M340 READ_VAR block freeze

[Jacek_P]

Hello,

We have situation with M340 2020 SV3.01 that Read_Var blocks freezes.

We have set time out to 500ms and we trigger block on the edge of %s6 with conjunction of lack of presence of activity bit.

When the block freezes it does not provide any error message.

We have tried to replace the block with another READ_Var function, Cancel operation, reset GEST variable to zero by hand upgrading Libset.

The only way to restore it to live is to change variable connected to GEST pin.

It generally takes three weeks of normal operation and after this time block is frozen.

Is there any solution against Read_var block freezing.

[MatthewM]

Sounds like you are doing it properly.

Is the Activity bit high when it freezes?  I assume you see that the high byte of the first GEST word stop incrementing and the Activity bit looks to be high all the time?

What you describe sounds like a classic state when the timeout is not set, but you say you are setting the timeout.  Please check that you are indeed setting the 'Timeout' (3rd word) in the GEST structure, and not one of the other words by accident.

I don't think this is your issue as you have no error messages, but check the number of open connections that the port has.  I can't remember exactly where to find that information sorry but the M340 can only have 16.  Maybe the READ_VAR is getting locked because there are already 16 connections open?  If that is the case I would expect you would see an error message of some sort, hence I don't think this is your issue.

Pop us a code and animation table snip of GEST.

[Jacek_P]

Hi,

I have checked number of opened connections.

It is visible in the NOE card debug window under Bandwidth button

In my case the number is 13.

I have noticed that in firmware version 3.01 there were some fixes on UMAS communication to CPU

Maybe it is this "feature"

In this moment we have upgraded CPU firmware to the latest version and we wait what is going to happen.

[MatthewM]

I can't comment on the FW version as all the M340s that we have that make extensive use of READ/WRITE_VARs are still FW 2.7.  Keep us posted on what you find.

[Ricard]

Hello Jacek

First of all, I can't see nothing related this issue in the history of the firmware of M340, nevertheless we never can discard update it.

For my experience, I would focus the problem on the application or the system.

Questions:
Does it only happen to you with a specific read_var?
Respond to ping?
Answer Modbus questions? Put a Modbus client on your PC and do some reading.
What version of Unity/Control Expert do you have? Have you updated all the HFs?
What version of libset does the program have?

Tests:
Try testing only the READ_VAR block to see if you reproduce the problem (Disable or delete all code that is not important for communications).
The other test would be to see the communication frames in case we have any problems on the network that affect the system.

If you are unable to draw any conclusions, contact technical support in your country.

BR

Ricard

[StevenCho]

I had a similar issue with the M580 through Modbus TCP (UNITY PRO).
What I did was use the DTM Browser.
After that, the freezing issue did not happen again.

https://www.youtube.com/watch?v=Igih7IFh5to