Real-Time Measurement and Control

cancel
Showing results for 
Search instead for 
Did you mean: 

Modbus TCP protocol violation

Ryan,

 

When you say the Master's address do you mean the slave address configured in the Master?

 

See attached picture grab of using the Modpoll utility. The slave is configured as address 1 and IP address 192.168.225.12. When polling registers 400-403 at address 1 (-a 1), it works fine.

 

However, polling at address 2 (-a 2), I get a timeout. This should not be the case. I should get a response regardless of the slave address for reasons I have outlined in earlier posts.Capture.PNG

0 Kudos
Message 11 of 21
(3,442 Views)

Yes, I mean the slave address as configured in the master.

 

I'm not sure what's going on there. I set up a master on my PC and configured the slave address on the master to be 2, and I set up a slave to run on a RIO device configured with a slave address of 1 and I'm able to establish communication.

 

What happens if you set up a master in LabVIEW and communicate with the slave from there? Does that yield the same results?

 

Regards,

Ryan K.
Product Manager, ATCA and BEEcube
National Instruments
0 Kudos
Message 12 of 21
(3,427 Views)

Ryan,

 

Tried using a Modbus master created within LabVIEW.

 

Slave was configured to address 1.

 

Verified using both Distributed System Manager and Wireshark that the slave does not respond when Master is set to talk to slave address 2. Works fine when Master is set to talk to slave address 1.

 

 

0 Kudos
Message 13 of 21
(3,415 Views)

That is odd. I've been doing some additional testing as well and while the code I used for my initial tests still works properly - I can establish communication between the master and the slave even though the address are mismatched - when I try to build new tests with mismatched addresses, the tests fail. It does seem like something odd is going on here with the Real Time module. I've passed this issue on to R&D, and if I receive any information from them about what's going on, I'll pass it along to you.

 

In the meanwhile, the best thing I can recommend is that you continue to use your current workaround. It will likely take some time for me to receive any information from R&D, but I'll post here if there are any developments.

 

Regards,

Ryan K.
Product Manager, ATCA and BEEcube
National Instruments
0 Kudos
Message 14 of 21
(3,400 Views)

Ryan,

 

When do you think you can get answers/resolutions for this? I have to answer to our customer. Do you think opening a service request will expedite this? (We have an SSP).

0 Kudos
Message 15 of 21
(3,376 Views)

Sorry for the delay responding. I've passed this issue on to R&D and they've acknowledged that this does appear to be an issue with the way that LabVIEW handles Modbus Ethernet. In terms of a time frame for a resolution, the best answer I can give you is that R&D is aware of the issue and they will investigate the matter and look into resolving it in a future version of LabVIEW. In the meanwhile, the best thing I can recommend is that you continue to use your existing workaround. It's unlikely that openining a service request would expedite a resolution, since the matter is in R&D's hands, but you're certainly welcome to open a service request if you want to.

 

Wish I could be of more help,

Ryan K.
Product Manager, ATCA and BEEcube
National Instruments
0 Kudos
Message 16 of 21
(3,358 Views)

Ryan,

 

Thank you for the feedback. I hope we will know more soon.

0 Kudos
Message 17 of 21
(3,354 Views)

Sorry about the bump.

 

Ryan,

 

Do you have the latest on this? Was the Realtime module updated with the new LabVIEW releases and if so, was this bug fixed?

 

Thanks

0 Kudos
Message 18 of 21
(3,218 Views)

To my knowledge, R&D didn't get a chance to address this issue in the current release in LabVIEW 2014, so the issue hasn't been fixed yet. However, the 2014 Real Time module did introduce a new Modbus API; if you wanted to test that out, you could always download an evaluation version of LabVIEW 2014 and the 2014 Real Time module and see if that would work better for your application.

Ryan K.
Product Manager, ATCA and BEEcube
National Instruments
0 Kudos
Message 19 of 21
(3,203 Views)

Thank you Ryan, I am certainly interested in learning more about the new API. Is there a white paper on this?

0 Kudos
Message 20 of 21
(3,191 Views)