Open source smart home planning with wired appliance control

Name: Open source smart home planning with wired appliance control
Uploaded: 2016-05-31T01:50:58+02:00
Description: Hello, this is my first post so welcome to the forum. The topic is quite vast - so I'd like to sort out a bit of what I've already read, and I wouldn't want to reinvent the wheel and duplicate the mistakes of other installers ;) First thing - at this stage I'm not interested in any off-the-shelf...

Question

Hello, this is my first post so welcome to the forum. The topic is quite vast - so I'd like to sort out a bit of what I've already read, and I wouldn't want to reinvent the wheel and duplicate the mistakes of other installers ;) First thing - at this stage I'm not interested in any off-the-shelf...

idepopizze · Accepted Answer

@Voder You are confusing terms. An electro-valve is not an electrically operated ball valve, because they are different things. Personally, however, I would refine this and meter the water meter + connect to the alarm. If you leave the house, the assumption is that no water will be drawn at that time - so there is no need to leave the laundry. If the water starts to run while the alarm is armed, the valve cuts off the supply and notifies you by e-mail/sms.

idepopizze · Answer

Ad9 - I think it's simple here - some kind of electro-valve

You'd better read up on solenoid valves or, better still, look at the cross-section of such a valve - you'll understand what's going on.

---------------------------------

When I look at these assumptions, it looks like a house and there is no boiler room. Am I to understand that you have a rudimentary boiler room in the style of a mop cupboard and in it you have a gas condensate unit taking care of this floor and DHW?

Voder · Answer

Re9 - as far as solenoid valves are concerned - well the subject of controlling RO water with a solenoid valve to fill a topping up tank in an aquarium I have covered. It's not some rocket science.

I had a preliminary look somewhere on google and I even saw one with a DN150 for drinking water ? I can't remember what diameter I get into the house anymore, but it seems much smaller to me. Let's say let's make it 1 inch which means if I checked correctly DN25 - such a solenoid valve for 16bar is 250zl on the first google search.
Can you write what exactly is involved ?

As far as the boiler room is concerned, I missed it - I abbreviated my thoughts too much - it will be a solid fuel cooker - pellets, although it is possible that it will be a gas cooker if we can get gas. Of course, the furnace will be controlled as much as possible, although I plan to have underfloor heating throughout the house and no radiators at all (well, maybe one in the bathroom or in the garage).

zmyslonyy · Answer

If you want to base such an ambitious smart home system only on relays or contactors, the idea is to be congratulated. I understand that with these relays and contactors you will also generate communication with audio/video equipment, alarm and others... (although I do not know how you will solve this in the other direction?) because if not, then this is where the stumbling blocks will begin, namely communication over relatively longer distances than at arm's length.
Another thing is you are asking others what this can be built on? You don't need an arduino to control the relays...but it's a question of what would it realistically do?

idepopizze · Answer

Domestic water systems are generally half-inch diameter.

The point is that at the solenoid valve you have a very small cross-section through which the water goes. Find one for your particular solenoid valve. A very small cross-section means high throttling, but what's worse is that any dirt can clog such a valve. I know that there are strainers, filters, etc., but I have cleaned the filter before the water meter a few times and I know what can flow in such "clean" water.

You should also read up on triacs since you are already at the relay stage.

A solid fuel boiler and direct underfloor heating?
No buffer?

Voder · Answer

@idepopizze - I've also seen a solution with triacs instead of relays. This is some solution that solves the issue of relays beeping. With relays it's simpler as I have a ready-made one and don't have to build a circuit. On the other hand, there is nothing stopping me from designing and preparing them in the form of interchangeable modules and having them in reserve in case of emergency.

As for the solenoid valve, well, I thought it was an ordinary, e.g. electrically controlled ball valve, where instead of a handle there is some kind of motor, a servo which opens and closes on command. I will look for more information on this, but such solenoid valves are used and I have read about them many times.

Of course, there will be a mixing buffer - unfortunately there is still gas, but it's on private land across the street, the neighbours don't agree to a connection and the gas company doesn't see any economic justification, as the nearest node I could connect to is quite far away. Unfortunately a heat pump is out of the question for the time being as it would probably be the best. I don't want to build or bury a LPG either. I'm left to choose the cleanest of the dirty - pellets :/

@zmyslonyy - read carefully - relays and contactors were mentioned in ad1/ad2 i.e. lighting/sockets - I haven't covered this topic anywhere else.

In the case of A/V, nowhere did I write about relays or contactors - I clearly wrote about Domoticz + Kodi as a media centre with NAS and possibly an IR transmitter for the other devices - controlled from the central unit. Having an IR transmitter - I am able to send any code to the A/V device. Most are probably passive and can't talk the other way - just instead of a remote, it broadcasts the central computer. I may be tempted to add an IR receiver to detect that someone has switched off the TV from the remote control, for example, so that the system is synchronised.
I don't know what else I can think of apart from the fact that when I turn on KODI or the TV, the appropriate lighting scene should start and the roller blinds should close if, for example, the light intensity sensor detects that it is too bright to watch TV. How do you solve this? (I think I've seen your threads here, because I remember your nickname, but I've already read that I don't remember).

Alarm ? - Domoticz supports chattering with Satel and can transmit to Satel and receive information from Satel from what I have read.

As for the question of what to use - as I wrote, I'm thinking of Domoticz + arduino or Raspberry as systems that collect information and pass it on to Domoticz.
The intelligence is embraced in Domoticz itself. But I am amenable to suggestions.

As for long distance transmission - I wrote that I was planning star type wiring so even with the bus I would have most of the actuators in the switchboard. RS-485 is nice - but I can use it to transmit, for example, information from the switchboard to the computer responsible for watering the garden, which will be in the garden. On the other hand, it's enough if I have a LAN there and don't need RS-485 because it will just send me the information to the port via tcp/ip. But if the computer controlling the garden is in the switchgear and it is a raspberry pi 3 or something more powerful, what will control the whole thing? Yes, one point of failure - but it's easier for me to keep one spare computer, especially if its cost is small, than a backup for hundreds of actuators in the form of uc cans.

See - the main point is that I can do everything on my desk - but a desk is not a house ;P Hence this thread. What I would like this avoids a lot of electronics, which can be equally unreliable and one discharge in a distributed architecture can cause a lot of SPOF ;P

Though for that I want a control stick that will switch everything to manual as if I never had a smart home.

zmyslonyy · Answer

Sympathetic colleague @Voder except that you reduce the whole idea of a smart home to either on/off control or a network infrastructure based and limited on an internal LAN. Just as on/off control alone may be too minimalistic in many cases, post-network control of further devices via the LAN is simply form over substance. From a software point of view, it is also much more complicated.

Voder · Answer

Because that's what the whole idea of such a house boils down to, maybe not always the typical binary 0/1 because there may be some intermediate states such as LED dimming or lighting, but in most cases it is like that. Even if I take the aforementioned WS2801 LEDs - switching on, switching off the appropriate colour, intensity will simply involve sending the appropriate control command to the controller of these LEDs to the output. That's it - any other philosophy in this ? I don't quite know how else you imagine it? Based on the given input parameters, something should happen at the output - and whether it is a digital or analogue value depends on the actuator and what it is supposed to do. Anyway, even in a distributed system controlled from the bus, it is also about switching on/off in the case of lighting - maybe only the elements which perform this switching are different. As I wrote - the intelligence of how to link the information from the sensors is to be taken care of by Domoticz and if I think to myself that I would like the temperature in the house to start dropping to 18 degrees at the moment when it is dusk outside but it is not yet 11 p.m. because I go to bed at 1 a.m., it will do it. It will receive information from the twilight sensor, check the time and, if the input conditions are consistent, check the temperature in the rooms and, for example, it will set the electro-valves accordingly, e.g. on the floor, or send information to the cooker controller, etc. This is what Domoticz can do, and if it cannot do it, it can be easily extended with your own solutions. It is like HC2 in the Fibaro system. In addition, it is basically able to replace HC2, as it supposedly supports Fibaro devices quite well. But I am eager to learn something new and what you mean by the idea of a smart home. Because for me it is mainly about replacing me in simple tasks - it won't iron my shirt or cook dinner. Of course - I can harness it to issue e.g. voice commands like here: but, for me, this is form over substance. It's just easier for me to do something over tcp/ip because I do it every day, rather than dabbling in CAN or RS-485 bus protocols. If I were using a Raspberry PI then the ethernet is built in. In the case of an Arduino or AVR, you'd already have to think about a module of some kind, and here it would actually be worse with a LAN, but there's no shortage of ready-made ones. As for on/off - in the case of lighting, that's rather what it comes down to - I don't expect anything else from pressing a switch button. Of course, this can be used in all sorts of scenes - but the links between sensors and scenes I would like to do is something I haven't written about. And seriously - as if you could give some examples of when it would be too minimalistic (and I'm writing this quite seriously) it would perhaps tip me into your line of thinking. I approach to the topic seriously - I have seen many such DIY solutions and my thoughts are some fallout from all of them. But it's not to say that you won't convince me of your approach. @idepopizze - well you see - I've always been a lousy plumber ;P I've always been rather attracted to bits I'll have to educate myself in that direction

zmyslonyy · Answer

@Voder well, that's not really what it's about... because in a so-called intelligent house system, in which the on/off system functions first of all, no other states are signalled, let alone failures, and you can live in the belief that everything works beautifully and here, for example, a light bulb burned out and it doesn't turn on because how? the roller shutters do not close because, for example, a child has put a toy there or even worse, the mechanism has burnt down...etc... Do you have to get into the CAN or RS-485 bus protocols? Oh mercy...and how much data do you send to use a LAN for this?

Voder · Answer

@myslonyy - no no ok, but that's more like the intelligence level of the system. I don't think I've seen a solution anywhere for the roller shutter controller to communicate the position of the shutters. You could add some kind of sensor measuring the distance between the roller shutter and the window sill or checking whether the motor turns the shutter (just like in a computer mouse with a ball ;P) - but I don't think it's so necessary at the moment. There's no substitute for thinking for the man in the house anyway, and I can always find something to check. Most roller shutters seem to have some sort of stop - the shutter closing time is also known and rather fixed. I can base what position the roller shutter should be in on time without adding more sensors. Maybe it won't be as intelligent as with the position sensor, but with the right algorithm and appropriate counting it should work. If the average closing time of the roller shutter is 10 seconds - and the limit switch does not switch off the roller shutter after 12 or 15 seconds - disconnect the power supply and send an alarm. Intermediate states when the roller shutter is not in the upper maximum position - the counter counting time will know that I closed the roller shutter for 5 seconds, so it should be halfway and it takes another 5 seconds to close it + plus some additional time before I send the alarm when the limit switch does not work - in this case, e.g. these 7 or 10 seconds (it is all a matter of repeatability of closing speed of such a roller shutter and the fork in which it closes). But if I have a twisted pair and a power cable at the point, I can extend the system with other sensors and add a position sensor to the roller shutter mechanism. However, it will still come down to reacting to the state of such a sensor - and to ... simple on/off. Unless you know of a method for such a smart building to remove the toy itself Same for a light bulb - if it burns out the system won't change it for me anyway. It can inform me that it's burned out. Anyway, I can probably implement this quite simply by measuring the current in the circuits already behind the relays in the switchgear. Relay on, no current - means no light and you have to raise the alarm. I don't need a controller at the bulb for this. But it still comes down to receiving information from some sensor and programming the response. I might as well, instead of just sending out an alarm - in addition, switch on another circuit to make it bright even though the bulb has burned out. It is a question of intelligence and what information I gather from sensors and how I use it to control circuits, most of which will be based on on/off status, whether controlled by relays, contactors or triac. Whether I trigger the relay with infrared, bluetooth, wifi or CAN bus not to mention RS-485 is a matter of implementation. I do not see any intermediate states as such, even if I decide to control the roller shutters depending on the information from my weather station, sunrises/sunsets, in the end I will only need to switch on the roller shutter motor and this must be done in some way, e.g. using a relay. To be clear, for each point or circuit, in addition to the 3x1.5 or 3x2.5 wires (although I am still thinking about the number of pairs to give more), there will definitely be at least one twisted pair, if not two (the second one as a possible backup if one of the wires dies). And I'm not talking about the LAN at the moment. So some possibility of expanding with more sensors I should have. And if I forget something and don't have anything nearby to what I could connect the new detector to is an arduino nano with a shield sending data on 433MHz and I close the topic But I also don't want to go to any extremes and saturate the house with electronics that may be unreliable - on the other hand the house is supposed to be functional. The fact that a light bulb burns out is also something that my sensor in the form of both pairs of eyes will notice. The more complicated something is, the easier it is to fail. For me, the most important thing is - to be able to quickly switch to manual operation and quickly repair what is damaged. A switch with a uC in the box, which will send information via RS-485 or wifi that I have pressed a button, does not really appeal to me. If it's hit, I'll have to replace it with a traditional switch. As for learning - well - if you haven't done something you always need to learn it - it seems that in one of the books about AVR and C - Tomek French or Mirosław Kardasio everything is laid out on a tray, but again, with Linux I have 20 years of experience so here I am able to figure out a lot of things yesterday But I have not written anywhere that the bus topics are impossible to understand

idepopizze · Answer

@Voder I suggest exploring the plumbing so you don't have to sit in the boiler room unnecessarily later. Hot water, after all, can do itself from spring to autumn. With the buffer, I'm not sure we understand each other either. I was thinking of something with a capacity of around 1,000 litres. Of course, everyone has their own priorities.

zmyslonyy · Answer

@Voder you see the whole issue is not at all as trivial as you try to portray it. If you don't adapt the right electronics and mechanics to the design of a so-called smart home, let's just hope no accidents happen in the process. It is obvious that a burnt-out light bulb is a trifle, but imagine, for example, roller blinds or actually their mechanism, which can, for example, crush a child's hand or a pet. There is no emergency stop mechanism, no reaction or information about an event. The same is true, for example, of the entrance gate to a property, where the mechanism, controlled by electronics, strives to open or close the gate regardless of any factors that may occur. It is not difficult to imagine that rather unpleasant accidents could occur. Of course, the electronic gates have appropriate safety features, but the on/off roller shutters do not... It's all well and good if it's controlled by someone who knows how to behave, but it's worse if it happens in the absence of such people. Besides, it's not difficult to imagine a child watching a roller shutter being folded and trying to block it by putting his finger in the guides. How could this end? It could look very bad indeed.... Of course, what can be predicted quickly...but the question is what can we not predict at all but can happen? according to Murphy's laws, anyway. It is also worth mentioning that the first cars with electrically opening windows also had no safety features other than information about the extreme position but only until there was a strangulation accident with such a window where there was no way to stop the action and someone thought they had time to take it away before it got stuck on their head. Also I would be very cautious about such solutions.

Voder · Answer

@myslonyy - except that external roller shutter mechanisms don't seem to have such solutions whether they are controlled manually or by the house automation. As I wrote I don't think I have seen a roller shutter solution that is more intelligent. The ever so popular Fibaro doesn't seem to have such a thing. Or at least I haven't seen in the manual that it can do anything more than control the roller shutter motor. It does have calibration so I suspect the 'position sensor' is based on a timed algorithm. Even if I have a window-opening detector, it too can fail, which I don't foresee, and disaster is afoot. And when it comes to children, you need to keep an eye on them at all times, because in this case it is not difficult to have a disaster. @idepopizze regarding the buffer, I thought you were asking if I was going to feed the underfloor directly from the cooker. It's well known that the underfloor is low temperature, the furnace reaches its highest efficiency for higher temperatures so I need some sort of mixing valve to feed the underfloor at a lower temperature. The DHW tank will definitely be there, I don't know if 1000l rather much smaller. Unless there is something else you have written about. Feel free to scribble on private so as not to get off topic.

zmyslonyy · Answer

@Voder and what happens if the shutter closing mechanism locks up? Will it burn out the drive motor?

Jado_one · Answer

Hello,
Let me join the discussion, as I am just at the stage of constructing electrically controlled roller shutters - but indoor ones - and this topic is close to my heart.
I've made a roller shutter controller (supports 5 roller shutters) and of course you have to provide for various safety features to avoid further damage in case of a failure (e.g. mechanical interlock).
And here we have several types of protection:

- limit switches (actually one photo-optical, but which detects both max up and max down positions).

- overload sensor - the motor controller has the ability to measure the current drawn by the motor. In the event of a blockage, there will be an increase in current and the controller will switch off the voltage going to the motor.

- timeout - measurement of the time needed to fully open or close a roller shutter. If the maximum time is exceeded, the voltage to the motor is switched off

- mechanical protection in the form of a friction clutch. If the roller shutter mechanism is blocked, the clutch will allow the gears of the drive mechanism to continue moving (albeit slightly overloaded). This avoids breaking the teeth in the gears.

In addition, the following situations still need to be foreseen:
If the roller shutter is in motion, movement in the opposite direction cannot be switched on immediately - the movement must be stopped first.
If the roller shutter is at the lower or upper end - the movement of the roller shutter can only be activated in the opposite direction (i.e. if the upper end, only downwards, and if the lower end, only upwards).

The controller allows you to control each roller shutter individually as well as all 5 at the same time (one pair of buttons for each roller shutter and one pair for all of them).
There is local control (buttons) as well as remote control (via radio).

I'm now at the stage of mechanically reworking the normal roller shutters - the corresponding bearing, motor, clutch, etc....
Of course this is fun for those who like to tinker a bit ;-)

So if the fellow author of the thread is serious about the safety of using roller shutters, he would have to implement similar safety features in his version of the controller.

Especially since we have here the possibility of automatic closing/shutting of the roller blinds depending, for example, on the intensity of light or at a specific time - without human involvement (the human may not be at home). Then there is no one to "fly up and switch off" when something happens ;-)

I am generally in favour of my own builds - especially those based on Open Source and Open Hardware - if one likes to tinker and has the right tools/skills/friendly workshops/ etc. etc....

zmyslonyy · Answer

...and knowledge mate @Jado_one . Of course I agree with you all the way and I would only add that if this was to be a so-called smart home it should nevertheless signal at least a problem. That is, there should be some kind of feedback channel to determine the operation or problem with the device. I am leaving aside more advanced solutions which identify the problem/failure and present it in a form readable to the user.

Jado_one · Answer

And, of course, the radio communication I have implemented in the controller makes it possible not only to send the appropriate command to switch on/off a given (or all) roller shutters, but also to send feedback on the status of the device or signal a failure. the radio communication I have implemented in the controller makes it possible, apart from sending an appropriate command to switch on/off the movement of a given (or all) roller shutters, to send feedback to the superior controller on the condition of the device or signal a failure.
The superior controller - if it has such a possibility - may then notify the user, e.g. via SMS or the Internet, about the situation - and, most importantly, it will stop further transmission of orders to the faulty device.
Further action requires direct intervention by the user, who must determine what has happened and restore full functionality of the device.

Generally speaking, any device which is beyond the direct control of the user (and this is what automation is all about - so that you don't have to keep an eye on the devices and their operation), and whose possible failure may lead to a dangerous situation, should be well protected against any eventuality
Although it is probably impossible to predict everything - e.g. what will happen if the cat jumps on our roller blind because it gets curious that "it moves by itself" ;-)

Voder · Answer

@Jado_one, @zmyslonyy - some external roller shutter controllers or motors have overload protection built into the roller shutter controller regardless of the control method, if only for protection in winter (freezing of the roller shutter). They also have in-built protections, e.g. to prevent up and down controls at the same time - so it is also a question of choosing the right motor with the roller shutter. The motors also have built-in thermals. So it's not quite the case that external roller shutters are so bare-bones. The disadvantage would certainly be the lack of transmission of information on the tripping of the security device to the central system. This is where an in-house A to Z controller is much better.

Yesterday, out of curiosity, I had a look at how Satel, for example, solves this:
http://www.satel.pl/pliki/ulotki-knx/MD-ULOTKA.JTJZ.ROLETY-PL.pdf

or
KNX/EIB - I haven't seen their actuator do anything more than apply a voltage, it only has a safety feature so that up and down voltages are not applied at the same time.

I'm not in any way negating what you write about because I fully agree - but it still comes down to sensors and reacting to them

- any one of the protections put in place by @Jado_one (other than mechanical) will disable the roller shutter actuator plus alarm. In my case, there may not be an alarm only with some of the security features that the timing algorithm will not grasp.

And if the cat jumps there will be kisiel or kebab for afternoon tea in the evening ;P

And when a person is not at home, the blinds are down, the alarm is armed, the house sleeps except for the necessary functions

@Jado_one - is it possible to find somewhere descriptions, schematics, charges to the procs of what you are creating ?

pawel3110 · Answer

As far as I have noticed, Domoticz and other systems of this type have one big disadvantage - they do not support confirmation of the execution of a sent command, e.g. a switch - we click, the command is sent and the bulb shows that it is lit, while physically there may not even be any communication.

Speaking of roller shutters, in your opinion, do group or central switches also represent the same risk? Let me remind you that roller shutters have their own security devices.

zmyslonyy · Answer

@Voder so I suggest you make another very negative assumption. What happens if, for example, a friend comes to you with a child and it is him who gets hurt as a result of a malfunction of, for example, roller blinds. If he takes legal action, you will not be paid out for the rest of your life.... Everything is fine when it's relatively new and running smoothly, but when it starts to age and break down, that's when the problems begin....

Voder · Answer

@myslonyy - you are going to extremes and making things up a bit by force. I might as well manually control the blind while looking at the TV for a mishap to occur. Second point - I have already written that roller shutters often have their own security features even when manually operated. @pawel3110 - I'm waiting for the raspberry pi 3 to play with it more. So far I've only played with dummies devices. Thanks for pointing this out, I'll definitely take a closer look at this as it may indeed be a flaw and one should think how to protect against it. Fortunately Domoticz is open source

pawel3110 · Answer

You can run Domoticz on windows and test it.

Jado_one · Answer

As far as the roller shutter controller is concerned, at the moment everything is still in the prototype stage - there is a development board created on a PIC32MX250 processor, there are other small boards with ICs connected (I2C expanders, RF tansceiver, H bridges for motor control, etc, etc...) Once I've got the roller shutter mechanics done, I'll have a look at the connection to see if everything works as intended - some of the features still need to be considered whether to implement them or not (e.g. time control to allow incomplete opening/closing of the shutter based on time measurement - whether this is necessary or not). Once the prototype is working sensibly, I will design the PCB and make the controller in usable form - only then can we talk about the final schematics and software. Another issue is the still unspecified communication protocol between the master controller and the roller shutter controller - the hardware layer of the protocol is done, but the interpretation layer (what command means what) has not yet been has yet to be invented. Probably because the master controller is also still in the prototype stage - for now it's doing a talking clock and an electronic DJ (The main idea of the master controller is to be able to generate voice messages - e.g. reading the current temperature, pressure or other parameters - including system error messages and any notifications).

Voder · Answer

@pawel3110 - I have both domoticz and homegenie running on my laptop right now and I'm playing around with them, exploring the possibilities, but for now locally. @jado_one - you can always use the time algorithm to calibrate a new roller shutter which has different times than yours and does not have a distance sensor fitted to have info on the position of the shutter. As long as the controller is to be universal. Two, you have the added security of having your photo-optic sensor go down for some reason.

zmyslonyy · Answer

@Voder that's life, not extremes...That's what the intelligence of automation is for, among other things, to prevent various unfortunate accidents or events. Roller shutters probably have current protection...but it is not difficult to imagine how something wedges itself in the roller shutter guide and shuts down the drive...and then cycles it just on and off...because the control signal has no feedback of the failure. Anyway, life writes scenarios that even people themselves cannot fully predict.

Voder · Answer

@myslonyy if the load is too high then a thermic will be triggered, for example. Protecting the roller shutter motor. Not to mention that you are mistaken in your assumptions - because if the roller shutter does not close in a certain time because it does not reach the limit switch, there will be an alarm in the system and the system will not control such a device, so it is not difficult to imagine that it will not switch it on cyclically until a person reacts. The feedback to the system could be the limiters and the timing algorithm.

zmyslonyy · Answer

@Voder Your original assumption you wrote about didn't involve some kind of timed switching at all, which means there would be fixed states. Anyway, if you wanted to control it with one line (controlling the relay) how would you get three states? Only by having fixed states on the H and L lines. I rule out putting a high impedance (processor pin as input) into state three because at such distances it would be an antenna collecting all the junk from the ether and introducing interference. Unless you further limit the functionality and the automation can only slide in but no longer push back. Otherwise, you need a roller shutter controller that converts the control signal into a timer and controls the process of shutting or pulling the roller shutter down at all.

Voder · Answer

@myslonyy you read not very carefully: i wrote: Ad8 - here usually relays to start the roller shutter motors. Dependent on dusk detector, wind, temperature etc. Note the words dependence on and etc. What I've thrown in here in the first post are ideas and not the final solution - it would be hard for me to write everything in detail in one post. Roughly as I envisage it. I'm not quite sure where I wrote about a single line either ? :o There's a roller shutter - I feed a 4 or 5 wire 220V cable to it - the control for the roller shutter motor is 3 wires - to be able to control up, down. I also wrote that I will have a twisted pair cable to each point - this is where I connect the limiters. Let's put the limit switches on the GPIO of the Raspberry (although it doesn't matter if it's a Raspberry or an Avr etc). The up-down switch to the next GPIO of the Raspberry, and another GPIO to a relay with two NO outputs, or two NO relays. The controller will be the raspberry - something has to control these roller shutters after all. It will also be able to control, for example, LEDs when it has free ports. And if I add some Domoticz to it, it won't be a roller shutter controller anymore? :o I may be taking shortcuts ... ... so maybe I will collect the main thoughts: 1) All detectors, circuits go down to one or two switchboards in a star arrangement 2) Basic functionality I need to be able to switch to manual operation with some kind of switch e.g. group switch etc. Here I have not yet looked how this can be solved. 3) The vast majority of the controllers are in the switchgear. If I managed to splice everything on one controller there could be one.Therefore here I was thinking about some shift registers or ready-made ones on i2c. If there is one controller managing everything, I will definitely have a second one as a backup. I want to limit the amount of electronics that can be damaged. I just wouldn't want to find myself in a situation where I have a distributed system, lots of microcontrollers on cans and after some discharge a 20pc failure with smashed electronics and half the house not very functional. I prefer to swap in one place with a piece waiting to fail. 4) Weather station, possibly garden controller will be 99.99% wireless.

pawel3110 · Answer

thoughtfuly I think you are exaggerating again. A typical solution in automatic control of roller shutters is to give a signal for a certain time (time of complete closure + some provision for increased resistance in winter) At this stage of the author's project do not require such details, which does not mean that such solutions will not be implemented. For now the discussion boils down to don't do it or you'll go to jail, or put a million five hundred sensors on one roller shutter and preferably make your own and I don't think that's the point. Group and central switches have been available for a long time and somehow nobody has had any objections to them.

Open source smart home planning with wired appliance control

Post #1

Post #2

Post #3

Post #4

Post #5

Post #6

Post #7

Post #8

Post #9

Post #10

Post #11

Post #12

Post #13

Post #14

Post #15

Post #16

Post #17

Post #18

Post #19

Post #20

Post #21

Post #22

Post #23

Post #24

Post #25

Post #26

Post #27

Post #28

Post #29

Post #30