💬 Sensebender Gateway

[alexsh1]

OK, I did manage to get debug working partially.
The gateway reads messages, but does not work as the gateway

0;255;3;0;9;MCO:BGN:STP
0;255;3;0;9;MCO:BGN:INIT OK,TSP=1
0;255;3;0;9;TSF:MSG:READ,10-10-0,s=1,c=1,t=17,pt=5,l=4,sg=0:348
0;255;3;0;9;TSF:MSG:READ,26-10-0,s=3,c=1,t=24,pt=5,l=4,sg=0:51219479
0;255;3;0;9;TSF:MSG:READ,10-10-0,s=1,c=1,t=24,pt=5,l=4,sg=0:51219482
0;255;3;0;9;TSF:MSG:READ,11-11-0,s=2,c=1,t=37,pt=7,l=5,sg=0:1717.62
0;255;3;0;9;TSF:MSG:READ,11-11-0,s=0,c=1,t=37,pt=7,l=5,sg=0:5338.44
0;255;3;0;9;TSF:MSG:READ,11-11-0,s=3,c=1,t=37,pt=7,l=5,sg=0:10.75
0;255;3;0;9;TSF:MSG:READ,11-11-0,s=3,c=1,t=37,pt=7,l=5,sg=0:10.76
0;255;3;0;9;TSF:MSG:READ,11-11-0,s=4,c=1,t=37,pt=4,l=4,sg=0:880
0;255;3;0;9;TSF:MSG:READ,6-6-0,s=2,c=1,t=1,pt=7,l=5,sg=0:64.8
0;255;3;0;9;TSF:MSG:READ,6-6-0,s=0,c=1,t=4,pt=7,l=5,sg=0:1015

I never had a message:

Gateway startup complete.

EDIT: Self-test with SensebenderSerial sketch gives me this:

Sensebender GateWay test routine
Mysensors core version : 2.1.1
GateWay sketch version : 0.2
----------------------------------

- > SHA204 Failed to wakeup SHA204
- > SD CARD SD CARD did not initialize!
 -> EEPROM PASSED
-> analog : 38 Failed

[tbowmo]

@alexsh1

Could you check the soldering of the samd?

SD card fails, if you haven't inserted a card, when running the self test.

[alexsh1]

@tbowmo

I do not have the SD card so that's fine

The samd looks fine - no bridges etc
I can try to re-solder it with a hot fan
Do you think samd may be at fault even though the serial GW is working?

Unfortunately, I do not have another Ethernet w5100 to try it. Probably have to order a new one

[tbowmo]

@alexsh1

I know that @Anticimex had problems with one of the pre-production samples that we got from China, with various problems, that was solved with "reflowing" the samd with a hot air gun.. So it was just a thought, that it could be a bad solder joint..

The W5100 is using it's own dedicated SPI port, so even though that the radio works, doesn't mean that there couldn't be any solder problems around the samd pins for W5100.

We have 3 different SPI ports in use on the gateway, one for the MysX connector, one for radio / sdcard, and one for W5100..

[alexsh1]

@tbowmo I do not have the oven to reflow the whole board, but can mount / unmount or re-work certain components 0805 or 0603. Below this it is a pain in the a***

Another observation - The GW behaves like this with or without W5100 mounted.
Seems like an SPI problem to me. How can I wire W5100 to MYSX please and what shall I can in the sketch please? Thank for your help - very much appreciate it

[Anticimex]

@alexsh1 The ATSHA is not connected to the SPI and is totally unrelated, so I am also inclined to suspect that perhaps one side of the SAMD is not entirely in contact with the pads. I had to hot-airgun mine to get USB working well but it was an early prototype like @tbowmo said.

[alexsh1]

@tbowmo OK, I manage to find the answer here -

MYSX
D11 - MOSI
D12 - MISO
D13 - SCK
D14 -CS

#define MY_SOFTSPI
#define MY_SOFT_SPI_SCK_PIN 13
#define MY_SOFT_SPI_MISO_PIN 12
#define MY_SOFT_SPI_MOSI_PIN 11
MY_RF24_CS_PIN 14
RF24_CE_PIN ???

Am I correct?

[alexsh1]

My current config for nrf24l01+ and w5100 is as follows:

#define MY_GATEWAY_W5100

#if !defined(MY_W5100_SPI_EN) && !defined(ARDUINO_ARCH_SAMD)
#define MY_SOFTSPI
#define MY_SOFT_SPI_SCK_PIN 14
#define MY_SOFT_SPI_MISO_PIN 16
#define MY_SOFT_SPI_MOSI_PIN 15
#endif

#define MY_RF24_CE_PIN 5
#define MY_RF24_CS_PIN 6

Why CE / CS pins are 5 and 6?
Should that be 27 and 3?

[alexsh1]

@tbowmo I have just re-worked SAMD and all the connectors (nrf24l01+ and NET_SPI), but still have the same issue sadly.

EDIT: If I change to rfm69, the same behavior is observed. It seems that Net_SPI is just not working.

[tbowmo]

@alexsh1

Configuration of the pins are handled by the hardware configuration file, https://github.com/mysensors/ArduinoHwSAMD/blob/master/variants/mysensors_gw/variant.h#L227

If you look in the gatewayW5100 sketch, the settings for CE / CS pins are enclosed in ifdefs, so they are only used if they are not defined previously.

If you want to connect the W5100 to the mysx connector, you can change line 174 in the hardware definition file, use SPI2 instead of SPI1 for SPI_ETH

It could also be a defective W5100 shield? Can you verify that it works on other hardware?

[alexsh1]

@tbowmo Swapping SPIs did not resolve the problem. I am now getting another W5100 to test further.

[alexsh1]

Unfortunately, a new W5100 shield did not work - it seems like a production issue with this sensebender gatewsay as the SPI2 did not work either with W5100.
Will get a new one ordered.

[tbowmo]

Is it possible for you to check the spi wires with a scope? Just to make sure if mosi/miso/ce/Ss lines are toggling as expected?

[tbowmo]

Another thing, are you using static ip? If yes, can you ping it? If no, does your DHCP server hand out an ip address?

Have you tried the basic ethernet examples for the W5100?

[alexsh1]

@tbowmo I am using the static IP and I can ping it, but the "Gateway Ready" message never comes up. Also I can only see the neg24l01+ log (incoming messages only)

[tbowmo]

@alexsh1

If you can ping it, then the W5100 is configured. Something else is wrong..

One thing that comes to mind, if you compile it as an ethernet GW, then it does not wait for serial console to be active, before initializing the radio. If you compile it as serial GW, then it waits with initialization, until serial is active.

This means that you probably is missing the GW ready message on the serial port, as it is "sent" before you have an active serial connection.

Look in MyHwSamd.cpp in the HwInit routine, remove the #ifdef around the while() loop, and try again.

[alexsh1]

@tbowmo Well, if I try the MEGA 2560 with W5100, I get the following:

0;255;3;0;9;MCO:BGN:INIT GW,CP=RNNGAS-,VER=2.1.1
0;255;3;0;9;TSM:INIT
0;255;3;0;9;TSF:WUR:MS=0
0;255;3;0;9;TSM:INIT:TSP OK
0;255;3;0;9;TSM:INIT:GW MODE
0;255;3;0;9;TSM:READY:ID=0,PAR=0,DIS=0
0;255;3;0;9;MCO:REG:NOT NEEDED
IP: 10.1.1.120
0;255;3;0;9;MCO:BGN:STP
0;255;3;0;9;MCO:BGN:INIT OK,TSP=1

I never get this message with Sensebender GW. I only get

0;255;3;0;9;TSF:MSG:READ,11-11-0,s=2,c=1,t=37,pt=7,l=5,sg=0:2019.12
0;255;3;0;9;TSF:MSG:READ,11-11-0,s=0,c=1,t=37,pt=7,l=5,sg=0:6275.51
0;255;3;0;9;TSF:MSG:READ,11-11-0,s=3,c=1,t=37,pt=7,l=5,sg=0:18.77

or similar messages from nodes.

[tbowmo]

@alexsh1

on atmega2560 you are resetting the device, when you connect a serial terminal to the usb device, so you get all the debug messages on the serial port..

On SAMD it uses a native USB port, and the chip is not reset when connecting a serial terminal. When using serialGw it will wait until a serial terminal is connected. before initializing radio transport (this is because it does not make sense to have a serial gateway that spits out data to a none existing port).

When using it as a ethernet gateway, we skip this, as we do not want to wait for a serial terminal, in case you are just using the USB as a power source. So in this mode, it skips the waiting, and just initializes radio gateway etc.

as stated in my previous message you could look at the MyHwSAMD.cpp file line 73. Remove the #ifdef recompile, upload, and check your serial terminal again.

Also the log shows that you actually are receiving data from your nodes, so the radio transport is clearly active.. And you can ping your gateways static IP, which means that the W5100 is initialized with the correct IP address.

[tbowmo]

@alexsh1

Did you try my suggestion about hacking the core? Or did you give up on the beast? :)

I'm curious to hear if it is solved..

[alexsh1]

@tbowmo I will try it tomorrow. However, given the amount of time I spent already, I am about to give up...