3 replies to this topic

[kennethone]

Good Afternoon,

 

I've been testing I2C on two netduino plus devices and I found that on both, after a certain time, the I2C port stops outputting any data and the SDA line stays high ~ 5 V for me and the SCK line stay at around 400mV (using digital scope). I can still breakpoint and step through all of the managed I2C code without any errors, but the status of the lines does not change. 

If I leave the device plugged in, and re-deploy the application, the lines stay in the non-functional state. It is only after I unplug and power cycle the device, the lines begin working again.

 

My SDA and SCL lines are pulled up using 4.7K resistors. Device under test is PF575 and functions without problems.

 

This problem also repeats itself when I use the netduino by itself and hook up the SDA and SCL lines to 5V using 2 15K resistors. When I do this, the SDA line does not get ACK response, but the activity is still visible on the oscilloscope. For some time, after which is stops working and requires a power cycle.

 

 

 

 

Is this an issue which may have been recently fixed by a firmware update? Should I update my firmware? Thanks,

 

 

 

 

 

DeviceInfo:

  HAL build info: 4.1.2821.0, Netduino Plus by Secret Labs LLC

  OEM Product codes (vendor, model, SKU): 34, 177, 4097

  Serial Numbers (module, system):

    00000000000000000000000000000000

    0000000000000000

  Solution Build Info: 4.1.0.6, Netduino Plus by Secret Labs LLC

  AppDomains:

    default, id=1

  Assemblies:

    mscorlib,4.1.2821.0

    Microsoft.SPOT.Native,4.1.2821.0

    Microsoft.SPOT.Hardware,4.1.2821.0

    Microsoft.SPOT.Net,4.1.2821.0

    System,4.1.2821.0

    Microsoft.SPOT.IO,4.1.2821.0

    System.IO,4.1.2821.0

    Microsoft.SPOT.Hardware.SerialPort,4.1.2821.0

    Microsoft.SPOT.Hardware.Usb,4.1.2821.0

    SecretLabs.NETMF.Hardware,4.1.0.0

    SecretLabs.NETMF.Diagnostics,4.1.0.0

    I2C test,1.0.0.0

    SecretLabs.NETMF.Hardware.NetduinoPlus,4.1.0.0

[philvr]

Interesting, we have a similar issue eith a Netduino PLUS MK2 . One moment all is well and I can obtain data back from I2C, the next the SDA and SCL lines are always at logic "1" and there is no activity. I have a brand new, unused Neduino Plus 2 which I swapped with and the I2C lines both remained high with no clock or data line actions. Next I simply left two 4k7 resistors between SDA,SCL and +3.3V - again no action. This convinces me the additional hardware ( a MAX6651 fan controleler) is ok and there is some issue with the netduino which causes I2C to cease. I am going to order a Raspberry PI as a blind test - to absolutely prove our circuitry is good and the netduino is flaky.

[Chris Walker]

Hi kennethone, I would recommend trying the latest 4.2 firmware. There have been quite a few I2C bugfixes for NETMF over time. I2C at its core is interrupt-driven, so conceivably an interrupt which doesn't fire could result in an unknown state. If that indeed happens and we can reproduce it, we can implement measures to work around that. Chris

[philvr]

FYI: The i2c issue seems resolved using the latest 4.2 firmware (and I haven't needed to order a Raspberry Pi)