site, I continued with oil and got a more efficient boiler.

On the matter of how much heat to give the floor .... knowing the usual use of each room a very close guess can be made without any measurements.
After that it is a matter of tweaking it.

I am unsure how good the wind & outside temp measurement will be in that tweaking ...... I might never need to bother

Dr4gul4 wrote: »

To bring our topic slightly back on track, with UFH there is or was a process devises called "step back" where by, in a 24 hour period, you allow 8 hours for cool down, and 16 for heating, this ensuring a constant average heat within the block.

It's actually very similar to what your proposing, and fully viable via the sonoff 4 chan device

Thanks, I had not heard of that "step back" or if I did failed to associate it correctly.
Any relevant links available?

On average my floors need 2 hours heating .... depends on room use and other factors of course.
Need an extra boost at odd times.

Dr4gul4 wrote: »

To bring our topic slightly back on track, with UFH there is or was a process devises called "step back" where by, in a 24 hour period, you allow 8 hours for cool down, and 16 for heating, this ensuring a constant average heat within the block.

It's actually very similar to what your proposing, and fully viable via the sonoff 4 chan device

I have some Sonoffs (4ch and also single) on order to do some tests, and have been thinking about this strategy for the UFH.

I think I will adopt this idea with some adaptations to suit my situation.

If each loop return (thus floor) can be kept to a minimum temperature a few times in the 24 hours, then it is likely a lower return temp will be needed, and give less likelihood of 'overshoot' of floor and room temp.

I will look at that in conjunction with the Sonoff timers.

Thanks for the ideas

deadlast wrote: »

My old APT analog immersion switch has given up again. What are my options with regard to fitting a smart immersion switch?

I've searched the thread and people are saying the SONOFF pow r2 is suitable, others saying they seem to be a fire hazard.
http://s.aliexpress.com/ERZnMZVr


I'm have very little knowledge when it comes to electrician work but my brother is law is a sparks so will get stuff and ask him nicely to wire it in....just need to know what to get first?

Use a Basic Sonoff switch to Switch a heavy duty relay which will in turn switch the immersion on/off.

The Basic Sonoff switches are not rated sufficiently well for such use.

The one you linked seems to be rated highly enough, but .... do you want such a device working close to its rated capacity?
I wouldn't TBH.

Hi

Just wondering if someone might be able to help me, I have 3 son-off basics controlling lights. A few months ago something happened my broadband router and I had to do a factory reset on my router, this in turn unlinked all my devices, I'm only getting around to linking the son-offs now but I'm having no luck at all, I've tried several approaches, sometimes the app recognizes the switches sometimes it doesn't, but it will never connect to it.
Would anyone be able to shed some light on this.

msmx5 wrote: »

It is most likely that router/phone connection is not operating on 5Ghz. The sonoff's don't support this. Most routers will operate at both 2.4Ghz and 5GHz so it should be possible to set the sonoff's up again but you need to make sure the phone you are using to set them up with is on the 2.4GHz wifi.

How to do this? Depending on your router/broadband provider you may be shown separate wifi SSIDs (wifi names) for each signal, if so, on your phone connect to the one with 2G or something similar in its name. Some routers (latest virgin media home hub being an example) use the same name for both 2.4Ghz and 5GHz signal - this makes it more difficult to know which your phone is connected to, you may need to log into the router and disable 5Ghz while you set up the sonoffs (after they are connected and working you can re-enable 5Ghz).




Good luck!

Hi

Thanks for your help, I bought another sonoff and linked this last night without any problems, it just seems to be the ones that were already installed that will not link.

yep, I tried both methods several times.

New relay switch with dedicated connectors for a switch now out.
As cheap as the basic, now no gunthering the basic to connect to a switch.
Will be getting a few of these, if they follow up with a dimmer version, it'll be a game changer for me in the Holy Grail of smart switches.

Edit, they confirmed in a tweet they hope to have a dimmer version out by the end of the year.

I can't seem to get inching and power on state to work together.

The options for power on are ON / OFF / Previous state.

Inching is in seconds up to 3600.

What I'm trying to do is set up inching for say 1800 seconds, so that once it's turned on, it automatically turns off after half an hour. This works.

But, when I enable inching, then the power up options (in my case, Power ON), seems to be ignored, so if I lose power, and the power is restored, the sonoff's output stays off.

Firmware is up to date.

Is it a bug..., Or am I missing something else?

The application I'm trying to use this on is as a time delay function for a landing light. So, when it is switched on, the sonoff comes online with output on, and Inching cuts the output off after half hour. And the landing light goes off.