Why do I get Error-22011 when I try to run a HW Watchdog Timer vi?

[Graziano]

Hi pmac,
   I've never done something like this.... but I think that you can do it easilly this way:

   Your host has to run a separate application, sending a ping via ethernet to cFP, say, each 500 ms.  In your application, you set a loop to acquire that ping, if you don't receive it for some seconds (fine tune this timing according to your needs), you make application return to a safe state (ie.: set a safe IO state) and stop application, or enter it a separate waiting state. (this is a kind of implementing a network watchdog)

    Actually, this is a really silly idea, butr quite easy to implement. 


    Just a question: when application will  be embedded, you will still control it from you PC someway?

   Have a nice day!

graziano

[RavensFan]

Hello pmac,

I think Graziano has a good idea.  But you may want it set up the other way.  (Or both ways at once).  If the critical code is on the cFP, have the cFP do the pinging to the host PC.  Then if it detects a lost connection, the cFP would be able to shut down the critical code.   If you set it up the other way, the host PC will be able to know the connection was lost, but it wouldn't be able to reset anything on the cFP because the connection was lost. 

[RavensFan]

For some reason there is a slow connection this morning and it submitted the above post twice.

Message Edited by Ravens Fan on 10-15-2007 09:35 AM

[TreyB_NI]

---

@pmac wrote:

My interest in the watchdog vi's originated when I decided it might be useful to know if the communication link is lost. Several documents I looked through stated that the network watchdog feature set does not support the FP2010 network controller. This led me to the hardware watchdog vi's.

pmac

---

Which documents are saying that it isn't supported?  On page 51 of the FP-2010 controller's manual, it reviews how to setup the watchdog timer with FieldPoint Explorer.  FP Explorer was used a years ago before FieldPoint control was integrated into Measurement and Automation Explorer (MAX).  To set it up in MAX, navigate to remote systems, as seen in the attached image.  If you hold your mouse over the watchdog timer, the help in the bottom left appears as in the image, and you can click on the circled link for instructions.

Trey B | Applications Engineering | National Instruments

[Graziano]

Hi Trey B,
   I didn't know that module, and I didn't investigate.... but anyway, I think it's a good idea to implement a self made watchdog, mainly to set a "safe state", with a safe IO configuration.  Actually I don't use cFP controlled by PC, so I don't use ethernet WD, is it possible to set a "safe state" when Network fails? That can be interesting (for me, later, for post's author.... today!).

   have a nice day!

graziano

[pmac]

Trey B.,

On that same page there is a note that states, "National Instruments recommends leaving the network watchdog disabled when you are running embedded applications on the FP-20xx."

Also, in MAX help for Fieldpoint under the heading "Configuring the Network Watchdog" there is a caution that states, "The network watchdog is not compatible with embedded applications on a Labview RT controller such as the [c]FP-2xxx. National Instruments recommends that you use the hardware watchdog functionality available in Labview RT. ..."

Based on the above I started looking at the hardware watchdog vi's. Unfortunately I haven't any success with that approach so I am currently following the advice of graziano and RavensFan by setting up a ping command using System Exec.vi. So far, this approach seems to work though I haven't fully implemented it yet.

I welcome any suggestions.

pmac

[Graziano]

Hi pmac,
   It's quite logical, as Tray said, that Network WD is not compatible with embedded mode, because if FP runs an embedded application, it isn't controlled by the PC, so it would be unnatural for it to stop if TCP/IP fails.  This behaviour could be dangerous in most cases, also in your.

   The hardware WD is something different from what you're trying to do, as I understand it.  HW WD simply RESTARTS controller if application hangs.  Of course, you can use it in a way compatible to your needs: for ecample you can "feedthe dog" only when a ping from the host is received, and if no ping is received, you reset controller, but setting first the IO to a safe state.

   I hope this can help.  Have a nice day!

graziano