Learning as you go is a big part of any DIY project. Seldom does a project go directly as planned, expected or even as I hope it will. THANK YOU for your work!!
Fantastic video, and refreshing that you show all the mistakes and learnings rather than typical on such howto videos where everything goes off without a hitch!
That is freaking awesome! I also love that it's a longer video where you explain everything because it helps me learn about the topic but also train the way I think about situations. Thanks!
Great video Jonathan. I love seeing the trials and tribulations of a project like this. Problems arise, solutions are found, directions change mid project. Fantastic!
Thanks Michael! The good news is that updated PCBs for the LIN module have arrived, so the RJ45 socket fits perfectly now. I still need to finish my integrated board that will replace the collection of modules linked together with jumper wires.
Thanks Chris! It's been so long in production that I've come to dislike it, because it felt like the episode that would never end. I'm glad it's useful though :-)
You are way down the road from where ever I might ever get, but I admire your thought process, when it doesn't work as expected, keep screwing with it until you figure it out. Your videos are great because you admit/show that crap happens and you just keep working on it, for a solution.
I love the Big Scary Laser label. As usual, you deliver quality content that is worth the wait. I can only imagine how much time and effort go into your videos. You always give such great detailed information including bonus hints and tips. It is great that you share your learning experience and how to debug. I love how well you future plan each time. Must be nice to have everything as organised as you do, button here, switch there, that on, this off, great times.
Great video, and I love how you showed your learnings along the way. I also like your rules regarding where to locate your automation gear. I built a walkway through my roof space when I built the house, so I thought it would be a great place to put my house control gear. Unfortunately I need a ladder to get up there to my main home controller unit. Never again!
Great video! Thanks for showing the change of thoughts, smoking your board, etc. Lots of good lessons from that. It is always easier to do the second or third time.
Another great video, definitely one of your best. Like many others it is refreshing to see someone document their mistakes and changes along the way, instead of just a glossy finished project that looks like you are a genius for getting it so perfect first time! Really good stuff, looking forward to what else you have in store...and trying to think of my next project...
Thank you Jon! I agree with you and the others that seeing the pros and cons is a fantastic learning experience! I believe your explanations of the possible solutions were very valuable, entertaining as really solidified the understanding! Appreciate your willingness to share your time, money, and effort for all our enlightenment! Hooked on SuperHouse Scott!
Effective pedagogy! Awesome and rich of lessons of life and approach of problems. Thanks you very much. Your videos are instructive, very helpful and give optimism and enthusiasm to learn. Explaining your design the way you do is fundamental for understanding the whole picture. I just discovered your channel, your gorgeous channel. I'll recommend it to my people. Vive l'Australie! from one another down-under island, Madagascar.
Looking forward to the new architecture set up video. I know there is one in progress. Also I thought you had a Netgear commercial PoE but i suppose you need a handful of them considering all the devices you have.... Great work and thank you for your time and knowledge
What a great sun room. The interior supports for the windows act as a heat sink. Window support structure needs to be on the exterior of the windows, preventing the reflection and heat sink effect.
Thanks for this video. I really enjoyed it being this long. A lot of good information you presented, and I am also glad you did not edit out your failures. That just lets everyone know that anyone can make a mistake. Maybe you can make a video where you can check what died on the Arduino Board, and repair it. I agree with Alan Woods comment, where you split the data lines up so that 2 data lines are not on the same twisted pair. If you have time, maybe you can see what uC is in the actuators, and substitute it with a nano with your own code. That way you can have a single bus, and just address them seperately. Also, as mentioned, the self test on power up is a big problem. I have a Wemos D1 Mini Controlling my Garage door motor from my openhab and mqtt server, and somewhere I have a discrepancy with the system, where sometimes if the power to the Wemos is cut off and restored (or sometimes even if the Access Point is restarted), that the garage door opens! But I at least have an openhab rule to automatically close the door after 30 minutes, so it is only a risk for a bit (but in South Africa, with our crime rate, it is actually a good opertunity for oppertunistic shoppers).
I've already started designing a PCB as a replacement for the controller shown in this video, so that everything is mounted together and there aren't any interconnect wires at all. It'll be much smaller and neater. I'm going to put a scope on the data lines and experiment, to see how much cross-talk there is both within the twisted pairs and between pairs. I definitely thought about replacing the control board in the motors, because my natural reaction when something doesn't work exactly the way I like is to think "I can do better than that, I'll make one myself" but I resisted the urge to go down that rabbit hole. Regarding the security problem, I totally agree that it would be a serious problem, but the motors haven't done it since I've had everything connected. I think they might only do it before they have been bound to a remote control, then afterwards they assume they are OK.
+SuperHouseTV your welcome. Keep up the great work and vids. Your teaching and explanation are one par. I like the fact you make and sell what you discover and share what you know and learn. It means a lot to me personally! 😎👍👍
Great vid and very informative. For your next project please leave in the mistakes and changes too once again - found it helpful to learn from your mistakes too with the pitfalls to avoid and thought processes to follow for my own projects. Loving the laser cutter by the way!!
Thanks Steven! I can't believe how long this episode turned out to be. Hopefully showing the adjustments along the way will show people that they should just have a go, and not worry if their first attempts don't quite work out :-) The laser cutter has been fantastic since I went to the trouble of setting it up properly and removing all the reasons for not using it. I used to have to drag out a ventilation pipe, set up the water cooling, etc every time I wanted to use it. So I spent about a day installing a permanent fume extraction system and connecting everything else so that it's permanently ready. I can walk up to it and be cutting less than a minute later. It's amazing how going from 10 minutes setup time to 1 minute setup time makes me so much more inclined to use it for even trivial things.
Great video Jonathan! Show how your project evolved was very inspirational. Could you think about a video covering your OpenHAB setup? Maybe discuss how you integrate various technologies via the different bindings etc.... Thank you for your contribution to the community.
Hello Jonathan, I can see that on the LIN BUS CHIP you have chip select on them, why are you not using that to select the individual chip you want to communicate to, bridge all the serial TX and RX to an UNO RX and TX so that you can cut down the cost of the project. I see that you need CS to be connected to VCC, so connect the with pull down resistor to ground and then using a current limiting resistor you connect the CS to digital ports on the UNO and then select the CHIP to receive commands.
Maybe using each solid colour for a separate data line, and its pair (strip line) for the power. This would give a bit more seperation between the data lines. Or - LIN1 Gr - LIN2 Bl - LIN3 Br - LIN4 Or/W - 12v Gr/W - 12v Bl/W - 0v Br/W - 0v
I'd love to quantify this. I'm going to follow up later with another video about making a more permanent controller with all the LIN drivers integrated onto a single board, so when I do that I'll put a scope on the lines and experiment with how much cross-talk there is both within the twisted pairs and between wires in different pairs. It should be interesting.
Great system dude! For the crosstalk issue i'd changeup your pinout, For-Example: Use Solid colors in the Cat5 cable to handle the LIM data then pick 2 stripe colors for DC+ and the other 2 for DC- That way you have minimized parallel capactive coupling between the pairs.
Great and inspiring video as always! Please do more videos like this where you don't cut out the encountered problems it makes the videos very interesting. Did you ever try visualmicro for your debugging and coding? Like you told in your video you do a massive amount of debugging with println in the serial monitor with visualmicro pro you can debug your arduino just like a pc based code. I am using it for quite a long time now and it is the most helpful arduino tool I have encountered so far.
Thanks for the comments! And thanks for the tip on visualmicro. I haven't used that before, so I'll check it out. I don't run Windows but I can set up a VM if necessary.
SuperHouseTV Yeah I am using Visual Studio since i use it every day at work but at home I run a windows VM using KVM and its working flawlessly. Depending on your OS it can be a bit challenging to get the serial connection to the Arduino through the virtual machine but once it is set up i haven't had any problems.
Awesome video Jonathan! Just awesome. To be honest I've so many other building projects going on I've been lazy with my home automation projects, but tonight I've sat transfixed with this video. It had a bit of everything I love - coding, soldering, pcbs, grinders, drilling and, of course, LASERS!! I've fallen in love again with home automation because of this one video. My wife won't thank you when the house scares her again and starts beeping at her, but I will thank you for this inspiring video. It must have taken you ages to film but it's so worth it! Look forward to many more when you are able. *googles* LIN bus...... :)
Thanks Graham! I was really worried that a video this long would make everyone give up after the first 10 minutes, but when it came time to editing there wasn't a much I could leave out and still cover all the important points.
I must admit when I saw the running time I did think, hmmmm. But when I started I just wanted more. As I said in my first comment, it had a bit of everything and to see a project conceived, designed, the pitfalls and the final result in one video was great. Keep going with the videos!!
This is great!!! It would be amazing to see more in-depth videos of this length! How long did it take for you to knock this one out? Also, whats with the audio (RF?) noise when you're in your lab? Are you using the wireless lapel in there, or is it wired? I'm guessing since you're in there with a bazillion scopes and power supplies, you're bound to get _some_ noise! :) Edit: Also, awesome to see Inkscape out in the wild!
It was recorded over a period of months, with maybe 4 or 5 days of solid work at the end to get it finished. I still have a couple of days work left to write the supporting documentation for the episode page on the SuperHouse site. I experimented with a few different microphones during this episode because getting clean audio is really hard, and I'm trying to find a way to do it that isn't so much work. In this video I used a wired lapel mic, a wireless lapel mic, and a Rode Videomic Pro. I also used 4 different cameras trying to find a process that would make it quicker and easier to set up each time I want to record something.
Hi Jonathan, really interesting video, well worth the wait. I liked the longer format; seeing the whole process, the decisions made, decisions changed, etc. I did wonder if power was going to be a problem, especially simultaneously driving 4 motors from a PoE powered device over CAT5, but you didn't seem to have any issues with that. I guess the little motors don't pull that much current then? I had never heard of LIN before, I thought cars used canbus?
There are some issues with power, but it's fairly minor. If I have two motors running and I stop one, I can hear the remaining motor speed up slightly. That means there isn't quite enough power getting to all the motors, which could be either voltage drop across the cable or it could just be that the little plugpack feeding in the power can't supply enough current. I haven't bothered metering it yet to find out but that's on my to do list for a follow-up. Cars do use CAN, but CAN is overkill for some simple applications so LIN was developed as a supplementary bus. CAN is fast and complicated, LIN is slow and simple. The usual topology is to use CAN as high-speed links between major subsystems, and then to use LIN for small local sub networks. For example, CAN will link the car's window controller module to the other major systems, but the window controller will then use LIN to talk to each of the individual window motors. This approach keeps low-priority local comms off the main CAN bus(es).
Interesting, with your setup you could simplify it with the following changes: Use the twisted pairs for the data Dedicated low voltage wires for the power All 4 chips on a single shield This would solve the interference issues, and allow you to bring power at the same time, though the main downside is that you would need a breakout system at the windows. My prefered setup would be to have 4 dedicated ethernet cables, which would plug into a single shield. (The connectors would be on the outside of the case, with wires connecting them to the shield, which makes it easy to rack mount the design in a 1U case.)
I LIKE this project. It does have many useful details and it is quite instructive. However, when you first mentioned the problem of the LIN bus(buss in some English variants), I immediately thought of the way that I might tackle the serial communications issue and my "answer" was significantly different than the architecture you used. Mine, too, would require some additional hardware, but would be a simpler hardware design (with a small compromise that shouldn't be an issue). You handled the mux/de-mux issue by using a separated serial path for each slave motor. However, one might continue to use the Arduino shield that used for your initial testing. The multiplexing and de-multiplexing could be handled with a single CD4066 analog switch IC to switch the DATA signal to the appropriate slave. A 4066 would handle 12V signal levels well. The control lines to each switch in the IC would have to be 0-12V, but a simple level translator ($1 on eBay) would handle level shifting. One would then enable each switch in turn to select the appropriate slave motor. The switch would have to be enabled for the interval needed to send a command and to await a response. So, sure enough, the Arduino sketch would be slightly more complex though : enable switch0 send open command to slave0 wait for "OK" disable switch0 enable switch1 send open command to slave1 wait for "OK" disable switch1 ... rinse and repeat. Total hardware cost? Perhaps $5 (includes the temp/humidity sensor), without consideration of the PCB cost. The interconnects would be much simpler. And, since you could use your favorite Arduino with Ethernet together with the LIN shield mounted - the extra 4 digital IO lines would be readily available to control the analog switch. You'd use no serial ports, so one would be available for use with the Arduino serial monitor, so it can be used for debugging though the entire design and debug process. Please don't take this comment as a criticism. I've learned over 45 years of hardware and software design that there are many ways to address a design and my own "back of the envelope" concept may have some warts. Dick (www.hardandsoftware.net)
I like that approach, it would have been a good alternative architecture. It would also lend itself to expanding easily to more channels with very small overhead for each additional channel. I did something quite similar to this in the Payload Processor Module that flew on ArduSat-X and ArduSat-1, with a pair of 16-way multiplexers allowing the supervising processor to communicate at will with any of the 16 slave processors that ran the experiments on the spacecraft.
Another, perhaps better(?), similar approach would be to control the 4066 - or multiple 4066's -- directly from the Arduino 0-5V DIO, and to use the bi-directional level shifters on the 12V data side. Though more commonly used for 3.3V to 5V applications, the level shifters would work fine for 5-12V conversion, and would have a much lower on resistance than the 4066 switch. Also, one might include an optional impedance matching network (if resistive, such as 75 ohms, we would have to boost the supply voltage from +12V to +24V to compensate for the network. But, this should reduce reflection, if that is the source of faulty signaling integrity. This might be tested on the bench before adding it to a design.
On second thought, this second alternative wouldn't work. The output of the LIN shield is 0-12V, so the analog switch would, perforce, have to use a 12V Vcc and switch level. The Rds on of the switch might upset the signal level at the slave receiver -- or, it might be fine. But, my first suggestion might be something to consider. Dick
I'm old-fashioned, I guess. So, the venerable 4066 came to mind first. It requires level shifting of the switch control lines, so that it can accommodate the 0-12V range of the LIN bus. However, a little study suggests that something like the Analog Devices ADG508 (www.analog.com/media/en/technical-documentation/data-sheets/ADG508A_509A.pdf) would be a better choice. It has TTL level control and 12V is well in range of its switching capability. In addition, it has a built-in 1 of 8 decoder, so 8 channels might be used with only three DIO controls from the Arduino. It does cost $7 each is small quantities, but this is still a reasonable choice, because it reduced the BOM to a single IC to mount on the PCB. Ron is comparable to that of the 4066, so that parameter needs some thought.
Great video as always! I just had a question about the wake up sequence. Can you turn this off? Sounds undesirable in the event of a power outage for security and noise reasons. Your scenario about the windows bashing in the middle of the night when windy might move to windows opening and shutting noisily in the middle of the night if the power drops.
That's a great question. To be honest I'm confused about exactly when it does that sequence, because I haven't seen the motors do it recently. I *think* it may only do it as a self-test if it's powered up without any remotes bound to it, but after it has been bound it assumes it's OK and doesn't do it again. It would definitely be a security problem otherwise.
Nice project! Maybe it was also possible with 1 arduino and 1 lin module and 5 relays.Control the relays with the same arduino. Use each relay to select the window motor (data line) and the rest as in the original control. So first you set the window you want to control and then send the command via arduino+lin module. Use the 5th relay to connect al datalines together and then you have control over the 4 windows in parallel.
SuperHouseTV I also thought you would use some sort of multiplexer IC when you told about the 4 wire solution... I think that would have been easier and more cost effective.
Great video. Those AXA units retail for about ~£120 in Uk. I suspect you got them free or in a semi-free deal? I think a much better option is using the ACK4 230v Electric Window Opener. What I have. Costs about £120 here in Uk. Use with relay and ESP8266.
John. I would think you would make a new single shield with all four ports on one board. Add some dc protection and some PTC’s. I would not trust the POE switch to manage any fault current.
Wouldn't it be more reliable to use 1 twisted pair for each of your 4 data lines and then use the 240v cables you have in the frames for running distributing the 12v? You could fix up a little connector in each empty battery box to translate the power and data into an RJ12 plug. That way no cross talk as each data connection would be on a pair.
Yes, that's a good idea, but I decided not to do it because it's very tight in the motor case and I didn't want to fiddle with breaking out the big flex cable. I'm also paranoid about using mains cable for anything except mains, because some time in the future when I'm long gone someone may find that cable in the ceiling and think it's meant to be at 240V AC. However, now that I'm going to follow this up with another video showing the crosstalk on the wires with a scope I think it's worth going to the trouble of doing that to see how much effect it has.
Is changing the firmware to allow individual communication to the motors out of the question or not possible(Similar to what you did with the sonoff)? It really seemed like daisy chaining the motors was a much more elegant setup.
I haven't pulled a motor apart to investigate the controller, but there's a possibility that they could be reflashed. I only have 4 of the motors at the moment and they're all mounted on windows, so I'd need to get another motor to experiment with.
Cool!! do understand correctly that the windows do not open automatically, they are released but as the chain is slack, you have to hope for the wind to open them, correct?
The chain has a cool design that's like a bike chain, but with little tabs on one side so that it can only bend one way. This has the effect of preventing it flopping around when the window is open.
I have noticed the AXA 2.0 is still available now. I am thinking of doing this and integrating to Home Assistant via MQTT. Do you know if the LINBUS adaptors are still available to connect to an ESP or Arduino? Did you fully solve the cross talk issue with the firmware? I was thinking maybe if you used all the "white" cables for bus and the solid colours for power (or vice versa)? Would having power wrapped around data cause an issue?
Great question, I often have people ask me that! Here's an explanation I prepared earlier :-) www.superhouse.tv/15-watchdog-timers-for-arduino-home-automation
I watched the video. The only reason I can see that you don't use the integrated hardware watchdog (the atmega's watchdog has its own dedicated oscillator and everything, but its basic pat or reset feature maxes out at 8 seconds) is that you need a long timeout and don't want to rely on any software solution to extend it. Is that correct?
That's about right, yes. Also, I don't trust software watchdogs to do the job as cleanly as actively asserting the reset line and forcing the whole MCU to do a full reboot. Software reset is *supposed* to do the same thing, but I have painful memories of doing software resets decades ago and having processors coming back in an unclean state with memory un-zeroed, etc. Since then I've had a strong distrust of software resets. It's probably better now, but the paranoia is still there.
Understandable, but as far as the atmega328x and atmega25xx series go. Up to 8s can be achieved completely in isolated on-board hardware. It is a completely isolated external yet on-die watchdog circuit. Very reliable in my experience. I love your videos I've watched nearly the whole series.
SuperHouseTV iirc one of Toyota's issues was a watchdog asleep at the switch when the throttle control glitched. Somewhere on the interwebs is a pdf independent software review on the portion of the code Toyota was forced to cough up.
Are the supplied batteries rechargeable? I wonder if it draws current because it has a built-in battery charger - so you could open the windows of the power is out e.g. in a fire or earthquake etc.
That would be a nice feature, but unfortunately it's not that smart. The batteries are regular Alkaline cells, non-rechargable, and the instruction booklet specifically says to remove the batteries if LIN is used.
Scott Hanson That was my original plan, but I don't think the software serial library supports 2 stop bits. It could probably be modified to do it, but I wanted to make quick progress so I used a Mega because it avoided that whole problem. That was one of the design decisions that I didn't cover in the video
Hi great video! Very interesting and informative. I've been struggling with the AXA motors myself. Especially since there is hardly any info about the 'home automation' part of these motors, whilst AXA does put 'very suited for home automation' in its brochures... I've even contacted AXA directly (I'm from holland), with no real results. I really want the motors to be addressed individually! I'm also in the possession of the linbus shield, but never installed it once I figured out that it is not possible to address the motors individually. Your video is a new eye opener how ever! What I've done is that I've hacked into the remote and routed additional IR leds to each motor, using transistors to control each motor separately. All connected to an arduino interfacing with domoticz. I'm currently adding some ultrasonic sensors to control the position in more detail and get feedback of the current position. Now You've got me wondering, I might be crazy, but I don't have all this knowledge so I'm just asking... Wouldn't it be possible somehow interrupt the linbus signal to all motors but 1, thus being able to control the motors individually. This would mean however that the linbus signal has to be able to pass thru a relay or transistor or what so ever, this I do not know, neither could I find any information about it. Could you give any comments on this idea please? Keep up the cool videos! Also really interested in your openhab setup! Cheers Sebastiaan
That's a great solution! I love the idea of using ultrasonic sensors to get feedback on the window position. I may give that a go. I'm definitely going to follow up on this video later after I've made an integrated controller. I'd like to get an answer on the addressing issue, too. Without any other information from AXA I think it'll be necessary to watch the data line with a scope to see if there's any address information being sent. I don't think there is.
What motors/equipment are you using to control the Shades in the conservatory? I’ve seen there is AXIS (Coming soon) and MOVEZ (Coming soon) but wanted to know which ones you used and I couldn’t see a post / reference to it elsewhere. Thanks.
Thanks for clarifying thongs as flip flops and not underwear.... My mind went to a scary place for a second. joking aside, dude I love this channel. I've been following for a while, now that I bought a house, I can start doing projects myself.
Hi watched a sonoff video and lo your channel popped up in my feed so have watched several of yours now......Google knows! I have similar windows (woooden noy metal) to you so I enjoyed this though (a) finance committee would not pass the budget (b) My skillset and toolbox in accountancy wouldn't fit well with going the whole way down this project and (c) How does your system handle sticky windows what is the torque? Finally I use Visual Basic code in a couple of my work functions and are surprised by the crossover in code syntax between what i use and what you used. Anyway looking at roller blinds and possibly using hacked RF toy car motors like the Soma Smart Shades - Soma Smart Home but Amazon dont ship and nothings cheap in Australia...maybe a project? Again Torque is an issue as I want one turns the blind cord not the tube itself also solar.
Could the firmware on the motors not be modified to include device IDs for the motors themselves? Or is that level of hacking a few levels beyond the skillset? I'm honestly not asking to be rude, quite the contrary. I am interested in the process of this should one undertake such a task. It seems that your problems would be solved if this could be done. Maybe even reach out to the manufacturer for a copy of the code (even under dev NDA or something)? My interest is starting to be piqued with home automation and getting ways to also include fire and intrusion detection ;)
Addressing them individually would be the best outcome, but I only have 4 of the motors and they're all attached to windows! If I can get hold of another motor (maybe a broken one) I'll pull it apart to see how they're controlling it. I did consider removing their entire control board and making my own to swap into their chassis so that I'd have total control over the firmware, but I have other projects to work on first.
Interesting. So, getting and hacking firmware isn't as trivial as it seems and making a custom board requires a fair bit or R&D. Any teasers for upcoming projects? :P
Modifying the firmware would require either the cooperation of the manufacturer (to provide the original source code) or writing totally new firmware from scratch. To do either of those things I'd need to learn about the specific processor used in the controller, such as what registers it uses, what timers, interrupts, etc. It's not a small task! There's a list of possible future topics here: www.patreon.com/posts/8033622
It seems strange that the manufacturer would go to the trouble of implementing LINbus but hobble it by not supporting device addressing. Maybe it was the cost of adding a DIP switch or somesuch for setting the device address is what killed it in the production model.
Hah, that's funny, I should pay more attention to my backgrounds! That's parts for an Arduino compatible kit that I make called the "KitTen": www.freetronics.com.au/kitten. There's a long story behind the name. I used to make an Arduino compatible board called the TwentyTen ("Duemiladieci") which was released in 2010 (hence the name) and designed to be a successor to the Arduino "Duemilanove" (which means "2009" in Italian) way back before the Uno was released. Then I deleted the USB-to-serial converter, changed all the parts to through-hole for easier assembly, and created a "kit" version of the TwentyTen. Hence "KitTen" with the punny name. Incidentally, that's also how the EtherTen was named. It began as a TwentyTen, then I added Ethernet, and EtherTen seemed like a good name.
Thanks for the explanation :) May I take the opportunity to thank you for your videos - they are a constant inspiration to me in my home automation projects. Keep up the good work!
Thanks Ed19601 for adding the link. There's also a list of parts with links on the project page on the SuperHouse site. The video description has more information.
What kind of tablet is that in the intro? I've been considering doing something like that with some Amazon Fire Tablets or whatever they are called (Amazon's 50 dollar tablet)
It was some cheapie Android tablet that I got in China quite some years ago. No brand, just something found in a market in Shenzhen. I also have some Microsoft Surface tablets on the wall, which have fantastic hardware (great screens) but not so nice software.
The mounting board was some random offcut that I pulled out of my pile of plastic that I use for the laser cutter. I don't know the source, sorry. Most hardware stores will sell something similar. The RJ45 breakout I used is this one: www.jaycar.com.au/rj45-side-entry-surface-mount-8p8c/p/YT6064. It's handy because it has screw terminals on it, but you could also use something like a regular RJ45 keystone socket, crimp some UTP cable into it, and solder the other end of the UTP to whatever you like.
Thanks, perfect. I've been looking for a screw terminal RJ45 connector for a while. Cheers! Keep up the awesome videos. PS: Do you do any paid board design? We are building a product at the moment and may need a bit of help designing a shield.
For anyone searching, the "Arduino Uno R3 Shield for HOME-LINBUS Devices Axa Remote 2.0 Window Opener" is available on eBay at www.ebay.nl/itm/Arduino-Uno-R3-Shield-for-HOME-LINBUS-Devices-Axa-Remote-2-0-Window-Opener/131524803718
Please, please, please get a video camera with face recognition. It's really difficult to watch your videos with your face slightly out of focus and the shelves in the background in focus. I find after about 5 minutes I have to look away, as I get a headache from my eyes trying to keep getting you in focus. I've watched a lot of your videos and the content is fantastic, really interesting and well presented.
I've included links to parts sources on the episode page, and I've started writing some supporting documentation that will be added to the page soon. If you're only watching this on RU-vid you won't have all the extra information to go with the episode. There's a link to the episode page in the video description.
Awesome, thank you for the swift response. I really enjoy keeping up to date with the SuperHouse. I personally would like to see smaller, multi-part series. Shorter videos more frequently, keeps fans more engaged in my opinion. By no means am i asking for this, just letting you know what a fan of yours prefers. Keep up the awesome content, thank you!
Thanks Sam. I'd love to have a more regular release schedule, and also to do more multi-part vids. Epic videos like this cause big gaps which is really frustrating both for me and for viewers. I'm finding that the process of filming and editing is a very big time investment, so I'm trying to figure out ways to make the process easier. Every step in the process is an extra hurdle that slows me down. For example, audio is a huge problem. I haven't been able to use an on-camera microphone because I've been using ancient video cameras with tapes, and the motor that turns the tape is so noisy that it can be picked up by the microphone. That means I have to wire myself up with a radio mic or a wired lapel mic every time I want to record, which means I can't just grab the camera and turn it on, I have to do a bunch of messing around with cables first. That's just one example. Each little speed bump that makes the recording and editing process more difficult may seem trivial, but when they all add up the result is that it takes me much longer to create a video than it should. Recently I've been trying to find ways to simplify the recording and editing process so that I can get out more content. This particular episode was made using 4 different cameras, 3 microphones, and 2 different editing packages while I experimented with ways to make it easier.
I can’t believe this, because we have blinds and fly screens it’s a hassle to wind the windows in and out so this might be my solution, my only concern is how long will the battery last
I really like your vids. How can I get involved to help your initiative grow?" :) Can you confirm that the LINBUS will allow for the definite state of the device to be known by the arduino controller? Have you tried the broadlink RM Pro with the axa window remote to verify its codes can be learnt? No wires that way no need for LINBUS...?. Im really astonished how many devices Ive been able to control with my broadlink. You can literally control any rf and ir based device via your phone, tablet and with the addition of some of the open source based android apps (RM Tasker plugin and Tasker) and an echo dot you have a way for Alexa to trigger any of the codes the broadlink has learnt so the devices respond as if their real remote button was pressed. The issue I have however is no state is maintatined and I am thnking of how I could somehow keep state stored via MQTT but how can it be definitely in that state? Hence the Q re LINBUS. the broadlink devices cost around 30-40 bucks they learn rf and ir commands. Another application on google play is RM Bridge which can be used to allow for a configured samsung smart things hub to have each device defined and controlled by Smart Things. I think RM Tasker uses the open source based Hue bridge emulator as when the rm Tasker plugin Alexa bridge is active the echo dot "Alexa" shows each of the devices setup in the RM Tasker plugin to be "Alexa devices" as hue bulbs. As a result I've been able to control pretty much every device I have with my echo dot :) The RM Tasker plugin also has an http server. this allows for any learnt IR and RF codes learnt on the broadlink to be called by making a simple http call for any of the devices. The rf range of the broadlink is very good. I've managed to get it working with logitech harmony as well through the http approach and a bit of a hack with the logitech harmony using applesscript.
I bought an RM Pro a few weeks ago, and it's been sitting in its box next to my keyboard while I tried to get this video out of the way. This episode has taken months and months to finish, so I didn't want to distract myself with a cool new toy! Yes, with LIN it's possible to query a connected AXA motor to get its status. In the video I showed how I used my laptop to send a command of "VERSION" which caused the motor to respond with its firmware version. In a similar way, there is a command called "STATUS" which causes the motor to respond with whether it's opened, closed, etc. The response statuses are very basic and seem to be oriented around the lock status, not the window position. I can't remember the exact messages, but if the open or opening it will say "UNLOCKED" or something like that. So you know it's not closed and locked, but you don't know if it's still in the process of opening or if it's already fully open. When it closes, it will say "SOFT LOCK" or "HARD LOCK" depending on whether the secondary lock has been engaged. Something like that, anyway. I'm going from memory here. But yes, you can get some status information from it.
SuperHouseTV thanks for getting back so quickly and I'm going to have a closer look at Linbus now. Where abouts are you based in Australia? Greetings from Waterloo/NSW :-) Do you hold any user group meetings for others in this space to meet and share their ideas? I'd be keen to come along if there is one taking place soon. Thx Paul
I'm in Melbourne. I don't know of any meetups specifically related to home automation, but my friend Justin runs the Internet Of Things Sydney Meetup which has lots of crossover into home automation. Perhaps that would be interesting: www.meetup.com/en-AU/Internet-of-Things-Sydney/
That is weird! If you cannot address the motors on the bus, it really does not make any sense to even have a LIN bus, I can simple do that with two wires and a DPDT relay (that's what I did in my green house). The people at AXA have to do some work to get this right.
On LINbus, any slave can assert for the master's attention, so while slaves cannot start communications, it can trigger the master's poll. There are a bunch of automotive AVR's like an ATtiny87 with direct LIN support, and even some "hybrids" from Atmel with LIN drivers. They are probably a better option for custom adapters. Perhaps you can modify the LINbus ID on the window winders so that they can operate individually on a bus. Or was that not an option? What? No manual?
Oh nice, I haven't looked at the ATtiny87! (Runs away to research it...) Perhaps the IDs can be changed on the motors, but I have no information about them other than how to mount them and plug them in. They come with no documentation about LIN at all. I figured it out by looking at how the HOME-LINBUS shield works, and then poking at it using the MCP2003.
The MCP2003 is significantly only the transceiver. LINbus frames need special feeding and care. ;-) The ATtiny87 does not contain bus-level drivers; only the LIN "engine". You need to add the bus-level stuff. If that's all you need at the "appliance end", then there's e.g. an ATA6625 that does the bus driving as well as provided some regulated voltage to your small circuit from Vbat. Prefer a kitchen sink? ATA6616C which put an ATtiny87 core into the same box as a regulator and transceiver - all in a tiny (not for soldering by quaking fingers) QFN38 package. Specially targeted to automotive applications. Alas; not easy to find in the usual distribution channels, unless you want to buy a reel of 4000. :-(
Could you do a teardown of the AXA module and maybe reprogram the modules with individual LIN addresses that way you could use a single bus to communicate with each unit individually? Isn't this what the manufacturer should have done to begin with? Thanks for the videos :-)
I'd like to do that if I can get another motor. I only have 4, and they're all attached to windows now. I might see if I can get a broken one from the importer so I can pull it apart.
Given that each device on the bus reports a 200 OK, wouldn't a more elegent solution to sending a command to all 4 devices on the bus be to wait for the 200 OK before moving on to the next device, Instead of using an arbitrary delay? This way you are guaranteeing receipt instead of assuming it.
Yeah to a certain degree pardon the pun. It a glass house buddy. Low-E helps but there is only so much you can do.... double, triple glazing , or how ever many is only effective as the size of the air gap. Timber windows are more efficient as they dont are not thermal conductors. Id rather have airflow and automation.
So I totally appreciate this guy is very intelligent and smart but I completely disagree with his philosophy. Adapter boards were great, but the he adds all kinds of complexity just to have wired control. I was him I would have just slapped a couple of esp8266s to the motors and called it a day.
Do you mean like this? github.com/SuperHouse/AXAWIFIESP01 Or maybe this? github.com/SuperHouse/AXAWIFI How much work do you think it is to just "slap a couple of ESP8266s in it and call it a day"? What's your threat modelling for controlling physical access to your building over WiFi?
Nice watch, thanks. So, this is where I can see that I'm much more lazy than you. Once I get v0.9 working, I call it done, and move on. Yes, a USPS cardboard box cut in a few pieces with otherwise exposed everything is a fine production system :) ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Qvsb9gRpLmw.html from marc.merlins.org/perso/linuxha/post_2013-01-30_2-way-talking-to-a-Lockstate-LS-DB-500R-lock-because-the-Lockstate-wifi-Remotelock-is-no-good-for-me.html And of course, I have professional I/O mapping on a yellow sticky :) By the way, I do agree about your rule on crawl spaces and going to the roof. The crawlspace is bad (and mine is clean), laying all my 1wire temp sensors in the vent holes and inside ducts, was a pain. My attic is usually super hot and full of insulation stuff that leaves fibers in the air that really you shouldn't breathe for long (if at all). I use a dust mask, but it only works so well. So yeah for everything going back to an accessible patch panel!