Micrologix 1100 to Keyence DL-EP1 Messaging

[IndCon]

I'm having a hard time messaging with a Keyence DL-EP1 using a MicroLogix 1100 PLC.

PLC is a 1763-L16BWA B/16.00.

I used the Molex Ethernet/IP Tools and confirmed that I can message the unit, but for some reason, the MicroLogix is MSG instruction is erroring out after the timeout with "Connection timed out by the network". I also monitored the channel status and all Ethernet/IP Replies are "Recieved with Error".

I've checked all my settings, got CIP Generic, Generic Get Attribute Single, Class 67, Instance 0, Attribute 90 (as per the manual) and also Mulihop is setup for the correct address of the Keyence.

I'm just stumpped, anyone come across something like this and/or have some ideas/suggestions?

Thanks in advance!

 

[IndCon]

I'm thinking maybe it's the "Size in Bytes (Receive)", but the manual shows it should be 32 and I've tried a bunch of other values just for kicks.. Nothing worked..

 

[AlfredoQuintero]

Can you get a Wireshark trace? You will need a manages switch with port-mirroring function.

 

[IndCon]

I have WireShark and was trying to use it to check, but didn’t have much luck. I also don’t have access to an available managed switch.

What are the main causes in the Chanel status that would cause Ethernet/IP replies to be rejected? Seems like it is communicating, just not liking the information coming back. Usually the size right?

In the Molex app, the data going out and back was 8 groups of 2 hex numbers. I would assume that’s 16 bytes - (1 byte per hex number x 2) x 8. But I’m not sure if there are trailing zeros or not. Also the manual states the information is send back as a Int32.

Thanks for the reply btw.

 

[AlfredoQuintero]

Invorrect connection size, incorrect assembly instace and incorrect product Id amongst others can cause such a behaviour. The Wireshark trace would show this clearly in the forward open failure redponse

 

[IndCon]

Thanks for the reply. I’m sure the class, instance and attribute is correct as I can see the data coming through on the Molex software. So, I agree that the size is more then likely the issue.

Do you, or anyone, have instructions on how to find the information you mentioned through WireShark? I tried to use it, but the data was a bit overwhelming not knowing what to specifically look for.

Thanks again!

 

[the_msp]

Post the wireshark info here and someone can analyse, if you're on a larger network filter the IPs to those in question

 

[IndCon]

Thanks, I’ll do that Monday when I’m back in the office. But I’d still like to learn one day.

I’m guessing I should do a capture with the Molex software connected directly to minimize other traffic. Should I also provide when the PLC sends messages too?

Thanks again everyone.

 

[AlfredoQuintero]

The Molex tool only causes explicit messaging. You need to analyze the implicit messaging, especially the forward open request of the plc and the forward open response from the field device adapter. The EtherNet/IP dissector in Wireshark is extremely good and decodes the CIP application layer in thorough detail. If you post that trace many in this forum will be able to point you to the cause of your problem.

 

[IndCon]

Thanks for all the assistance so far everyone. Please find attached two Wireshark capture files.

"KeyenceDL_EP1_Molex" is direct between my computer and the DL-EP1 using the Molex Ethernet/IP Tool

"KeyenceDL_EP1_ML1100" is with ML1100 and DL-EP1 connected to a switch and sending MSG instructions (I think the switch is blocking out the packets, but I'm not really sure).

ML1000: IP: 192.168.1.10
DP-EP1: IP: 192.168.1.15
Laptop (VM): IP: 192.168.1.247

Let me know if you need any other information.

Thanks!

 

[IndCon]

Good morning, just wondering if anyone was able to have a peak and see if they can assist me. I appreciate your time and support.

 

[IndCon]

Just a quick update, seems the 32 was the correct byte size, the issue was RSLogix V10. I downgraded to V9 and everything is working as expected.

Thanks for everyone's assistance, hopefully this helps someone in the future.

I'm going to try V11 and 12 when I have time.

 

[Ken Roach]

The "_ML1100" Wireshark capture contains only broadcast and multicast traffic, so it must not have been connected to a mirrored switch port.

It's normal for a switch to send ordinary unicast Ethernet packets only between the involved devices; you need to use an old-style "hub", or a dedicated purpose "tap", or a managed Ethernet switch with the ability to "mirror" all the data to a specific port where you can plug in with Wireshark and capture the traffic.

Reaction to your edit:

Oh, fricking RSLogix 500 version 10. It had a bug that corrupted some kinds MSG instructions, and that bit you. The defect is documented in RA Knowledgebase document #BF23408 (Access Level: Everyone).

 

[IndCon]

Thanks Ken.. Yeah, even Rockwell didn't mention that in the 1000's of emails and online chats. When I mentioned it, their reaction was "Oh, should have asked what version your using...."

Guess 11 and up are OK in case anyone else ever comes across this issue.