Communication problems v2.0.0 between GW and nodes.

[bisschopsr]

Hi All,

After stugling with V2.0.0 for a few days now, it's time to ask for help (again). I had troubles with radio's before, those are supposed to be fixed now.
I have A GW with W5100 (nano) and two sensors (Uno and Pro Mini) running. I changed the channel to 1 (#define MY_RF24_CHANNEL 1), as I also have a 1.5 network running on the standard channel 76. That is all I changed in the standard scripts.
The is no controller attached yet.

I now get the following log from the gateway (taken from MYSController):

[2016-08-30 14:44:27.217 Info] INFO	Connected to COM3
 [2016-08-30 14:44:28.705 Info] RX	0;255;3;0;9;Starting gateway (RNNGA-, 2.0.0)
 [2016-08-30 14:44:28.706 Info] CHILD	New child discovered, node id=0, child id=internal
 [2016-08-30 14:44:28.706 Info] RX	0;255;3;0;9;TSM:INIT
 [2016-08-30 14:44:28.706 Info] RX	0;255;3;0;9;TSM:RADIO:OK
 [2016-08-30 14:44:28.706 Info] RX	0;255;3;0;9;TSM:GW MODE
 [2016-08-30 14:44:28.707 Info] RX	0;255;3;0;9;TSM:READY
 [2016-08-30 14:44:29.008 Info] RX	IP: 192.168.0.211
 [2016-08-30 14:44:29.008 Info] ERROR	ErrorFlag=255, Msg=IP: 192.168.0.211
 [2016-08-30 14:44:32.017 Info] RX	VMDPV_1|1_VMDPV
 [2016-08-30 14:44:32.017 Info] ERROR	ErrorFlag=255, Msg=VMDPV_1|1_VMDPV
 [2016-08-30 14:44:32.017 Info] RX	0;255;3;0;9;No registration required
 [2016-08-30 14:44:32.017 Info] RX	0;255;3;0;9;Init complete, id=0, parent=0, distance=0, registration=1
 [2016-08-30 14:44:36.381 Info] RX	0;255;3;0;9;TSP:MSG:READ 1-1-255 s=255,c=3,t=7,pt=0,l=0,sg=0:
 [2016-08-30 14:44:36.382 Info] RX	0;255;3;0;9;TSP:MSG:BC
 [2016-08-30 14:44:36.382 Info] RX	0;255;3;0;9;TSP:MSG:FPAR REQ (sender=1)
 [2016-08-30 14:44:36.382 Info] RX	0;255;3;0;9;TSP:CHKUPL:OK (FLDCTRL)
 [2016-08-30 14:44:36.382 Info] RX	0;255;3;0;9;TSP:MSG:GWL OK
 [2016-08-30 14:44:36.962 Info] RX	0;255;3;0;9;!TSP:MSG:SEND 0-0-1-1 s=255,c=3,t=8,pt=1,l=1,sg=0,ft=0,st=fail:0
 [2016-08-30 14:44:38.447 Info] RX	0;255;3;0;9;TSP:MSG:READ 1-1-255 s=255,c=3,t=7,pt=0,l=0,sg=0:
 [2016-08-30 14:44:38.447 Info] RX	0;255;3;0;9;TSP:MSG:BC
 [2016-08-30 14:44:38.447 Info] RX	0;255;3;0;9;TSP:MSG:FPAR REQ (sender=1)
 [2016-08-30 14:44:38.449 Info] RX	0;255;3;0;9;TSP:CHKUPL:OK (FLDCTRL)
 [2016-08-30 14:44:38.449 Info] RX	0;255;3;0;9;TSP:MSG:GWL OK

The radio gives the following log (from direct serial log):

Opening port
Port open
Starting sensor (RNNNA-, 2.0.0)
TSM:INIT
TSM:RADIO:OK
TSP:ASSIGNID:OK (ID=1)
TSM:FPAR
TSP:MSG:SEND 1-1-255-255 s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=bc:
TSM:FPAR
TSP:MSG:SEND 1-1-255-255 s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=bc:
TSM:FPAR
TSP:MSG:SEND 1-1-255-255 s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=bc:
TSM:FPAR
TSP:MSG:SEND 1-1-255-255 s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=bc:
!TSM:FPAR:FAIL
!TSM:FAILURE
TSM:PDT 

As you can see failures on both ends. As the two nodes show the same behaviour, I suspect the problem lies with the GW.
Does anyone have a hint where to look.
Thx

Ralph

[tekka]

@bisschopsr Looks like a power issue (no or too small cap, add at least 100uF) or radio interference (radios too close, AP nearby)

[bisschopsr]

@tekka Thx for the response. I will look into both in more details.
As for power: This is always a critical one, both gateway and nodes are power with a solid supply that I used before on all my nodes, so I don't expect much issues there. But I must say that I don't use 100uF as cap, so will see what that does.
As for interference: The GW and node are on one desk, so that might be a problem. I did not have this problem playing with earlier versions of MySensors, but since this is V2, there might be a change in comms that is of influence. Also there is GW on channel 76 nearby. This setup is on channel 1, so I don't know if the other GW gives interference. AP’s are all on a distance.

[bisschopsr]

@tekka I more and more start to believe i'm dealing with a crappy Nano. Whatever power I use, I noticed it is assining a very strange IP address that is changing all the time. Also when I put it in my (exisiting) gateway board, it's getting hot.

[bisschopsr]

This indeed looks like a bad Nano. I rebuild it with a spare UNO and working. There is however one thing. The W5100 can be powered on 5V right? I do so on my "old" gateway. The W5100 does have its own step down regulator. This one is not working well at 5V. It's assigning the strangest IP adresses on 5V. On 3.3V it works ok. Anyone experienced this also, or is this also a bad W5100?

[TheoL]

@bisschopsr Haven't experienced it. But whilst preparing Fridays workshop I did notice one thing. If you use the shortest Dupont cables you can buy you don't need capacitors perse. It's ofcource even better to have a cableless connection like soldering the radio directly on the pcb - or use headers.

Not sure how it will work out Friday, so I'll bring some capacitors with me. But I was actually able to build nodes without capacitors that remained communicating over at least 20 or 30 meters. For this we use 10 cm dupont cables.

So I've added an extra reason for communication failures to my list. A too long distance between the radio and arduino to which the radio is connected.

[bisschopsr]

@TheoL Hi Theo, thanks for your feedback. Interesting workshop you do on Friday, unfortunately I can't participate due to work obligations. Regarding the cables, I understand your experience. When using Dupont cables, it's always good practice to keep the connections as short as possible. Capacitors are always a good addition. In designing electronics, its standard practice to at least add decoupling capacitors (generally around 100nF) closed to the power connections of your integrated circuits or modules. For the radio's in MySensors I would advise to add a larger capacitor of 4.7uF or 10uF over the power connections of the radio, preferably soldered on the radio module itself. I think this is also something to share with the participants coming Friday as design rules. Regarding the performance of mySensors V2.0.0, I have my thoughts. Version 1.5 was running flawless out of the box, V2.0.0 does not in my experience. Like to hear the thoughts and experiences you have especially after the workshop.

[TheoL]

@bisschopsr You're of course right, I was just a bit surprised to see it work. But I don't think it'll work great without capacitors on the long run.

[bisschopsr]

@TheoL Agreed, better safe then sorry :-). I'm interested to see how the sensors will perform if they are close together (they will probably be in a class room kind of situation). I found that this is one of my biggest pitfalls. 1.5 was always ok, even on the same work bench. 2.0 is very critical in this in my experience. For myself I wonder how this is going to work in real life situations where sensors might be close range to eachother and to the GW. If have not enough sensors now to test this on 2.0 unfortunatly, but you will have I guess some during the workshop.

[xydix]

@bisschopsr
Have you solved your problem?
I have nodes running fine on 1.5 but i have problem with 2.0
I get the same as you in my serial monitor from my GW and nodes.
My issue was (still is) that i use pro minis with voltage regulator removed and a battery booster.
My battery booster give me 3,36V -3,38V.
This cause problems under the presentation on 2.0.
If i get lower voltage, closer to 3,3V the problem is gone.
It turns out that i little to high voltage can cause this.

[rubyan]

I had lots of problems with receiving the ACK messages of a node. I was sending a switch off command with ACK. The ACK was never delivered 50% of the time.

 [2016-09-07 17:16:44.241 Info] RX	0;255;3;0;9;TSP:MSG:SEND 0-0-100-100 s=1,c=1,t=2,pt=0,l=1,sg=0,ft=0,st=ok:0
 [2016-09-07 17:16:44.241 Info] RX	0;255;3;0;9;TSP:MSG:READ 100-100-0 s=1,c=1,t=2,pt=0,l=1,sg=0:0
 [2016-09-07 17:16:44.241 Info] RX	100;1;1;1;2;0
 [2016-09-07 17:16:44.490 Info] TX	100;1;1;1;2;0
 [2016-09-07 17:16:44.537 Info] RX	0;255;3;0;9;!TSP:MSG:SEND 0-0-100-100 s=1,c=1,t=2,pt=0,l=1,sg=0,ft=0,st=fail:0
 [2016-09-07 17:16:44.755 Info] TX	100;1;1;1;2;0
 [2016-09-07 17:16:44.771 Info] RX	0;255;3;0;9;TSP:MSG:SEND 0-0-100-100 s=1,c=1,t=2,pt=0,l=1,sg=0,ft=0,st=ok:0
 [2016-09-07 17:16:44.771 Info] RX	0;255;3;0;9;TSP:MSG:READ 100-100-0 s=1,c=1,t=2,pt=0,l=1,sg=0:0
 [2016-09-07 17:16:44.771 Info] RX	100;1;1;1;2;0
 [2016-09-07 17:16:48.052 Info] TX	100;1;1;1;2;0
 [2016-09-07 17:16:48.099 Info] RX	0;255;3;0;9;!TSP:MSG:SEND 0-0-100-100 s=1,c=1,t=2,pt=0,l=1,sg=0,ft=0,st=fail:0
 [2016-09-07 17:16:48.899 Info] TX	100;1;1;1;2;0
 [2016-09-07 17:16:48.946 Info] RX	0;255;3;0;9;!TSP:MSG:SEND 0-0-100-100 s=1,c=1,t=2,pt=0,l=1,sg=0,ft=0,st=fail:0
 [2016-09-07 17:17:00.021 Info] TX	100;1;1;1;2;0
 [2016-09-07 17:17:00.068 Info] RX	0;255;3;0;9;!TSP:MSG:SEND 0-0-100-100 s=1,c=1,t=2,pt=0,l=1,sg=0,ft=0,st=fail:0
 [2016-09-07 17:17:00.614 Info] TX	100;1;1;1;2;0
 [2016-09-07 17:17:00.645 Info] RX	0;255;3;0;9;TSP:MSG:SEND 0-0-100-100 s=1,c=1,t=2,pt=0,l=1,sg=0,ft=0,st=ok:0
 [2016-09-07 17:17:00.645 Info] RX	0;255;3;0;9;TSP:MSG:READ 100-100-0 s=1,c=1,t=2,pt=0,l=1,sg=0:0

When I removed the (aliexpress) NRF24L01+ Socket Adapter Board on the node and connected the radio directly to the 3.3v nano the acknowledge rate went to 100%. Strange...

[bisschopsr]

@xydix NRF modules are sensitive to power fluctuations, so take care that you have a stable source. Is the booster of good quality or producing noise on the power lines?. The converter can produce a high frequency signal on the powerline, also know as noise. You can try to filter the noise with a small capacitor like 100nF or maybee even smaller depending on the freqency of the noise coming from the booster.
My problem was/is the distance between the GW and the sensor currently, other problems were related to power as well, solved for now (see above my findings)

@rubyan Sounds like power trouble and/or distance between GW and sensor. Powering from the Nano is in genaral not a good thing. The supply might be to week. Use an external power source and some capacitors (see above) Also try to see if putting the GW and sensors at a distance helps.