WiFi bridge in Device Level Ring (DLR)

tduryee

Member
Join Date
Mar 2020
Location
Washington
Posts
7
Hi folks.
A client replaced a slip-ring some time ago with a WiFi point-to-point bridge, connected to Flex I/O. While it works most of the time, it drops sometimes and they've asked me to spec redundant WiFi.

I've looked at a number of (more expensive and more complicated) scenarios before realizing it could probably be done with 2 point-to-point bridges connected between 1783-ETAP devices. (See attached img)


I suspect the WiFi devices may behave like non-DLR devices between DLR-enabled devices, but I haven't found any publications on this particular application.

Do any of you have experience with WiFi in a DLR?
Are there reasons why this architecture would not work, or be unreliable?
Thanks!
 
I work with DLRs on daily basis but never heard about WiFi DLR enabled device. All devices within DLR must be DLR aware and not only pass data, but must handle all housekeeping info. Also default beacon is 400usec with 1960usec timeout. I don’t see how standard WiFi can handle that.
 
Last edited:
I work with DLRs on daily basis but never heard about WiFi DLR enabled device. All devices within DLR must be DLR aware and not only pass data, but must handle all housekeeping info. Also default beacon is 400usec with 1960usec timeout. I don’t see how standard WiFi can handle that.

Well that was quick.
I appreciate the feedback.
Any thoughts on alternatives to create wifi redundancy?
Could the ETAP or some other device be configured with STP (or something else) to avoid an unmanaged ring?
 
Last edited:
First choice solution would be to fix the existing wifi. That can be tricky though, physics can be cranky.

There are other redundancy protocols that could make more sense. PRP (Parallel Redundancy Protocol) is specifically intended to allow for parallel networks and combine the data at the end so that only one packet gets through.

I know Siemens wifi systems implement specific support for it (as well as EIP).
 
Physics is definitely an issue. Between magnetism, high voltage, rotational and distance movement, I think they need something redundant or wired. They've nixed using another slip-ring, so I'm trying to help with the redundant wifi (obviously not my specialty).

On first look, PRP looks like what I'm aiming for. The Rockwell Stratix 5400 has it and would allow me to grab data directly from the switches. Now time to read through all these manuals. Thanks for the help.
 
Physics is definitely an issue. Between magnetism, high voltage, rotational and distance movement, I think they need something redundant or wired. They've nixed using another slip-ring, so I'm trying to help with the redundant wifi (obviously not my specialty).

On first look, PRP looks like what I'm aiming for. The Rockwell Stratix 5400 has it and would allow me to grab data directly from the switches. Now time to read through all these manuals. Thanks for the help.


Something to watch out for, although it probably won't affect IO traffic:


PRP adds a few bytes of header to the normal Ethernet frame. If the packet is already max size (1500ish bytes), and those extra bytes are added, it becomes a "jumbo frame". Many networking systems (wifi included) drop jumbo frames, although some may be configured to allow them.
 
Last edited:
Hi & Welcome to the Forum!

tduryee said:
...A client replaced a slip-ring some time ago with a WiFi point-to-point bridge, connected to Flex I/O. While it works most of the time, it drops sometimes and they've asked me to spec redundant WiFi...

May I ask what WiFi hardware is currently in use?

Regards,
George
 
You could also use maximum values for beacon timeout and beacon interval on the DLR supervisor(s) instead of default values, if you have not tried this yet.

20200324_DLR_Parameters.png
 
tduryee said:
Thanks!
Currently they're using Engenius ENS500EXT.

Thank you. As you may or may not be aware, Rockwell Automation have many "Encompass Partner" companies. They will use these companies to develop and provide us specialized or bespoke solutions. One such company you may be aware of is ProSoft Technology who specialize in connectivity between disparate communication systems.

Among ProSoft's many product lines is their tried and tested wireless solutions. Among these products is their Fast Industrial Hotspot RLX2 range. Rockwell, partnered with ProSoft, have provided many companies with a wireless RLX2 replacement solution instead of existing, expensive and troublesome slip ring assemblies. They have been offering such solutions for many years now with great success. While I have not used the latest models, I had used a previous model a few years back while working in and around the Water Treatment industry and also Grain Handling (Maltings). We replaced the slip ring assembly on a number of large outdoor Clarifiers and Grain Silo augers with a Rockwell/ProSoft wireless solution and I must say it was rock solid.

Anyway, I would advise you at least take a look or perhaps have a chat/visit with a rep and see what they may be able to do for you. It may prove better here to put this specialized solution in the hands of the experts rather than us trying to design something up here based on your own "seemingly unsure" specification? If you can find a reliable wireless solution then redundancy may not need to be considered here at all? Also, looking at "Engenius", whom I'm not that familiar with, and I could be mistaken, but they do not scream "industrial wireless supporting the likes of EtherNet/IP and Modbus" at me? ProSoft eat such staples for breakfast.

802.11abgn Fast Industrial Hotspot RLX2-IHNF

Note: There's a Logix 5000 Add-On Instruction (AOI) in the Downloads section.

Another technology to consider is Radiating Cable (Antenna), which ProSoft also provide, in case you are fighting RF interference. You "wrap" it around the machine as a flexible antenna to boost the range and cancel out RF signals...

Is Radiating Cable Right for Your Material Handling Application?

Find out how a radiating cable can improve wireless communications in your facility!

Radiating Cable White Paper - It’s not a Cable, It’s an Antenna!

Some testimonials I've found...

Wrapping five bottles per second

Conveying reliable communications in Italy

Wireless Radios Help Reduce Costs and Downtime for Manufacturing Company

I don't work for any of these companies. I've just used some of these products and can testify that they "do what they say on the tin".

Regards,
George
 
Thanks George. In fact I speak with the Prosoft people frequently and have had good successes with their equipment for many years (mostly Modbus devices). The radios you mentioned (outdoor version) are exactly what I have in mind for this project. They actually mentioned the radiating cable and I plan to discuss that with the client as well.
The old WiFi units are definitely not what I would call industrial, and my first suggestion to the client was that we just do a simple replacement, tune the network a bit, and see if it solves the problem. C'est la vie.

A little update on my thinking:

In reference to PRP:
According to Wiki, "The maximum MSDU size is 2304 before encryption. WEP will add 8 bytes, WPA-TKIP 20 bytes, and WPA2-CCMP 16 bytes."
That suggests, with an MTU of 1506+header, frames theoretically shouldn't be dropped. However, According to the following site, if the MTU is above 1500, 802.11 fragmentation will occur.
https://inet.omnetpp.org/docs/showcases/wireless/fragmentation/doc/index.html
I'm not sure that's a problem, but honestly PRP may be an over-build for this application. Most of the FlexI/O is operator input, and nothing critical that requires zero latency failover.

The build I'm considering is: 2 Stratix 2500 switches with redundant Prosoft WiFi units (see attached img).
This at least provides RSTP.
If the time-out for reconvergence takes 3 x 2 seconds, it will annoy the operator enough that they will let the techs know, but it won't interrupt production.
The hope is to get the new installation dialed in enough that it won't fail-over.
Also, this parts-build allows access to device diagnostics through Logix 5000 Designer which should make it easy to provide indication of failures.

Alfredo: The current RPI is 100ms. I may bump it up to 200ms, but I'll do some testing and calcs to see what the traffic looks like. Thanks for that information.

I'm still reading/researching and would love feedback.
 
Last edited:
I would question "PRP" note on your drawing as Stratix 2500 does not support it.
PRP requires physical separation between two LANs in configuration called PRP RED BOX. This is only supported on Stratix 5400.
 

Similar Topics

Hello, I want to establish LTE communications (2 S7-1200) with common Workstation. All i have is 1200CPU with 6GK7243-7KX30-0XE0...
Replies
0
Views
270
Have a project coming up which will use a tablet HMI in addition to the two fixed panels. Curious if anyone has a tablet/router/AccessPoint combo...
Replies
13
Views
1,372
I have a customer that wants to make a system. Think garage door opener. That functional part is simple. They want to connect it to the customers...
Replies
8
Views
1,606
My PLC project I have been working on for a few years now has a new home. It is a Micrologix 1400 and I have a C-more Micro 6 inch panel...
Replies
8
Views
1,885
Good afternoon, I'm building a test station for temperature mapping that will log 12 data points to SQL via Open Automation Software. This station...
Replies
4
Views
1,900
Back
Top Bottom