My experiences with MySensors
-
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)
-
You are right. Mysensors based on NRF is dead. I started my adventure with home automation based on mysensors but gradually I replaced all my sensor with Wemos. Now everything work fine. Sorry mysensors...
@rodaman
Although fully agree with @mickecarlsson and @tbowmo (I have been lucky with really cheap nrf24L01+ modules I think) and have a similar experience as @TheoL (MySensors works fine with nr24L01+ modules) I can state that mySensors is very much alive.However, I'm interested in your Wemos setup.
I've got some questions from which I'm sure you can give me the right quidance:
I guess it used some bandwith in your Wifi network? can I use it similar as mySensors topology or even with mySensors libraries?
Can it be used with batteries (sleep modes and so on) for a long time?thank you in advance for your values answers,
BR, Boozz
-
Well, unfortunately I cannot agree with the stories of success.
It's ok that a tool works for someone, but it can't be so weak that things like radio clones, track shapes or cap variations renders something unusable. And still did not mention interference or houses construction materials. All things that should be tested. WiFi/mobile interference should not be a problem, everyone has em.
Just only think that if Chinese manufacturers are making these clones in that volumes, and they are sold, it's because they took a look at the original and made a product profile with what specs are acceptable for the expected application and clones are the result. I mean that perhaps mysensors expected to much from them.
I guess everyone already knows that I only had problems trying to use mysensors, and now I have a defeating gw problem while using original radios all around. I have nrf52s in the box for testing but I'm so defeated, that I have no strength to try them.
I've ordered parts for making a good ole rj45 gw, but I'm already working in a custom base sketch for moving to esp+mqtt.
My influx data is full of "holes"Now I'm thinking in the fact that my gw started to make this more or less after the installation of a Google home and a echo Dot, for testing. It already seems paranoia.
-
Well, unfortunately I cannot agree with the stories of success.
It's ok that a tool works for someone, but it can't be so weak that things like radio clones, track shapes or cap variations renders something unusable. And still did not mention interference or houses construction materials. All things that should be tested. WiFi/mobile interference should not be a problem, everyone has em.
Just only think that if Chinese manufacturers are making these clones in that volumes, and they are sold, it's because they took a look at the original and made a product profile with what specs are acceptable for the expected application and clones are the result. I mean that perhaps mysensors expected to much from them.
I guess everyone already knows that I only had problems trying to use mysensors, and now I have a defeating gw problem while using original radios all around. I have nrf52s in the box for testing but I'm so defeated, that I have no strength to try them.
I've ordered parts for making a good ole rj45 gw, but I'm already working in a custom base sketch for moving to esp+mqtt.
My influx data is full of "holes"Now I'm thinking in the fact that my gw started to make this more or less after the installation of a Google home and a echo Dot, for testing. It already seems paranoia.
Just my "little" two cents on this because I don't think it's MySensors fault only..afaik MySensors started with rf24, but got compatible with better radios years ago!
it can't be so weak that things like radio clones, track shapes or cap variations renders something unusable. And still did not mention interference or houses construction materials. All things that should be tested.
when you learn RF and electronics you'll discover what's the purpose of decoupling and buffering power. Same about the shape of the board, gnd plane size, enclosure, parts close to antenna et, will affect antenna frequency like shifting it, and decrease sensitivity (noise).
Manufacturers of end device passes certification to prove their device are RF "green" and calibrated. Even when you change the external antenna on a device, and don't buy the same original one, you may lose CE/FCC, because there are chance it detuned a bit RF, more or less.
These are things that a beginner would not imagine when picking a rf module, arduino, some dupont cables and playing.. So there can be variations in users nodes which can weaken link quality (Pity when the radio is only 0dB max). Far from reference design concept. Not easily coverable by MySensors team.
Don't trust me?? Just read, manufacturers datasheet, application notes, check with good RF tools or even watch some youtube videos.
Just only think that if Chinese manufacturers are making these clones in that volumes, and they are sold, it's because they took a look at the original and made a product profile with what specs are acceptable for the expected application and clones are the result. I mean that perhaps mysensors expected to much from them.
those who clones and sell volumes, why do you think they care if people continue to buy because it's cheap. they just simply sell. And when you get a problem, they show you the github from someone else who could help..
WiFi/mobile interference should not be a problem, everyone has em.
No sense, not scientific. Coexistence of multiple protocols in one RF band isn't magic. What happens when people in same room all talk at same time, and some louder?? You can catch louder people and rest is just a mess. Sort of jamming, you could even pick esp8266/32 for short range jamming 2.4gz, never tried..but in this, the stronger wins.
See this pic, especially for those who would like to use different products on same band..
(source: https://www.researchgate.net/figure/Spectra-of-multiple-wireless-technologies-in-the-24-GHz-ISM-band-The-colors-indicate_fig4_326598630)
why not adding MySensors protocol in this 2.4ghz recipe?? Just pray that the freq of your node is not shifted/detuned too much.. I hope too you choosed the right freq for your network.
- Which protocols is either fixed freq, or freq hoping based??
- What if there more or less powerful repeaters for these ?
The only way would be to run protocols concurrently (with time slots). Not possible when there are separate gw for each protocols, it just get messy in air, lost packets, etc.
Was it like this 10-15 years ago in homes, I don't think so.I hope I don't look old at repeating but finally, I think this problem is mostly about user choices..and I understand it maybe doesn't solve all problems.
-
Just my "little" two cents on this because I don't think it's MySensors fault only..afaik MySensors started with rf24, but got compatible with better radios years ago!
it can't be so weak that things like radio clones, track shapes or cap variations renders something unusable. And still did not mention interference or houses construction materials. All things that should be tested.
when you learn RF and electronics you'll discover what's the purpose of decoupling and buffering power. Same about the shape of the board, gnd plane size, enclosure, parts close to antenna et, will affect antenna frequency like shifting it, and decrease sensitivity (noise).
Manufacturers of end device passes certification to prove their device are RF "green" and calibrated. Even when you change the external antenna on a device, and don't buy the same original one, you may lose CE/FCC, because there are chance it detuned a bit RF, more or less.
These are things that a beginner would not imagine when picking a rf module, arduino, some dupont cables and playing.. So there can be variations in users nodes which can weaken link quality (Pity when the radio is only 0dB max). Far from reference design concept. Not easily coverable by MySensors team.
Don't trust me?? Just read, manufacturers datasheet, application notes, check with good RF tools or even watch some youtube videos.
Just only think that if Chinese manufacturers are making these clones in that volumes, and they are sold, it's because they took a look at the original and made a product profile with what specs are acceptable for the expected application and clones are the result. I mean that perhaps mysensors expected to much from them.
those who clones and sell volumes, why do you think they care if people continue to buy because it's cheap. they just simply sell. And when you get a problem, they show you the github from someone else who could help..
WiFi/mobile interference should not be a problem, everyone has em.
No sense, not scientific. Coexistence of multiple protocols in one RF band isn't magic. What happens when people in same room all talk at same time, and some louder?? You can catch louder people and rest is just a mess. Sort of jamming, you could even pick esp8266/32 for short range jamming 2.4gz, never tried..but in this, the stronger wins.
See this pic, especially for those who would like to use different products on same band..
(source: https://www.researchgate.net/figure/Spectra-of-multiple-wireless-technologies-in-the-24-GHz-ISM-band-The-colors-indicate_fig4_326598630)
why not adding MySensors protocol in this 2.4ghz recipe?? Just pray that the freq of your node is not shifted/detuned too much.. I hope too you choosed the right freq for your network.
- Which protocols is either fixed freq, or freq hoping based??
- What if there more or less powerful repeaters for these ?
The only way would be to run protocols concurrently (with time slots). Not possible when there are separate gw for each protocols, it just get messy in air, lost packets, etc.
Was it like this 10-15 years ago in homes, I don't think so.I hope I don't look old at repeating but finally, I think this problem is mostly about user choices..and I understand it maybe doesn't solve all problems.
@scalz Regarding WiFi I'm not saying that mysensors should overcome the problems of crowded or noisy environments. I'm just saying that if that's the war field, it should have slided to another one.
I spent more than 4 years on those problems. I'm physician and software architect/analyst, and my father was ham radio fanatic. Not boasting, only saying that I have a minimum knowledge. I created a nice environment with only one WiFi device, good ubiquity equipment, with nodes without boards or ground planning issues, studying the band and choosing channels and still had nightmares.
Yesterday I learned how to easily hang a node by sending a message to it. That shouldn't be possible.
I'm not saying that this is crap or that this doesn't worth playing with it. Only think that it needs some love on the low level even if it has to move to slightly powerful devices. Or adding capabilities on powerful architectures while still supporting the good old ino.
Perhaps combined with a custom nrf52 development board or something to make easier burn and solder them. Make an nrf52 mysensors starter kit, and convert sensebender to it.
Sorry, as you said I don't have the knowledge to make long run boards to contribute with them.
-
I think the problem is less about the radio and more about the transmit power. Even for official Nordic nRF24l01+ chips, without an external power amplifier, the maximum transmit power is just 0db. For a lot of people, that just won't be enough, which will inevitably lead to disappointment. Compare that to other radios which offer 20dbm tx power, and it's night and day difference in range, coverage, and satisfaction.
For beginners who want to use the nRF24L01 design, I humbly suggest that mysensors change its website to guide them to using modules with an external power amplifier and LNA, such as:
or similar. There are also much smaller modules with trace antennas and pa + LNA on them, and even so the cost remains low. If someone doesn't need the full Tx power, they can always scale back.
I think a big reason why people like the hoperf rfm69hw modules is simply that they provide 20dbm Tx power. So, you can switch to that or another radio, or you can simply amplify the type you've got. For comparison, an ESP8266 has a maximum 20.5dbm Tx power, which is set as it's default tx power. So, for the guy who moved everything over to wemos, maybe that's why you like it better.
Of course, there are differences in receive sensitivity, and that's obviously equally important, and even preferable. And datarate and coding schemes can be other large factors too. Still, if the max Tx power is capped at 0dbm, you're pretty limited in most cases as to your range and coverage and, in my view, a bit crippled for most things other than very short range.
-
@neverdie said in My experiences with MySensors:
I think a big reason why people like the hoperf rfm69hw modules is simply that they provide 20dbm Tx power. So, you can switch to that or another radio, or you can simply amplify the type you've got.
Too simplistic. It is not just about power but also the behaviour of different frequencies and how noisy the RF band is. 2.4GHz is not noisy for some but is for others, add in fakes and you have a nightmare. Your suggestion reminds me of a texan ham who used to claim he was turning his LF beam when he was slowly screwing up his linear.. ;)
I deliberately went with 433MHz rfm69s because of range and structural penetration, I'm in the sticks so the Wifi band is relatively quiet, no regrets, no fakes (yet), just MySensors doing it's stuff... :) -
@zboblamont said in My experiences with MySensors:
@neverdie said in My experiences with MySensors:
I think a big reason why people like the hoperf rfm69hw modules is simply that they provide 20dbm Tx power. So, you can switch to that or another radio, or you can simply amplify the type you've got.
Too simplistic. It is not just about power but also the behaviour of different frequencies and how noisy the RF band is. 2.4GHz is not noisy for some but is for others, add in fakes and you have a nightmare. Your suggestion reminds me of a texan ham who used to claim he was turning his LF beam when he was slowly screwing up his linear.. ;)
I deliberately went with 433MHz rfm69s because of range and structural penetration, I'm in the sticks so the Wifi band is relatively quiet, no regrets, no fakes (yet), just MySensors doing it's stuff... :)I think you're missing the point. How many people using, for example, primarily rfm69 are capping the Tx power on all their modules to 0dbm or less? I'd wager the answer is very, very few, if any.
-
Of course, if it were just for me to decide, I'd recommend all beginners start with LoRa, because if you can't find a way to be happy with the LoRa radio module (which, incidently, can also do everything an rfm69 module can do), you probably can't be happy with anything.
Is LoRa even supported by mysensors as yet? It's been a long while since I last checked.
-
@neverdie You're not related to my erstwhile Ham contact are you? :joy:
I believe you missed the point that it is not purely a power issue, TX/RX is a combo of frequency, noise, and pre-amp/antenna gain, environment, etc, etc..
I've no idea what the default setting is for my 433MHz RFM69s, perhaps when I update the system and the nodes I might be able to advise. They are standard devices not the high power version, all using 2dB whips, none have failed to communicate to date. I do have an HW version, it's still in it's antistatic bag...
I can track my wifi signal outside the house to about 5m, my Nodes are well beyond that. I could add gain to the wireless router to try matching the performance but the question remains why? -
@neverdie You're not related to my erstwhile Ham contact are you? :joy:
I believe you missed the point that it is not purely a power issue, TX/RX is a combo of frequency, noise, and pre-amp/antenna gain, environment, etc, etc..
I've no idea what the default setting is for my 433MHz RFM69s, perhaps when I update the system and the nodes I might be able to advise. They are standard devices not the high power version, all using 2dB whips, none have failed to communicate to date. I do have an HW version, it's still in it's antistatic bag...
I can track my wifi signal outside the house to about 5m, my Nodes are well beyond that. I could add gain to the wireless router to try matching the performance but the question remains why?@zboblamont
The problem is that, assuming FCC compliance, the allowed uses for the 433Mhz band in the USA are pretty limiting for narrowband communications. -
Of course, if it were just for me to decide, I'd recommend all beginners start with LoRa, because if you can't find a way to be happy with the LoRa radio module (which, incidently, can also do everything an rfm69 module can do), you probably can't be happy with anything.
Is LoRa even supported by mysensors as yet? It's been a long while since I last checked.
@neverdie said in My experiences with MySensors:
Is LoRa even supported by mysensors as yet? It's been a long while since I last checked.
Yes, we support RFM95 radios and various LoRa settings: https://www.mysensors.org/apidocs-beta/group__RFM95SettingGrpPub.html
-
@zboblamont
The problem is that, assuming FCC compliance, the allowed uses for the 433Mhz band in the USA are pretty limiting for narrowband communications.@neverdie No knowledge of FCC 433MHz restrictions, in the EU it is limited also, from vague memory on %age TX and probably ERP as it is a shared band. But you would need to be pumping a shedload of data for these constraints to take effect, and crucially it suffers none of the frequency crowding issues or penetration constraints of 2.4GHz.
-
I started out with NRF 2.4 MHz radios for my 8 nodes MySensors and Domoticz. As I also install WiFi networks for customers and have access to our equipment at work I did a measurement at home because I could not get a reliable network.
First measurement revealed that my DECT wireless phone occupied the same channel that was the default for the NRF radios. And that my neighbors run their WiFi occupying all other channels. Equipment used was Ekahau.
I tried to tweak the channels but no avail.
So, I bought new radios, RFM833, and that was the end of my troubles. For two year now my eight nodes have not skipped a beat. I measure temperature, humidity, light (lux), UV and radioactivity.And all my WiFi at home is at 5GHz as the 2.4 was exhausted.
I have about 40 NRF radios in a box with no use for them. -
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..
-
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..
@tbowmo You see, it's a lottery.
I also live in the countryside, and also having problems.
What about ble capabilities of the nrf52? Some claim that a Bluetooth network works.
I really doubt it, it was the most affected band by the microwave, when I analyzed interference.Bs:
- Adaptive signal power
- Low level checksum error control and retry
- Channel analysis and automatic selection
- Double channel / twin radio gw 😓
- Nrf52 ble advertising/handshake and then regular 2.4g transmission? 😱
-
@tbowmo You see, it's a lottery.
I also live in the countryside, and also having problems.
What about ble capabilities of the nrf52? Some claim that a Bluetooth network works.
I really doubt it, it was the most affected band by the microwave, when I analyzed interference.Bs:
- Adaptive signal power
- Low level checksum error control and retry
- Channel analysis and automatic selection
- Double channel / twin radio gw 😓
- Nrf52 ble advertising/handshake and then regular 2.4g transmission? 😱
@sergio-rius said in My experiences with MySensors:
Bs:
- Adaptive signal power
- Low level checksum error control and retry
- Channel analysis and automatic selection
- Double channel / twin radio gw 😓
- Nrf52 ble advertising/handshake and then regular 2.4g transmission? 😱
duh, it looks like a subset of feature list we studied for gateways we designed (Janus, Halo) ;)
Still,
- Adaptive signal power
Available for RFM modules only, csma as well. Whereas nrf24 has no true rssi feedback.. - Low level checksum error control and retry
Afaik it's available for rfm69, no ideas for nrf24 I don't remember - Channel analysis and automatic selection.
We know this, actually it's not really hard to do. It doesn't protect against potential unfortunate jamming of course, or when 2.4ghz is crowded, but it's good start for a network. - Double channel / twin radio gw
No dual channel on same band, or freq hoping planned for the moment afaik. For the moment, only local dualrf pocs - Nrf52 ble advertising/handshake and then regular 2.4g transmission?
doable with non arduino environment only, so for the moment it's out of MySensors scope.
BLE is using freq hoping if not wrong, so in theory it might be "more resistant" to interferences than non-freq hoping protocols (depends on their strategies for reliable packet delivery). Freq hoping may imply change on hw too, for precise timing etc
that's why I'm saying old tech may get limited at some points, this implies clocks and mcus, radio etc.. By limited I mean, like a subset of features available for less powerful mcu etc. But it's too soon for saying, we miss time too..
Always suprised when someone reports rfm69 poor range..sounds crazy to me, but I guess it was a problem with sw (like the old rfm69 lib, a good test is to use radioheadlib/lowpowerlab libs for checking link, then you know if it's sw or hw) or hw (like sometimes you can receive the wrong freq for a module etc). no idea, pity!
-
Well, unfortunately I cannot agree with the stories of success.
It's ok that a tool works for someone, but it can't be so weak that things like radio clones, track shapes or cap variations renders something unusable. And still did not mention interference or houses construction materials. All things that should be tested. WiFi/mobile interference should not be a problem, everyone has em.
Just only think that if Chinese manufacturers are making these clones in that volumes, and they are sold, it's because they took a look at the original and made a product profile with what specs are acceptable for the expected application and clones are the result. I mean that perhaps mysensors expected to much from them.
I guess everyone already knows that I only had problems trying to use mysensors, and now I have a defeating gw problem while using original radios all around. I have nrf52s in the box for testing but I'm so defeated, that I have no strength to try them.
I've ordered parts for making a good ole rj45 gw, but I'm already working in a custom base sketch for moving to esp+mqtt.
My influx data is full of "holes"Now I'm thinking in the fact that my gw started to make this more or less after the installation of a Google home and a echo Dot, for testing. It already seems paranoia.
-
That is just the point I had in my mind. I'm convinced that mysensors is a great project and the work done here is golden, really. But I would like to know why the documentation always seems to recommend hardware that when you fail to implement, at some point someone tells you that it was expected.
That happened me again this week with the Ethernet gw. Following the documentation, hours and hours trying to make the module work, to share spi port, etc... and then someone tells me better getting a w5100. Then why is this board in the official docs? Let's take it out and put a warning.
The docs are full of pinout pics of nrf24. But if you use the search (for hours) you get plenty of posts from vip telling... Better going with rfm69.
I hope it's not the same with rfm69. -
@rodaman
Although fully agree with @mickecarlsson and @tbowmo (I have been lucky with really cheap nrf24L01+ modules I think) and have a similar experience as @TheoL (MySensors works fine with nr24L01+ modules) I can state that mySensors is very much alive.However, I'm interested in your Wemos setup.
I've got some questions from which I'm sure you can give me the right quidance:
I guess it used some bandwith in your Wifi network? can I use it similar as mySensors topology or even with mySensors libraries?
Can it be used with batteries (sleep modes and so on) for a long time?thank you in advance for your values answers,
BR, Boozz
@rodaman: May I remind you of my questions:
@rodaman
...
However, I'm interested in your Wemos setup.
I've got some questions from which I'm sure you can give me the right quidance:
I guess it used some bandwith in your Wifi network? can I use it similar as mySensors topology or even with mySensors libraries?
Can it be used with batteries (sleep modes and so on) for a long time?thank you in advance for your values answers,
BR, Boozz
Unfortunately no reaction from your side till today. In the meanwhile I've searched the internet for a ESP8266 or ESP32 mesh network as an alternative for the mySensors network. I cannot find a good working open source (read Arduino based) startpoint, so I'm wondering if you could direct me to a good starting point.
Thanks,
Boozz
