sbRIO, no Ethernet after reboot with USB or SD plugged in

[uwe.parchatka]

My sbRIO-9636 doesn't make any network connection attempts if started with plugged in USB-stick or SD-card.

-          Lamps of LAN don’t lit

-          No communication over LAN possible

-          At power-on self test normal behavior with yellow status LED on for some secs

-          MAX doesn't recognize or connect

-          NI-Network browser doesn't find the module

Operating system: NI Real-Time VxWorks-PPC603 6.3

Static IP                     

Sometimes it works, but very seldom

Have two boards doing the same

I want to:

-          Reach the Module after power failure over LAN

-          Store measured data on the card or USB-stick

Does anybody know how to solve the problem?

Kind regards,

Uwe

[AnMu]

Dear Uwe,

I am not really sure about the relevance between what you are observing and what you want to work.

Is it correct, that the sbRIO is not working at all as soon as you start it up while plugged in with a USB drive or SD card?

So what does this have to do with how to reach the module after power failing and storing the data?

Thank you

[uwe.parchatka]

Dear Anoj Mubarak,

as there is no startup-exe on the sbRIO it does not do anything what you can see. But it might be working perfect only you can't connect to it over LAN.

I have done many tests and got the board running under some strange conditions like:

- crossover cable to switch, which is not usual

- fixed speed rate 10 or 100 MBit on the switch not auto negotiate

At the present state I have send one of the boards to NI to check out the problem.

To me it looks like a problem with the Ethernet communication chip/interface.

I wait now for the response from NI-service.

Regards,

Uwe

[tannerite]

Do you see any network activity on the Ethernet LEDs when the USB or SD card is plugged in? Do you have access to the Console Out (Serial) of the sbRIO-9636? If you do not see Ethernet LEDs when the USB drive is plugged in, how do you have this device grounded? Is COM (negative power supply input of the sbRIO) connected to Chassis Ground? Is Chassis Ground connected to anything? Is it possible that this device was susceptible to ESD?

- Tanner

[tannerite]

Also, what was the Serial Number of the device sent in and what is the Serial Number of the device that is currently seeing this behavior?

[uwe.parchatka]

Dear Tanner,

- As written above the LEDs don't show any activities they are constantly off if the module has the problem.

- Meanwhile it seems not to be a problem of the USB or SD because the actual behavior is independent of these devices. Only when the problems started to appear they showed a dependency of plugged in USB/SD. The Title of this thread may have to be changed accordingly.

- Yes, I have access to the console output. What comes out of this port looks very normal to me.

- The two chassis have different groundings:

    - One was a Evaluation Kit from the "Build Your Own Embedded System - Seminar". It has no ground at all.

    - The second one is mounted inside an air monitor system for measuring NO/NO2/O3 in ambient air.

      This chassis is connected to the power ground with a low impedance connection.

- ESD: I handled the board itself with care. Due to the dry air in the lab and offices there may have been discharges to the air monitor's chassis. I am used to discharge myself before I touch lines inside the monitor.

Let me mention again... Both modules work now, but under some strange but reproducible conditions like:

- I have to switch of the auto negotiation of the connected PC or switch

- mostly it works only if I use a cross over cable, even with the switch. Switch shows errors/warnings in the connection but works.

- not every cross over cable works even if it has the same pinout and both Cat 5

regards,

Uwe

[uwe.parchatka]

Dear Tanner,

here are the serial numbers of the boards:

Eval-board, sent in:  sbRIO 9636, M/N LV RIO EVAL, S/N: 18A67CA (18A0EE1), P/N: 153982A-01L (151866E-12L)

2nd board, still at work:  sbRIO 9636, S/N: 18B775D

Attached is the protocol of some tests I did with the two boards. It's written in German. I can translate it if you find it necessary/helpful.

Regards,

Uwe Parchatka

[tannerite]

Uwe,

With the device you still have that is experiencing this odd behavior, could you please tell me the voltage between chassis ground and COM while the device is powered on? Please have everything else disconnected so it is just the sbRIO being powered. I have attached a picture you can use as a reference:

If COM and Chassis Ground are tied together, externally, and the polarity of the power connector is reversed, damage could occur to the protection diodes on USB and Ethernet. Is there a possibility that this happened with your device?

Do you have other devices that are functional and do not experience this behavior?

[uwe.parchatka]

Tannerite,

sorry, did not recognize your message until now.

I have only two of these devices. One (Evaluation Kit) is sent to NI with RMA. The other one is running in the lab.

Pretty sure that polarity of the power plug was always right. The Evaluation board had no extra grounding connection to the power GND as it was delivered together with a mated power supply ready to start.

The one I still have here in the lab is connected to the instrument with many lines. It is not good to stop working at the moment and disconnect everything.
The measured DC voltage between the power GND and the chassis GND is about 7 mV during full operation.

Regards,

Uwe

[tannerite]

7mV is expected. Is the device you currently have in the lab working properly at the moment? Can you see if the FPGA is extremely hot? I do not recommend touching it but if you can measure that temperature at it is above 85-100C, I would say there is an internal short in the FPGA due to DIO overvoltage or ESD. I do not think this is the case, however.