Figured it out. I had to send HVAC_FLOW_STATE to HA, in order for it to figure out that it was a "thermostat". Only problem is that it shows up with states "Auto" "Cool" "Heat" and "Off" in the dashboard, even though I only have the possibility to heat, or turn it off.
I have started digging into home assistant again. I've been using node-red for the last 7-8 years, for all my automations.
I have one mysensors node setup as a thermostat (HVAC) controlling my floor heating. I have it working in a node-red dashboard, but would like to have it in HA as well.
When I present the child as a S_HVAC type sensor, it is not shown in the entity overview of HA. Only the childs defined as S_TEMP, but those I can't change from HA.
So is there anyone that has a working thermostat example for mysensors in HA?
I deployed 5 sensebender nodes just after release in 2015, all with 2 AA batteries from ikea.1 died within a couple of years, because it was placed outdoors. But the rest kept going, and all have been going until this year. So 8 years of battery time on one set of batteries..
I think that is excellent battery life.. 
@Yoshu you could boot linux off a usb, and check there..
@Yoshu Does the same device work on another machine, running win10 / linux?
Unfortunately I do not have any hardware compatible with win11, so can't test myself.. But if I remember right it should turn up as a generic cdc device..
Is there any info in the windows device manager?
@TheoL said in Merry X-mas:
Hey all,
I wanna wish you all the best X-mas - I know it's a little but early - and I wish you all some awesome projects in 2022. I hope we can get a creative spirit in 2022 and create cool things together.
Merry x-mas, and (an early) happy new year to you too 
@keithellis
They should be available here https://itead.cc/product/mysensors-micro/
@monte said in Best PC platform for running Esxi/Docker at home?:
@tbowmo I don't know your setup, but optimizing the code that serves frames to the detector can help more then you may think.
I'm using zoneminder, and zmeventserver for this.. All running in docker at the moment.. Haven't had the time for digging in to this machine learning, other than scratch the surface, and trying to apply prior work to my own setup..
@monte currently I'm at the order of 0.2fps, would be nice with a bit more throughput, as I'm planning for more cameras along the way. (currently I have 2 cameras running 24/7,but have 2 more that I just need to put up..)
@NeverDie i prefer to keep video footage local. Not saying that I need an rtx3080/3090. They're too pricey. Looking more towards 1050ti or the like. Should be enough, if I'm not going for coral tpu. Waiting for a friend to get hold on some tpus, and give his verdict.
I've been looking at building a new server myself.. Currently I have an old hp desktop machine with an I5 / 16Gb of ram, running as server.. Works fine (and it was cheap). But I want to play more with docker / kubernetes, and perhaps do some more machinelearning on zoneminder images, so could use a cpu with a bit more oompf.. Perhaps an nvidia gfx for the ML part..
Only problem is that I'm not ready to spend 1000$ or more on hardware, so I'll probably just wait for the prices to drop again..
(Currently I just run ubuntu server, and then docker on top of that.. No VMs or the like..)
Perhaps my definition about "controller", is a bit broad here.
What could be done, is to have a thin "controller", that receives MQTT from the existing gateway, hands out IDs if they are requested from a new node, and translate the message into homie compliant topics. That is all it has to do, should not care about the rest of the automation at all. Should also be able to translate back from homie compliant to mysensors.
This could be written in python, c#, javascript, rust etc. The benefit is that you have (relatively) easy editing of configurations at hands, without re-uploading sketches to your gateways / nodes, if you decide that node X should be named Y instead.
And I am in no way trying to treat anyone as a fool!
@Peter-Loeffler
I still see this as a "controller" part, as it's mqtt you are totally free to subscribe to the messages from the gateway, transform it to something else, and then publish that onto mqtt as well.
That's the way I would do it, as it's much easier to edit config files etc on a computer (rpi etc).
My reaction was mainly towards the part of your statement "very snobbish and autocratic".
I haven't seen a response about paying someone to do a feature request (I've been away from the forum, so maybe I have missed something?).
You are absolutely free to create a PR for any feature requests. We might fight back a bit but we need to have a good use case for the features that make it into the core of the project, otherwise we get the project filled with features that is only used by 1 or 2 persons..
You still need to tell the gateway "This node is 'living-room' and it transmit 'some-odd-property'" This needs to be configured somewhere, that be the node itself (and transmitted over the air/wire), or the gateway (or a computer running some software).
The current iteration of mysensors let the gateway run on an UNO. So that is the lowest common part that the gateway firmware needs to be compatible with (at the moment), and that mcu is too small in supporting the above schemes. Let alone, have all in the different sensor types programmed as text strings.
It might be doable in some future version, that is only in the talks (We've had some brainstorming on v3 of the protocol a couple of times).
There are many examples where of the shelf equipment downstairs make the cut. Of course if you only want temperature, or door switches, then zigbee devices are easier to implement. But if you want specialized Sensors, for measuring depth of water in a well, automate an existing garage door, automated the chicken coop, etc. Then mysensors is there to the rescue. Because you can't find any of the shelf non cloud solutions for that.
Perhaps we are behaving like old farts in the core , but it's not intentionally. As mentioned a lot of times, we are using our spare time for the project, but the spare time is limited..
The mqtt / serial protocol adhere very closely to the OTA protocol, where we are using byte sized nodeIds. This keeps the overall transmission payload down, and limits the time needed for transmitting a sensor value, and also keeps the likelyhood of errors down (Larger payloads = larger chance of failure). And currently chosen radios have a limited payload length
If we should implement the homie-convention (see below for an example) we need to be able to tell the gateway that nodeId 10 is device123, and childSensorId 1 is myThermostat. Then the gateway needs to store that info somewhere in a lookup table, to be able to do the conversion on the fly, as it's not feasible to send that info over the air for each transmission.
Also, we support the use of the same sketch on multiple devices with the current nodeId assignment. IE when you have 10 temperature sensors that all use the same sketch. The nodes get a nodeId from the controller, that it stored in eeprom, and re-used. Even if you re-program the device.
Homie-convention examples:
homie / device123 / $homie → 3.0
homie / device123 / $name → My device
homie / device123 / $state → ready
homie / device123 / $nodes → mythermostat
homie / device123 / mythermostat / $name → My thermostat
homie / device123 / mythermostat / $properties → temperature
homie / device123 / mythermostat / temperature → 22
homie / device123 / mythermostat / temperature / $name → Temperature
homie / device123 / mythermostat / temperature / $unit → °C
homie / device123 / mythermostat / temperature / $datatype → integer
homie / device123 / mythermostat / temperature / $settable → true
I'm not sure how to address your statement. But beeing one of the core group members myself, I feel that it's a bit harsh. We all have a life, and fulltime job, besides mysensors! This means that we can't spend every minute of our (spare) time on mysensors all the time. Some of us have scaled down on activities for a period, due to children, work, house renovations etc.
Maybe we (the core group) say no to some new features in the project, but that is because we are uncertain about the following support / bugfixing of that particular feature. Specially in the light of the above, with limited spare time for the projects.
For my part, I've had a couple of rough years, that took the steam out of my envolvement in a lot of projects, including MySensors. But I'm slowly re-emerging from the deep dark now. So I hope that I get more time to do new fun sensors / hardware again. But things take time, and do not want to rush things, that could send me back into the dark again.
I admit that I have been pretty silent the last couple of years on the forum. But daytime work have drained my energy reserves. I have however just picked up mysensors again, and started jumping into platform Io for arduino. And converted one of my old mysensors sketches.
I hope that I can return to the community again, still have some fun hardware projects planned, but also need to constrain myself, so I do not have to many projects at the same time, otherwise my energy is depleted completely
/ Thomas
In my opinion it's not feasible to do this on memory constrained devices, we should keep the topic structure as simple as possible. If we implement the homie-conventions we need to keep all the sensor names etc as string values in the code, in order to create a proper topic.
As an alternative, you could use nodered to transform topics to another format. I'm doing that in my setup, so I have the following topic structure:
house/<room>/<device>/<sensor>
to set a value to a device I have the same topic as above, with /set appended.
When I said that you needed the package "build-essentials" I assumed that you where running on a barebone ubuntu distribution, and not in docker. This complicates things, but it seems that others do have the same issue with node-red and sqlite on home-assistant forum.
One solution is given here: https://community.home-assistant.io/t/how-to-get-access-to-sqlite-on-hassio/241439/3
As said earlier I do not know anything about the internals in home-assistant and their specific docker / node-red setup.
What I did way back when I got it up and running in docker, was to jump into the container shell, and install the package manually (So you are using the same node.js version that node-red is using).
You could also (in that docker shell) do a apt install (given that it builds upon a debian based image).
something in the line of:
$ docker exec -it <id of container running node-red> /bin/bash
# now you are in docker..
$ apt install build-essentials
$ cd <to your node-red data directory>
$ npm install node-red-contrib-mysensors
$ exit
Above is taken purely from what I think is the right commands, as I have not tried working with home-assistant for a while.
NPM is trying to build sqlite, because it cannot find a suitable pre-built one for your architecture, it fails because you do not have the needed tools installed for this on your machine.
Might help if you install build-essentials (sudo apt-get install build-essentials, assuming you are on a debian like install)
I currently use it on node-red 1.1.2, without problems. But I'm using plain node-red in a docker container..
Just heard they put one of the vaccines, astrazeneca, on hold here in Denmark, as there apparently where reported some serious health issues with it (someone apparently died after getting it). My wife was in line for getting one of them at her job.
For my part, I think I'll probably be working from home, until after the summer vacation, as my employer has asked us to stay at home. But I do miss the daily walk and talks to the coffee pot with the colleagues. Our sitting together when looking at code issues.
/ Thomas
Unfortunately we do not have any other manufacturing partners, so all sensebender boards are produced by itead.
How did you confirm that the board where broken? Have you tried starting the self test in the standard firmware that the micro is delivered with?
The factory firmware is available at
https://github.com/mysensors/SensebenderMicro/blob/master/Arduino/SensebenderMicro/SensebenderMicro.ino
Had a look at our itead dashboard, total number of ordered units until today is:
Sensebender Micro: 3575
Sensebender GW: 587
which, if I remember the figures right, has given the project something in the order of 5330$ (before paypal get their cut).
PayPal has certainly earned their share on mysensors as well, every 3 months we get commissions from the sensebender sales at itead (sensebender micro and gw) the payments are split in two, 50% to the project, and 50% to the creator (me). Both paying fees to PayPal, for receiving the money.
Sensebender sales are also slowly decreasing, but the micro is also 5 years old now. (they stated shipping the first batch end of May 2015). I never thought that we could sell the amount that we did (I'll have to look at the numbers when I'm at my pc tomorrow).. To be updated
/Thomas
@alex28 It's always the complete product that you need to do certification on. It will probably be easier with a complete module, as the rfm69, but a lot of things can change the RF characteristics. For example the ground plane layout on your own board, chosen MCU, xtal frequency, decoupling etc. and firmware...
Just made a fun (and useful) sensor / device, using a arduino pro-micro (atmega32u4), and VL53L01 TOF distance sensor.
It's an auto lock device for my pc, I have mounted the VL53L01 to the bottom of my center monitor, and then have it measure the distance to my body, whenever it exceeds 1m (or is out of range), I increase a counter. When the counter reaches 20 (seconds) I send keypress GUI+L, to lock the screen (works equally well on ubuntu and windows).
When I return to the desk again, it is detected by the arduino (distance is now under 1m again). And it then sends CTRL+ALT+DEL to start login procedure.
The arduino sketch is available on https://gist.github.com/tbowmo/7e9934796d47566dc09e7b3bc5b2f208
next project should probably be to find a better enclosure, and build one for when I return to the office in a month or two, when the corona lockdown is lifted more
Took a quick look, and one thing I noticed, is that you switched around the names for ftdi / isp pin headers.
That is, you have MISO / MOSI etc. routed to the FTDI header, instead of the ISP header.
@echi We are just pointing out potential pitfalls if you are going to sell your pcb pre-assembled to the masses..
That is, if you are selling a PCB with radio on, then you are obligated to follow the rules and get the radio verified by a notified body.
@echi for my part I'm not concerned about it, because I'm not the designer/seller of those modules. So if the notified bodies find that they are not compliant, then it's not me that they're gonna bug about it.
It's all about responsibilities. I'm not making a radio module, so I'm not responsible for the rf part, and as such, they can't sue me for that part.
/ Thomas
@echi the thing is, that if you design/sell a board with an radio incorporated, then you are obligated to get fcc certification (for the US), and other approvals for the rest of the world, as @bjacobse explains.
But if you do not include the radio module, then it's not your problem if an end user use a cheap Chinese radio module together with your product.
@echi nice product, but what about rf conformity of the radio? Is there any fcc / ce approval? It's not enough to just slam a radio chip on a board and sell it (unless you are a cheap Chinese eBay seller ;))
That is one of the main reasons to why I did not incorporate the radio module with sensebender micro, as I do not need to make any radio measurements.
/Thomas
@paqor said in Ceech Board MOSFET Pin DIGITALWRITE Problem:
You have to set pinMode for the same pin as you want to use as a output.. That is, if you use pinMode(5, OUTPUT); you also need to use digitalWrite(5, HIGH); See documentation here
If you do not set pinMode for a particular pin, then that pin will be set as input by default. And then digitalWrite() will enable / disable the internal pullup on that particular pin.
I also suggest that you use a #define to make sensible names for your pins, that will make your code more readable:
#define RELAY_PIN 5
void setup() {
pinMode(RELAY_PIN, OUTPUT); // Setzt den Digitalpin 13 als Outputpin
}
void loop() {
digitalWrite(RELAY_PIN, HIGH); // Setzt den Digitalpin 13 auf HIGH = "Ein"
delay(1000); // Wartet eine Sekunde
digitalWrite(RELAY_PIN, LOW); // Setzt den Digitalpin 13 auf LOW = "Aus"
delay(1000); // Wartet eine Sekunde
}
There is actually two "products" that are sold.. Not complete with plastic box etc. And you have to solder a socket for the radio, and add the radio yourself, as well as do a bit of programming, but the SMD parts are already mounted. I think that this is the closest that you come to a pre-assembled unit these days.
Sensebender gateway - Can be bought from Itead
Sensebender micro - Can be bought from Itead
I don't know if any of you have seen this before:
https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6
@bjacobse thanks, hope so too.
Just looked back into the thread, I posted a picture of my old workshop back in 2016, still the same tools around, just a smaller desk to work on.
/ Thomas
@monte said in Your workshop :
@tbowmo is that KDE?
Nope, Ubuntu / gnome, but with a tool bar at the bottom of the screen.
resurrecting this old thread
I got thrown out of my old mancave, and in to a small corner in the main building (the oldest son, aged 21, needed a place to stay, as his 19 year old sister didn't want to share room with him)
Anyways, I have moved from a 240x60cm desk to a 120x80 desk, and no storage space around me, other than a small office cabinet on wheels. Which means that I have to constrain myself, and find some alternative setups. As I have to work from home these days due to corona, I treated myself with a monitor mount for my 3 screens.
As a bonus, I can now use the space below the monitors for my tools (power supply and soldering iron,and probably the oscilloscope as well, when I get to it). I'm re-arranging stuff every other day, to figure out the best setup, and remove cable clutter (which isn't easy).
Finally I am starting to do a little electronics again.. First thing is a prototype assembly of RASLE (Rpi Arcade Sound and Light Extension). It's a custom made arduino "coprocessor" for a raspberry pi, built into retropie arcade cabinets. It's a joint project with a couple of friends that are building arcade cabinets (I built mine a couple of years ago, I think that there is pictures earlier in this thread).
Features:
Here in DK, we started with closing down all public schools, and educational institutions 2 weeks ago. For my part, I have been working at my home office for the last 2½ weeks.
The government is taking a more and more strict approach to the current situation, limiting the freedom of movement, the latest is that (unless you'r a family living together) you are only allowed to be 2 persons in the same area (taking a walk etc.). And the schools etc. are closed at least until 14th of april (after the Easter holidays in denmark)
Luckily I live in the countryside, so we have lot's of space around us, with forrests etc. So we can get out in the fresh spring air, almost without meeting other people, or if we do meet other people, we can keep a safety distance of 2m. This also means that our local grocery store is not overly crowded with people, when we do shopping, as opposed to the shops in bigger cities.
As a kind of bonus, I get the opportunity to order some gadgets for my home office, now that I work full time here. "But honey, this thing will unclutter my desktop / make it easier for me to switch between work and hobby".." 
I was forced over to react, as the new project I was assigned to is using it (one of the downsides / luxuries of being in a large corporation). But yes, "Another framework". I can't decide if react is better than angular though. It's not as strict as angular on how to design things, but then the "strictiness" could be one of the strengths of angular.. I have experienced it a couple of times, where there are a lot of different ways to accomplish the same thing in react, so it's a bit harder to find the "Best way (tm)", whereas for angular it's a bit more strict path to implementations.
for the websockets, there might be security issues with it but if you want to have an open connection where you push data to the client, you don't have that many options available. I have thought about ditching mqtt for the frontend, and implement my own (websocket) protocol. Mainly to limit access to my mqtt broker, but there is only so many hours a day, and work, wife, kids etc. all want a part of them..
The project is already on github, don't expect anything fancy though..
Perhaps, but we are a large household (6 persons, where 4 is above 18), so laundry has to be done on a regular basis, and just filling the baskets isn't an option.
That's an interesting chip indeed.. And could also prove useful (how filled is the basket at current time) and collect it in influxdb for some nice graphing Could also be used above the table to detect if excessive amounts of clean laundry has been put there, instead of being folded, and distributed to the different persons in the house..
Just need to make something durable so that the sensor isn't destroyed if someone decides to overfill the basket, and force the drawer in.
a couple of sensors put into the buy now on ebay
My plan is to disable internet access for user and guest vlans (the complete vlan, both wifi and wired). I think that it's our 21year old son that is most likely to fill up the laundry basket, he also use the most of his free time with online gaming, and his room is next to the laundry room. I think that he will be pretty quick figuring out that laundry piles will block his internet
Actually he volunteered to be on the laundry team, last time we had a "planning meeting" where we had the different tasks in the household up for discussion. The only issue is that I think that he forgot about this.
anyways, started out ordering some infrared beam sensors (adafruit ADA2168), that I can play around with..
Strain gauge could be an option, but is probably a bit over the top, so much that I won't be coming around to it before the kids move out
The laundry is their own problem, Neither I or my wife enter their rooms to pick up their dirty laundry, they are adults now, so they have to figure things out by themselves (their younger siblings is another story, they are 10 and 8, so we still cater a bit for them).
The whole contraption can be hidden inside the drawer, the only thing that is visible is a patch of some sort of reflective material on the inside of the drawer front, and a box in the back of the cupboard / closet (the laundry sorting station is made up of ikea kitchen elements, with large full height drawers..)
I just came up with an idea for a sensor.. In our house, we have 3 large drawers for dirty laundry, with sorting baskets inside. One for White, one for red and one for other colors.
The problem is that some people in the household (mainly the two eldest kids, aged 19 and 21), doesn't know when it's time to do laundry, they fill up one of the baskets to the brim, and beyond, even though they know how to start the washing machine. My idea is to add a small sensor inside the drawer cabinet, that detects if the basket is full and then cuts internet to the whole house, until it's fixed, it should only take 1 or 2 attempts before the said persons figures things out, and start to do the laundry by themselves.
Now for the actual sensor, I have thought of a simple laser / light sensor mounted in the back of the drawer cabinet, that bounces off the back of the drawer front, and if the line of sight is blocked, report this to the home automation. There is 5-10 cm from the top of the basket, to the top of the drawer front / table top, so there should be some space for electronics. Alternative mount a ultrasonic distance sensor above the basket, but it could potentially be knocked off, if the basket get's filled too much.
As a secondary thing, when people take out the clean laundry from the dryer, they just put it on top of a table, and thinks that "Someone else is going to fold it, and put it aside". I have thought about a simple ultrasonic distance sensor, that detects distance from the ceiling to the table above the dryer. If the distance is below a certain point, it alerts the home automation setup.
Only issue I fear coming up, is the WAF, as I think that she wouldn't approve of the "disciplinary punishment"
Any other thoughts about the above sensors?
It's more or less the script below, I removed some stuff from my utility meter, that I also handles in my script (but isn't needed for just mysensors).
#!/usr/bin/python3
import serial
import paho.mqtt.client as mqtt
import time
mysensors = serial.serial_for_url('/dev/ttyNRF24', 115200, rtscts=0, timeout=1000)
mysensors.isOpen()
mqtt_connect = False
def mqtt_reception(client, userdata, msg):
payload = msg.payload.decode("ascii")
topic = msg.topic.split('/')
topic.remove('mys-out')
mymsg = ';'.join(topic) + ';' + payload + '\n'
print("out <- ", mymsg)
mysensors.write(str.encode(mymsg))
def mqtt_connect(client, userdata, flags, rc):
print('Connected')
mqtt_connect = True
client.subscribe('mys-out/#')
def mqtt_disconnect(client, userdata, rc):
print(rc)
client = mqtt.Client('mys')
client.on_connect = mqtt_connect
client.on_message = mqtt_reception
client.on_disconnect = mqtt_disconnect
client.connect('192.168.1.64')
def handle_data(data):
print("in -> ", data)
s = data.split(';')
payload = s[-1]
s.remove(payload)
topic = 'mys-in/' + '/'.join(s)
if mqtt_connect:
client.publish(topic, payload.rstrip())
if __name__ == "__main__":
client.loop_start()
mys_data_str = '';
util_data_str = '';
while (True):
if (mysensors.inWaiting()>0):
char = mysensors.read(1).decode('ascii')
mys_data_str += char
if (char == '\n'):
handle_data(mys_data_str)
mys_data_str = ''
time.sleep(0.01)
@magpern this module is more if you want to only use nodered for your home automation, without Hass.io or the likes.
Most of the standard controllers (domoticz, openhab, home assistant etc) are all able to handle the mysensors network directly (more or less), and handle the device registration for you.
My module for nodered is implementing a crude controller, that only handles out node id's to the mysensors network, when asked for, and then leaves everything else up to your imagination, for how to wire things together behind the scenes.
That being said, if you only need to convert your serial mysensors gateway into mqtt, I do have a small python script that I use for that particular case, as my nodered instance can't access the serial Ports from my docker swarm setup. So this script is running besides everything else. I can post it tomorrow if interested, as my pc is closed for the night now.
/ Thomas
Not hardware though, but I'm working on my dashboard solution for a wallmounted tablet.
I kind of already have this working in angular, but lately I have switched to react (joined a new project at work where they use react for frontend development), so wanted to see how "easy" it would be to create a something similar to my angular dashboard, but in react instead.
Made a small screencast of the dashboard, and are now playing around with "fully kiosk browser" on an old android tablet. Next step is to create a wooden frame, so it doesn't look like a tablet that is mounted on the wall (for WAF'ines).
home automation tablet view – 02:07
— Thomas Bowman Mørch
All data, except weather forecast and channel lists, is from my mqtt broker, where I use mqtt over websockets.I have a bunch of python scripts, and a node-red instance, to wrap things up for the display, and also do a lot of magical stuff, like turning on TV and amplifiers automatically when casting youtube, and then turn the TV etc. off again, when I haven't streamed anything for a couple of minutes. (So the kids doesn't forget to turn off the TV when they're finished watching cartoons)
The AS3930 is designed to be used with inductive coupled antennas, as in RF-ID devices, where it can feed on the energy generated in the magnetic loop..
It's stated on page 12 in the datasheet (detailed description)
The AS3930 is capable of detecting the presence of an inductive coupled carrier and extract the envelope of the ON-OFF-Keying (OOK) modulated carrier
So unless you want to make a gigantic coil, covering the whole house, it's not usable for waking up low power (my)sensors devices.
Version 3.3.1 is now released
A bug with the Encapsulate node was reported, where child sensor types didn't have the correct IDs. Also updated dependencies as a few where reported to have security vulnerabilities.
Can't remember if there where any specific books.. Probably some magazines like elector electronics in the 80's. There where a couple of danish magazines back in the 80's, that I read at the public library (I'm turning 46 in a couple of weeks).
Perhaps a broadcast once an hour, with "I'm here and I'm alive". I'm doing this with all my sensors, even battery powered ones.
@alowhum it's SPI, so can probably share it's bus with nrf24/rfm69, only one extra pin for CS is needed.
You could probably also have a supercap which supplies just enough power for a "last will" message from the node, just before it dies.
When using a WDT you're spamming the network with "I'm here" all the time..
@sergio-rius
Just be aware that with W5100 you can probably not share the SPI bus, as there is a hardware limitation in the W5100 (depending on how the manufacturer has wired the chipselect lines)
This forum is about sharing ideas, and open source software / hardware.
So that is the reason people ask you to put your questions out in the open here. So all can collaborate, and enjoy the teamspirit
I've been away from the forum a couple of days.. But the W5100 is actually the one that is promoted around the site. It does however have a problem with sharing SPI port in some circumstances, as it doesn't let go of MISO (if I remember right).
For ENC28J60, it does say that you might run into trouble with memory, as it does have a larger memory footprint.
mysensors "store" : https://www.mysensors.org/store/ethernet
build instructions: https://www.mysensors.org/build/ethernet_gateway
As an example, I have 4 sensebender micros, which has been deployed since june 2015, I have not changed batteries in them yet, and they still report temperature / humidity (that's more than 4 years battery life!). The batteries are cheap AA alkaline types (2 pieces)..
I'm not doing anything special (hardware wise) to power off the radio, or any other peripherals on the board. Only using standard software powerdown where available.
Perhaps I should make a 433Mhz transmitter, with a 1kW (60dBm) pa on it (which is the limit of my ham license)
The problem with higher power output on the radio side, also consumes more power from the (battery) supply.. In case of battery operated sensors I'm always aiming for the lowest possible current draw, which is with non-lna devices.
Luckily I live in the country side, so the 2g4 band is not that crowded, and so my nrf24 is good enough. I tried to use rfm69 as well, but never got the range above 50cm, so I ditched those radios again. And yes, I added decoupling capacitors etc. all over the thing..
As others say, cheap clones of the nrf24 is probably the issue here..
I'm using mysensors with nrf24 throughout the house. Some sensors have been in operation for more than 4 years now, with the same set of 2xAA batteries! And they are still happily reporting temperature / humidity. I admit, that I do not have many sensors around, as I only have 15 or so units deployed (mix of both sensor-only and actuators)
many retirement homes here in Denmark, has special pressure sensitive floors, which can notify the staff if one of the elders have fallen. But that is probably a bit over the top, to put those into your parents house
I sort of decided that a GW never has to sleep, as it is usually connected to a good powersource (usb, wallplug etc), so we do not need to conserve power by going into sleep..
@nalith the sensebender gateway does include both 10uF and 100nF decoupling capacitors on the board, near the radio.
I haven't used the GW with rfm69 that much though, so I have no experience with that radio.
@zboblamont is the battery and switch mode psu just in parallel? Or do you have a switchover / charge circuitry?
just for reference, 4 out of 5 sensebender micros is still running strong, after 4 years on the same set of AA batteries, that they where deployed with.
The 5th sensebender is my outdoor sensor, that one eats a set of batteries within 48 hours.. But i suspect that it's the air humidity that has destroyed it.
I opted to use a Si7021 instead (there might be others that are better suited now), as it was better suited for low power operation, and runs down to 1.2V supply (if I remember right).
@bjacobse something is wrong, it's way to organized
What @scalz is trying to explain, is that we are not doing mysensors as our daytime job, we do have a "normal" job to attend to, in order to get money to pay the bills. And we have families that demands our presence when we are at home, and then we have mysensors and other projects (house rebuilding, old cars / tractors etc) during the small sparetime that is left.
In other words, we are not super humans, that can work 30 hours a day
The feature request forum part, is there so others can chip in on the development, and we can discuss features that might be a good idea to implement. But it doesn't say WHO should implement it..
You are welcome to create a PR with the improvements that you need. After all this is an opensource project, which means that the source is available to be hacked on, it doesn't need to be one of the few core group members that do all the code changes. We are happy to do code reviews as well on incomming PR's.
One of the benefits of opening the documentation, is that others can review your design, and give you new ideas for features/implementation details, specially if you are "new" to electronics.
The link you provide for current mode led driver, is not meant to be switch on/off like the one I provided.
It also puzzles me, that you have chosen a 3W diode manufactured for nightvision? Normal small 3/5mm LEDs are just fine (that's what the RC manufacturers are using). Look at the datasheet, and peak forward current. For example this datasheet. Constant power iFwd is 100mA, but if you are using it with short pulses (like a IR transmitter, where you have a carrier freq. of 38Khz which you modulate), you can actually safely drive it with up to 500mA iFwd.
One benefit of using multiple diodes, is that you can put them on an arc, like this:
And then spread your IR signal on a larger area.
fair points..
One thing, you mention that it's 1$ for the chip.. Now imagine if you want to sell this puppy by millions, then 1$ is a lot of money
Have you considered a constant current driver instead, like this one https://arduinodiy.wordpress.com/2015/12/12/driving-an-ir-led-constant-current-source-or-not/
To be honest, I think that having a dedicated 1.5V rail (with a step down switchmode) just for IR transmitter, is not an ideal solution.
I would use 2-3 ir leds in series, depending on vFwd, and then a series resistor to limit the current (might not be needed). and then run the thing from the 5V rail, and use a n-fet as a LED driver to power the LEDs on / off, like the schematics shown in the first response here https://electronics.stackexchange.com/questions/67775/driving-led-strip-from-microcontroller
You do know that this forum is build upon open source, and open hardware, right?
I see that you have a step up supply, which rail is it connected to? And what is the output voltage? What purpose does it serve? The same with step down converters? You have both 5v and 3v3 available all ready?
No schematics anywhere?
To my best knowledge, you should be able to use all the standard features, together with mqtt, including auto-id assignment
But then again, I have never used OTA yet so I can't tell you if that part works.
I'm not expert on the networking details in the gateway (even though I designed the hardware ).
But if you want to use multiple controllers, I would recommend that you use MQTT as transport layer instead, and then have a mqtt broker, like mosquitto, on your HA server.
@rollin said in not working Ethernet Gateway (sensebender and/or nano gateway):
the pinout of the jtagice 3 is not compatible with the cortex m0 debug port pinout. So check the pinout of both devices first to be sure that they match.
Yes, they do not match.. I created https://www.openhardware.io/view/192/Jtag-ICE3-breakout to make them compatible
@boozz said in 
Arduino UNO NRF24L01+ Shield:
You mean this item?:
I never used the CE pin via a level converter up till now. I use 3 of them and some have ran for over 3 years now. So I guess they must be 5V tolerant.
Of course, power pin should be 3v3.
Yes..
Some of the chinese modules, have the CE pin wired to the PA/LNA chip RXEN, which is not 5V tolerant, like the NRF24 are. So it is not safe to assume that it is OK to just apply 5V to the input pins.
You are just lucky, that the designer of the radios that you are using, didn't do this..
You should always check the datasheet from your vendor, as they might do things that are less than ideal
The design assumes that the 3v3 on the arduino board is enough to power the NRF24, which it isn't on the cheap china clones.
Also be aware that if you add a high powered NRF24 with PA/LNA onboard, the LNA is not 5V tolerant, and you need to have a level converter on the CE pin (if I remember right)
I'm using linux and have arduino 1.8.8 installed. There is a library manager in it :).
I have not had any problems in programming / using mysensors libraries in my setup, and had the libraries installed through the library manager. Works without any issues. This is on a ubuntu 18.10 install though, and not mint.
Today I have released version 3.3.0
New in this release is that battery level messages is now stored in the database by myscontroller, and are also shown in the table of mysensors nodes, when you open the myscontroller node.
MySensors library initializes the serialport itself, and set the baudrate to 115200 on standard arduinos. Regardless of what you are setting in your setup routine.
Another thing I noticed, is that you are trying to run arduino as root on your linux. This is not recommended, you should always run as a non-priviledged user, and only switch to root when it is absolutely necessary for the thing that you are trying to do.
New release: 3.2.1
It's a minor update, as @pansen discovered that node IDs wasn't shown correctly in the controller view.
Also behind the scenes a couple of module dependencies have been updated to mitigate a couple of (low-risk) security issues that has been reported against some of the npm packages that was used.
FYI, I just released a small bug fix for the mysensors node-red module.. So now it should show the correct node ID's in the controller view.
@pansen said in Howto use myscontroller from node-red-contrib-mysensors?:
Hi,
@tbowmo maybe you can help with this:
I would like to use the myscontroller node from node-red-contrib-mysensors to assign and track my node IDs, but I am struggling with the details.
myscontroller seems to read the incoming msg, but when double clicking on it, id is "undefined".
Just double checked my install.. And it too is showing nodes as undefined ID.. It seems that a bug appeared, and I have to dig into it..
When I set the topic in myscontroller to the publish topic, the following message spams my broker:
mysensors-out/1/255/3/0/6 : msg : Object object topic: "mysensors-out/1/255/3/0/6" payload: "M" qos: 0 retain: false _topic: "mysensors-out/1/255/3/0/6" _msgid: "f5d62d35.d8df2" topicRoot: "mysensors-out" nodeId: 1 childSensorId: 255 messageType: 3 ack: 0 subType: 6 origin: 2 messageTypeStr: "C_INTERNAL" subTypeStr: "I_CONFIG"
This is the controller nodered node responding to an I_CONFIG request from your mysensors node. It sends an "M" to the node (for "M"etric). Does the node receive the data?
It should only send one for each request from your mysensor nodes..
I haven't looked that much at my system the last 6 months or so, other than doing a restart now and then, because it drops either my ikea lamps, or my chromecast integration. Been to busy with the paid work thingy
One extra question, what type of filament did you use to print the adapter plates in?
Cool setup, is the adapter system stable enough for the different tools? Mainly the extruder for the 3d printing?
it should not be necessary to use atmel studio to erase the SAMD21, before you reflash with arduino. The erase cycle is included in the flashing process..
I haven't updated any of my SAMD21 boards the last year, so don't know if the latest arduino breaks something.
Is it the standard gateway sketch, that you use for the SAMD? If not, could you post it here? Then I could try programming a spare samd21 board at some point in time (I do not know if I get around to it this weekend, as I have lots of projects running through my head at the moment)
The SAMD21 doesn't use fuses in the same way that the atmega328 does. For example, the clock source is set by the running program, and not the fuses. The only fuses awailable is BOD, WDT and bootprotection. (See page 43 in the datasheet)
So it should not be needed to set any fuses in the SAMD21 in order to get it working.
btw. why did you replace the atsamd21 in the first place?
Interesting, please note that you need to be a licensed ham radio operator, in order to use the APRS network.
You probably need to have the bootloader as well.. I might be able to find the binary/hex file that is used for production at itead in one of my old emails.. If interested, I can send it in an email (PM me if interested)
Could you please keep it in English in here, as more people would be able to help, and benefit from the conversations.
Also try reformatting your code, using the inline code format, it makes it much easier to read your code
@rafael156
I'm using the atmel jtagice3 myself, however I had to make a converter, because the pinout of the jtagice 3 is not compatible with the cortex m0 debug port pinout. So check the pinout of both devices first to be sure that they match.
In order to program a fresh atsamd21 without bootloader, a CMSIS-DAP / SWD debugger. P4 on the front of the PCB (unpopulated) is used to connect with a standard cortex swd pinout.
Alternative connection is on the back, where there are testpads for the SWD / CLK pins.
You do know that there is solder pads for RFM69HW on the back of the gateway, right? And the board defs for the gateway are configured to use those pins for the RFM69 by default.
