💬 Building a wired RS485 sensor network

[rejoe2]

@nofox The problem you describe sounds pretty similar to my observations in this post.
There are also some suggestions how to solve this kind of trouble you may apply.

[nofox]

Hi! It’s me again. I want to precise one thing. Are there anyone here who have succesful working RS485 network without any issues? We are talking here about issues but maybe someone have fully working sensors network?

[pjr]

If some node hangs what happens if you recycle the power of rs485 transceiver only?
I'm just thinking could it be the protection of max485 that kicks in if there is two nodes sending exactly the same time.

[nofox]

Hi! My problem still exist. I switched to newest MySensors Library and to HWSERIAL and some nodes acts the same as before... Now i'm thinking that maybe its related to DE_PIN 2 ?? Someone tried to switch to other pin for Enable function for RS485 ?? Now after almost a year I'm thinking about changing my transmission from wired RS485 to wireless nRF.. I choose RS485 because of its stability (thats what I taught over a year ago ). Now I think that RS485 its a piece of garbage..

[nofox]

Hi! I think I came up with the solution for all our problems but I don't know if it is worth and possible at all. The thing is that RS485 isn't a multimaster bus, thats what we all know. So the best option is to turn the gateway in to master and all other nodes to slaves. I'm thinking if there are any possibilities to make the gateway to ask nodes in the order one by one for its states (child states). The nodes can only talk to the gateway when the gateway ask for their states. This process should take place in endless loop in the highest possible speed (depending on the count of nodes in the network). I think that in this option the baud rate matters, cause higher baud rates makes the time of a single loop shorter minimizing the delay that may occur in the networks with many nodes. Maybe I'm reinventing the wheel now and this type of transmission needs to rewrite all MySensors protocols..

[wallyllama]

@nofox absolutely that is one of the problems. Polling as you describe could be a fix, another is having each node listen before transmitting, there is a (not mysensors) library that does that, CAN bus( a mutli master 485 comaptible bus) is another option. I think that running a common ground to all the rs485 adapters is also wise. This is all in theory as I havent wired up my home yet....

[pjr]

@wallyllama

another is having each node listen before transmitting

How would you describe this functionality in MySensors rs485 code?
https://github.com/mysensors/MySensors/blob/f5ed26c778c18107d1516bd86704f8c8f99ff571/hal/transport/RS485/MyTransportRS485.cpp#L246
If I'm reading the code correctly it is exactly reading the bus and making sure no other node is sending before sending anything.
The place where collision can happen is those 5 microseconds what the node is waiting to get driver tx enabled pin up.

So the collision avoidance is already there. So what MyS R485 transport is missing is collision detection and recovery from collision.

CAN bus( a mutli master 485 comaptible bus) is another option.

Its disadvantage is small payload size.

@nofox Its quite like how the Modbus works. By polling.
I think there is some better ways to do that. Check these out from google:

• RS-485 Token ring
• uLAN
• RS-485 RRP (Round Robin Protocol)
• CSMA/CA

[wallyllama]

I think you have found most everything that I have. Something akin to csmacd is what I was thinking. Listen, transmit, back off for random time if a collision still happens. Of course full duplex rs485 is another option, hard to find off the shelf modules.

I think this is the library I had found before.
https://github.com/MichaelJonker/HardwareSerialRS485/wiki

[wallyllama]

https://github.com/dukelec/cdbus_doc
This one has some is interesting, not sure if it does multi master or not. There are a few projects that seem to do multimaster/collision avoidance, the trick it to find one with low enough overhead and automatically retransmits if there is a failure, and for this project isolates this enough that my sensors doesnt need restructuring.

Good wiring is crucial so you can use higher speeds. The common ground wire in addition to the differential signals seems almost required.

[kimot]

https://github.com/gioblu/PJON/wiki

Some works for MySensors done:
https://forum.mysensors.org/topic/3977/transport-based-on-single-wire-pjon-protocol

[kimot]

But PJON uses completely sw bit by bit sending or HW sending with bad bus arbitration.

My idea - combination CAN bus like arbitration and UART message sending.
With ACK possibility.

Time T1 ( two bytes send time for example )

Node sends message and waits T1 time for ACK.
Receiving node must send ACK within T1 interval.

Node, witch needs send message listens bus.
BUT not by HW UART or SoftwareSerial - use directo IO operation.
If recognize bus free time greater then T1, it start sending its own ID.
BUT again not by HW UART or SoftwareSerial - but bit by bit by direct IO pin operation.
And listen if bit I am sending is bit I am receiving.
If not - it means, that other node with higher priority ( lower ID ) is sending - then immediately stop sending other bit and wait for next bus free T1 interval.
If yes - bus is free for me end send my message normally by UART ( HW or SW, maybe with 9 bit addressing support - time less than T1 between characters allowed).
Wait max T1 time for ACK.

CAN bus driver usage TJA 1050 or MCP2551 - no DE RE pins.

[Andreas Sieber]

I have the problem that I always get this error: !TSM:FPAR:NO REPLY

Once it worked briefly yesterday.

What have I tried everything?
All hardware exchanged - without success.

Now I have just written two small scripts that send RS485 back and forth. without problems. So wiring has to fit. I spent then in the serial montior the data which the node sends me, there comes some. Then I once tried to display the data that sends the GW, because nothing comes.

Does anyone have an idea why my GW does not answer?

[mfalkvidd]

@andreas-sieber could you post the debug output from the node and the gateway?

[Andreas Sieber]

---

| / |_ / | ___ _ __ ___ ___ _ __ ___
| |/| | | | _ \ / _ \ _ \/ __|/ _ \| _/ __|
| | | | || || | / | | _ \ _ | | _
|| ||_, |/ ___|| ||/_/|| |/
|__/ 2.3.0-alpha

17 MCO:BGN:INIT NODE,CP=RSNNA---,VER=2.3.0-alpha
26 TSM:INIT
27 TSF:WUR:MS=0
29 TSM:INIT:TSP OK
30 TSM:FPAR
49 TSF:MSG:SEND,255-255-255-255,s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=OK:
2056 !TSM:FPAR:NO REPLY
2058 TSM:FPAR
2075 TSF:MSG:SEND,255-255-255-255,s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=OK:
4083 !TSM:FPAR:NO REPLY
4085 TSM:FPAR
4103 TSF:MSG:SEND,255-255-255-255,s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=OK:
6111 !TSM:FPAR:NO REPLY
6113 TSM:FPAR
6130 TSF:MSG:SEND,255-255-255-255,s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=OK:

0;255;3;0;9;0 MCO:BG
0;255;3;0;9;29 MCO:BGN:INIT OK,TSP=1
0;255;3;0;9;0 MCO:BGN:INIT GW,CP=RSNGA---,VER=2.3.0-alpha
0;255;3;0;9;5 TSM:INIT
0;255;3;0;9;7 TSF:WUR:MS=0
0;255;3;0;9;9 TSM:INIT:TSP OK
0;255;3;0;9;12 TSM:INIT:GW MODE
0;255;3;0;9;15 TSM:READY:ID=0,PAR=0,DIS=0
0;255;3;0;9;18 MCO:REG:NOT NEEDED
0;255;3;0;14;Gateway startup complete.
0;255;0;0;18;2.3.0-alpha
0;255;3;0;9;22 MCO:BGN:STP
0;255;3;0;9;29 MCO:BGN:INIT OK,TSP=1

[gohan]

why don't you use the stable 2.2? Anyway the node is missing the my_node_id define and second it seems the gateway isn't receiving anything, so I guess there is something very wrong in communication

[Andreas Sieber]

in the 2.2 the same problem, i have now updagradet to alpha.

When you have a solution on 2.2 i will downgrade. no problem!

Guys you are good! I had set the mynodeid. But befor i include the my sensors.. that's the mistake..

hour of hour to search... thx!

[Omemanti]

Is it possible, or will it be possible to build a combination between wired and wireless nodes?

[gohan]

one network, one gateway

[Pavel Polititsky]

I just trying to work with examples but my node always transmitting some piece of shit from hardware serial:

 
 __  __       ____
|  \/  |_   _/ ___|  ___ _ __  ___  ___  _ __ ___
| |\/| | | | \___ \ / _ \ `_ \/ __|/ _ \| `__/ __|
| |  | | |_| |___| |  __/ | | \__ \  _  | |  \__ \
|_|  |_|\__, |____/ \___|_| |_|___/\___/|_|  |___/
        |___/                      2.3.0

16 MCO:BGN:INIT NODE,CP=RSNNA---,VER=2.3.0
25 TSM:INIT
26 TSF:WUR:MS=0
28 TSM:INIT:TSP OK
29 TSF:SID:OK,ID=5
31 TSM:FPAR

⸮X⸮⸮w34 TSF:MSG:SEND,5-5-255-255,s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=OK:
2041 !TSM:FPAR:NO REPLY
2043 TSM:FPAR

⸮X⸮⸮w2045 TSF:MSG:SEND,5-5-255-255,s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=OK:
4054 !TSM:FPAR:NO REPLY
4056 TSM:FPAR

⸮X⸮⸮w4058 TSF:MSG:SEND,5-5-255-255,s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=OK:
6066 !TSM:FPAR:NO REPLY
6068 TSM:FPAR

⸮X⸮⸮w6070 TSF:MSG:SEND,5-5-255-255,s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=OK:
8078 !TSM:FPAR:FAIL
8079 TSM:FAIL:CNT=1
8081 TSM:FAIL:DIS
8083 TSF:TDI:TSL

From software serial too

⸮X⸮⸮w

⸮X⸮⸮w

⸮X⸮⸮w

⸮X⸮⸮w

Consequently the gateway can't find any device.
I have tried also with RS485 with different baudrate. Finally I just connect my USB TTL directly to ouput pins, but always same result.
How to fix it?

[rejoe2]

@Pavel-Polititsky
To me this looks like you mixed between-node communication (RS485) with debug output (and maybe also serial output of the gw).
The corresponding serial interface for RS485 should be dedicated to RS485 exclusively.