AB7697-F Configuration Help

Anybus Support Anybus Gateways
#26

So that process produced a much larger ESI file (233KB vs 3KB). I can still see/connect to the gateway with it. It added in an for each of the 128 bytes of Rx and Tx. It also added a large <Dictionary><DataTypes> section with a lot of datatype names and bitsizes. It also added PdoUpload="1" CompleteAccess="0" to the <CoE> tag list near the end of the file (in addition to SdoInfo="1" PdoAssign="0" PdoConfig="0" that was already in the base ESI file). I’ll attach a copy here, you can compare it at greater length to the base ABX_V_3_29.xml yourself. HMS Industrial NetworksESI.xml (232.2 KB)

At any rate, this seems to have worsened my problem. I can see the device, I can tell when it’s turned off, but now I can’t transition from Idle to pre-op, it doesn’t even throw an error, just nothing happens when I hit the transition button.

#27

Sorry for the delay. I’m going to escalate the case so you will get an email from our ticketing system. I should be able to get you some information on Monday morning and some other things to try.

#28

I was trying to explain the LED status’ to a colleague and from what you’ve said, it sounds like the RUN LED never lights up? I think this is a problem with the controller failing to communicate the operation mode correctly to the gateway. Are there any other diagnostics that you can give me from the controller?

#29

No, see message 12 from Aug 4. The 3rd LED does light up with ~50% duty cycle (maybe 1-1.5 Hz?) when I transition from idle to PreOp. At the time I had misidentified that as the ERR LED due to the ordering in the documentation, but the order is reversed on the label of the actual product, so that was the RUN LED. The ERR LED lit up at the same time I got the error from my controller when trying to transition to safe op. ERR LED would go back out either when I issued the ‘clear error’ command or when I transitioned back to idle.

#30

Any updates?

#31

Have you just tried using the OMRON file and changing your data size to 256/256?

Have you tried using this master with a different slave device?

It really appears to be an issue with the master not being able to change the operation mode of the slave. I would recommend working with NI to find out what the error messages are caused by and why this is not working. There isn’t much else we can try from the slave point of view.

#32

The OMRON file has the Outputs ControlByte set to x24 and not x64 like the ‘master’ ESI file. Which is correct? It seems like through some brief experimentation, that I cannot even transition from idle to pre-op if the ControlByte on that line is set to x24.

Also, the OMRON file defines PDO’s for 128 Rx and 128 Tx, not 256/256, which should I be using?

I have not used this particular controller with another Ethercat slave since this is the only Ethercat device I’m trying to talk to. However, I have used this controller model with other Ethercat slaves in the past (mostly motor drives) as have my coworkers on at least a dozen projects without this issue.

The error message indicates that it’s the slave that has thrown an emergency. I will reach back out to NI in order to try to get the exact technical reasons why that error would be thrown.

#34

So this took a while, but we finally got some Wireshark traces of the interaction between the gateway and the cRIO Ethercat master.

In packet 486 of the trace, the Anybus gateway communicates back that the Mailbox data length is incorrect (Expected:10 Actual:0).

It looks like that was preceded by a bunch of successful 1e00 mailbox queries, but it was immediately after the SafeOp request (packets 483&484).

Can you help me locate what is going wrong using this information?

Thanks,
Mike

AnybusGateway_TransitionFailureTrace.pcapng (66.8 KB)

#35

I am going to open a case in our ticketing system so that I can escalate it to the proper engineers. You will get an email about a new case being opened.

#36

Hi @mwebster,

I checked with a colleague and he confirmed some things:

It appears that the EtherCAT master sets the size of sync manager 2 & 3 to 0 and disables them, this causes it to become impossible to enter the SAFE-OP state.

It might be because of incorrect PDO mapping in the ESI file.

#37

I submitted a reply on the mysupport.hms.se site with the support ticket you opened on November 9th, but received no reply back. I redid the ESI file in several configurations and submitted the Wireshark capture on that ticket. Reference 202010-22329
Should I still be expecting a reply there or do I need to funnel everything through this forum?

#38

I think you may have already tried this, but I was looking at another thread where a customer had a similar issue to yours and was able to resolve it:

In your case though, it seems there may possibly be an additional issue with the Sync manager. From the pcaps we can see the EtherCat controller sets the size of sync manager 2 & 3 to 0 and disables them.

The error message you got translates to this:

ESM Transition Error

  • Transition PRE-OPERATIONAL to SAFE-OPERATIONAL not successful.
  • “Sync Manager 3” Settings Error, Settings of the Sync Manager channel are not matching.

The “64 00 64 00” portion of the emergency data further describes the “sync Manager Settings Error” which can be interpreted from ETG1000.6 specifications table below:

MicrosoftTeams-image

Do you have any documentation from NI that explains how the Sync Manager setting is handled?

#39

Even better - Can you export an ENI file from the controller?

#40

I forwarded your previous email to NI to see if they have a response. I don’t know what an ENI file is.

#41

It’s the EtherCAT Network Information file that describes the full configuration of the network that the master will apply, and the actions to take on each state transition. We should be able to find out why it’s setting Sync Manager 2 and 3 to 0 with it.

#42

I don’t see anything in the NI literature about being able to produce an ENI file and only one post asking if it’s possible to import one, which it wasn’t. I’ll forward the question along though.

#43

Any response from NI?

#44

They sent this last Thursday:

The sync managers are not manipulated in any way on our controllers besides reading the configuration data in from the ESI file and saving it to memory. Also, there are no ENI files that are produced from our NI controllers since the EtherCAT functionality of our controllers is to read in the ESI files and communicate with the desired device using the provided configuration.

This really feels like one of those issues where the knowledgeable engineers could hash it out in 10 minutes, it’s frustrating being the middle person between L1 tech support on both sides who have to relay messages back and forth.

Could you possibly isolate the commands coming from the cRIO in the trace file that are problematic? I also sent a number of messages a couple of weeks ago to support@hms.se after someone there emailed me directly with the subject line CRM:0357002028, but they immediately went dark again.

Also, can you confirm that the Omron XML file is incorrect for this ethercat/devicenet gateway? It won’t even allow transition between idle and preop and I suspect it’s because the “Outputs” mailbox is marked with a controlbyte of x24 instead of x64 (the base/”naked” gateway ESI file from your website says it’s x64 and attempts using that base file allow the idle to preop transition, but not, obviously, the preop to safeop)

Best Regards,

Mike

#45

Hi Mike,

I can’t tell which part of the response is quoting NI and which part you’ve added.

I would like to verify something quickly because I’ve gone back and looked at other cases involving our EtherCAT slave and most people have been successful in creating a working ESI file, but you need to set the correct sizes in the gateway using the Anybus Configuration Manager first. Also, when running the ESI Generator, make sure you are directly connected to the gateway via Ethernet, and there are no other devices connected to the PC by Ethernet.

I looked at some working ESI files, like this one and they have the x64 output control byte that you noticed was different in the OMRON ESI file. As far as I know that file still works with the OMRON PLCs, but I’ll check into the discrepancy.

I’m going to see if I can get one of our EtherCAT experts to work with you directly, but can you verify that the ESI from the Generator was created after the sizes set and no other Ethernet devices connected? Also, make sure that the VednorID in the ESI file is identical to the one set in the application in the EtherCAT Host Object.

Thank you,

Kyle

#46

Are you available for a call this morning?