Introductions and Range Issues

[Carywin]

Good luck! I tried sticking with NRF long past when would be sensible, and still ended up changing to RFM69 (433 MHz in my case).
Inner-city areas are just too crowded on 2.4G now, but paradoxically 433 MHz is becoming freeer thanks to the slow death of landlines and DECT phones.

[zboblamont]

@titvs Attenuation of 2.4GHz signals is your primary problem as @scalz pointed out, and here is a real example:
My Gateway is in the centre of the masonry/concrete house, with 433Mz communications to Nodes currently at 10-15m radius, but previously tested out to 30m without any problems.
The Router maintaining the Domoticz link to my Wifi laptop is at the front of the house nearest the Gas Meter Node, if I go to reprogram the Gas Node I lose the Wifi signal even at 10m, but the Node continues comms flawlessly.
Boosting the signal with a reflector behind the Node or changing to higher power output might increase the attenuated signal sufficiently to make the connection, but failing that, change to lower frequency devices.

[Yveaux]

@titvs said in Introductions and Range Issues:

i already orderer a NRF24/PA/LNA

Please make sure to buy decent quality modules, like the ones from Ebyte. They are only marginally more expensive but will save you a lot of trouble.

I'm only using nrf24 at home and covered my whole house. To counter range issues I've setup a few gateways (one on the ground floor and one in the attic for example)

[sundberg84]

@titvs - a gateway could be run through a good quality (genuine) phoneadapter without issues. To point out a regulator in the MySensors shop use the AMS1117 and not LE33ACZ because the later is to weak. You can also use pretty much any other linear regulator that can handle about 1A or more. Dont forget to check the datasheet for recommended setup with capacitors.

The only thing you need to do is to setup for a repeater is (example here)

// Enable repeater functionality for this node
define MY_REPEATER_FEATURE

I would also really check the capacitors again and try some different values.

[rozpruwacz]

@titvs I also would recommend to go for rfm69. I had exactly same problems as you, and tried a lot of differenr nrf24 modules and still couldnt get a descent result. As soon as i switched to rfm69 all my problems disappeared :) after all i spent more money on nrf24 modules than i would spend on rfm69 if i would start with them. Use nrf24 to rfm69 adapter board for easy transition.

[joaoabs]

@rozpruwacz said in Introductions and Range Issues:

@titvs I also would recommend to go for rfm69. I had exactly same problems as you, and tried a lot of differenr nrf24 modules and still couldnt get a descent result. As soon as i switched to rfm69 all my problems disappeared :) after all i spent more money on nrf24 modules than i would spend on rfm69 if i would start with them. Use nrf24 to rfm69 adapter board for easy transition.

Hi,

I'm also experiencing some troubles with NRF24+'s, but I already have a considerable number of nodes with it. Most of them have the 2x4 pin header like the ac/dc ssd relay and the slim node.

Is there any adapter that could easily adapt ( ! ) from the NRF24+ format to the RFM69?

Thanks,
Joaoabs

[rozpruwacz]

https://www.mysensors.org/hardware/nrf2rfm69
I tested it and it works. It is just a little bit larger than a typical nrf24 module.

[titvs]

Thanks for all the input guys and the vote of confidence regarding the NRF24! :D

So, since there are a lot of people telling to go for the RFM69, I've ordered a couple of them to do some testing. Im still going to test the NRF24/PA/LNA but after all you guys said, im not feeling very confident…

If everything doesnt go according to pIan, i guess im going to have to try to fit the RFM69 inside my already printed boxes and have an antena protruding out of the box… not very pretty… Ah well. Nothing is perfect. :)

Anyway, I think there should be somekind of note on the radio page of the main website, stating this kind of problem regarding range and construction type.

Will keep in touch.

[rozpruwacz]

@titvs said in Introductions and Range Issues:

If everything doesnt go according to pIan, i guess im going to have to try to fit the RFM69 inside my already printed boxes and have an antena protruding out of the box… not very pretty…

You don't have to :) just use a flexible, insulated 1/4 wavelenght wire antenna and keep it inside the box. It does not have to be straight.

[Yveaux]

@joaoabs said in Introductions and Range Issues:

Is there any adapter that could easily adapt ( ! ) from the NRF24+ format to the RFM69?

As @rozpruwacz said it works, but keep in mind that the RFM69 is not 5V tolerant on its IO lines like the nRF24.
This means that if you plan to use it connected to a 5V arduino you have to put a level shifter in between.
Also, make sure to connect the IRQ line, which is optional for nRF24 but required for RFM69.

[sundberg84]

@titvs said in Introductions and Range Issues:

Anyway, I think there should be somekind of note on the radio page of the main website, stating this kind of problem regarding range and construction type.

I have around 25 nodes using Nrf24 radio. Once you have found your way and how to test them before deployment they work very good. Sure, its easier with the RFM69 but expensive and as @Yveaux said not 5v tolerant.

For me I have a fixed node on my lab-bench with a socket for the radio. This way I can test any new radio before I use them. Most of them work great with a 4,7uF cap. In your setup I would guess it will work much better once you get the gateway a proper and stable radio.

[alowhum]

isn't the NRF24 able to use a higher frequency than wifi? I read that if you choose channel 100 you get much better results in Europe.

[mfalkvidd]

@alowhum it might work better, but you'll need to consider if it is worth breaking the law.

In EU, the highest nrf24 channel that can be used legally is channel 83.

[rozpruwacz]

@mfalkvidd I also tried that outlaw trick and the results weren't satisfying :) I think that your microwave is more affecting the comunication than the wifi. This is another issue that I think that 2.4GHz is not suited for HA - can't turn on lights when using microwave is not acceptable :D

[titvs]

@sundberg84 said in Introductions and Range Issues:

I have around 25 nodes using Nrf24 radio. Once you have found your way and how to test them before deployment they work very good. Sure, its easier with the RFM69 but expensive and as @Yveaux said not 5v tolerant.

Yeah, they are really more expensive than the NRF24. But still cheaper than the cost of a repeater node (NRF24 + Nano) even without considering the power supply... Well, i guess that each case is different. :)

As for the 5v logic vs 3.3 thats also a problem with 5v Nanos. I think they need something like "this" ?

I'll still do tests with the NRF24/PA/LNA and with repeater nodes but i think i wont be so lucky with a weather station node (needs to be outside) that im planning to add to the system. Lets wait and see. :)

I'll also prepare the STL files to upload to Thingiverse of my box. Need to take some pictures of the assembly first. :)

Thanks for the input!

[sundberg84]

@titvs said in Introductions and Range Issues:

Yeah, they are really more expensive than the NRF24. But still cheaper than the cost of a repeater node (NRF24 + Nano) even without considering the power supply... Well, i guess that each case is different. :)

Both RFM and NRF modules needs power supplies and if the plan to have more than 2 modules the cost for a repeater is less than building two RFM modules. Its more about if you want to add more money for less hassle.

[titvs]

Hi everyone,

As promised, the box over at Thingiverse (smaller version):

https://www.thingiverse.com/thing:3019468

[Mathea90]

@sundberg84 said in Introductions and Range Issues:

@titvs said in Introductions and Range Issues:

Anyway, I think there should be somekind of note on the radio page of the main website, stating this kind of problem regarding range and construction type.

I have around 25 nodes using Nrf24 radio. Once you have found your way and how to test them before deployment they work very good. Sure, its easier with the RFM69 but expensive and as @Yveaux said not 5v tolerant.

For me I have a fixed node on my lab-bench with a socket for the radio. This way I can test any new radio before I use them. Most of them work great with a 4,7uF cap. In your setup I would guess it will work much better once you get the gateway a proper and stable radio.

Can I ask which kind of power supply you are using for the transceivers?
I have now tried various LDOs in a combination with various different kind of capacitors, RC filters, LC filters... but in the end I'm still not really satisfied and sometimes results are not repeatable.

I currently have the best connection with a node whose radio is directly supplied by the power regulator of the Arduino pro mini. The Pro Minis are using MIC5205 LDOs which can only supply 150mA but are absolute low-noise regulators. The node is communicating with a gateway using the large NRF24L01 with external antenna, AMS1117 power regulator, an LC filter and at least 40µF worth of various ceramic capacitors. But this only started working properly since I shielded the radio with aluminum foil and moved the device away from other electronic devices (I assume the DIY style of homemade devices cannot cope with EMI as good as FCC approved consumer electronics can).

All other nodes are custom designed PCBs with the Atmega328 directly soldered onto it and a dedicated LDO for the radio. So far I've tried the LE33, AMS1117, LP2950ACZ and I have still unassembled boards designed to use an MCP1700 and an MIC5207 which I have not tried yet. Before I read into the theory of linear power regulators basically none of those variants worked reliably. Now that I realized that the LP2950ACZ for example has very very specific requirements for the ESR of the used filter caps, I desoldered my 47µF electrolytic cap on the radio and changed it to a 10µF ceramic cap. Now it works much more reliably but the connection is still spotty.

I am very eager to assemble the board with the MIC5207 because it has similar low-noise characteristics as the LDO of the Arduino Pro Mini.

In the end I am afraid that a lot of dangerious half-knowledge is being thrown around on this topic. Sometimes you hear that the problem with unreliable communications is caused by insufficient amount of power. My experience leads me to believe that in most of the cases the amount of power is absolutely not the problem (the LP2950ACZ can only supply 100mA but I still got them to kinda work). The main problem most people are fighting with is the power supply's noise. And this one is a bitch because the normal person without an oscilloscope cannot really measure and test this.

Now I am working on a revision of my standard PCB layouts which will include feed-through EMI filter capacitors, ferrite beads and a two-stage power regulation to filter the DC for the radios. Hopefully this will prove to work without failures.
Nevertheless I am really interested what other people who solved this problem are doing differently.

[sundberg84]

@mathea90

Can I ask which kind of power supply you are using for the transceivers?

Very different... mostly 2xAA but for AC-DC i use samsung and iphone charge/phone adapters. From 5 to 3.3v i use the ones in the Mysensors shop (LE33 and AMS1117).
The most important part is the capacitor as you know.

in the end I am afraid that a lot of dangerious half-knowledge is being thrown around on this topic. Sometimes you hear that the problem with unreliable communications is caused by insufficient amount of power.

Trying to drive a PA/LNA with a LE33 is a to common issue and that is why you hear this alot but bad clones is even worse common issue. You never mention this, and if you are unlucky you can buy a batch of really crappy radios. I have had good batches (10/10), and some with 7 out of 10 worked but I have had a discussion with people with 10 bad radios.

It can be frustrating and hard to understand all this... but are you sure creating an advanced pcb with all those extra components will make it easier? Sure you can prove a point but it will be hard to do this for you unless you buy genuine radios and compare them.

[Mathea90]

@sundberg84
Thanks for your answer. Currently I don't have any battery powered nodes but I assume that they cause a lot less headaches as you get a clean DC without the ripple and noise of AC/DC buck converters.

If you are using LE33s, what exact capacitors are you choosing? Do you have low ESR electrolytics? Or does the LE33 also work reliably with normal ones? Maybe it's also a good choice of yours to use high quality charger / phone adapters. I bought a bunch of cheap phone chargers from Aliexpress and I suspect that they have a really dirty output since they even supply 5.4V and not around 5V. I have read that power regulators like e.g. the LE33 have a parameter called "ripple rejection" which states how much of the input power's ripple gets through to the regulated output. So if the input is bad enough, the output will also carry this noise.

What kind of behavior do you notice from bad radios? Don't they work at all or do they work and produce bad range? I personally also received two bad radios once but in that case they didn't work at all. Of course, this is easy to see, but diagnosing performance issues is much harder. Since I've heard this statement a lot in combination with the NRF24L01 radios I also changed my transceiver against a different batch, but I always get the same performance, independent from the specific radio. Nevertheless all my radios were bought from the cheapest supplier at AliExpress at that time.

Now the last thought: Most of my tryouts with my custom PCBs were done with the 2.3.0 version. My most stable node is still running 1.6. I have read here that there seems to be communication problems with the latest version. Also here I have found a bug report which maybe correlates with my bad experience.