AKET

NX IO Unreliable?

7 posts in this topic

Has anyone found Omron NX IO to be unreliable?

I have 2 racks of various NX analogue and digital IO where connections between the IO modules appear to fail when modules are prodded with a finger. System will be running fine without error, then prodding/pushing IO modules will cause TS lamp to flash red on module and cause CPU to throw various errors such as image below. After cycling power, system will start up and run fine with no errors present.

Doesn't appear to be related to any particular module and where a particular module looks suspect, fault symptoms will not follow suspected faulty module when swapping with other modules of same type in the rack. Contacts on sides of modules are clean with no dirt or corrosion visible. System has been installed for 24 months and these issues started appearing at around 12 months and seems like it is getting worse.

Local Omron tech support are saying solution is to "re-seat" the modules by sliding modules in and out of the rack 10 times. Strangely enough, this actually improves situation for a while and then modules will start playing up again.

I am wondering if this is typical of NX IO modules? The connector arrangement on sides of the modules on close inspection looks like it may be prone to spring/wire side of connection jumping out of shallow grove in mating module.

ts_220927_121636.thumb.PNG.16560e58a55fd

 

20201023_081411.jpg

Edited by AKET
wrong photo

Share this post


Link to post
Share on other sites

Any time a manufacturer tells you to seat / reseat the connectors to improve connection, there is a design problem.  It sounds like the connectors are prone to oxidation and/or the spring connections are weak.  Many years ago I worked with a motion control company that had a servo controller and drive in one unit that would typically work fine for 12-18 months and then would start failing.  It was found that connector pins were oxidizing.  There was such a problem with them that they recalled the units and replaced the internal connectors with ones that would not oxidize over time.  It was an expensive fix for them, but I imagine they never repeated that mistake.  I would recommend being the "squeaky wheel" with Omron and encourage others with similar problems to do the same.  Maybe if they have enough customers complain, then they will fix the problem.  

Share this post


Link to post
Share on other sites

You say you have 2 racks of I/O cards, I am assuming that one rack is the CPU with some I/O cards and the other rack is the EtherCAT coupler with the remaining cards (as opposed to 2 EtherCAT couplers, with the CPU located elsewhere in the installation).

If this is the case, does the pushing/prodding on the CPU cards create the same kind of issues, or is it just the cards in the coupler that have the issue when prodded?

Is there anything else in the EtherCAT network or is is just the CPU and EtherCAT coupler?

Are the PLC components installed in a cabinet free from dust etc? The photo you showed does look quite clean, but best to ask. Dust never helps :-)

I would suggest that if the CPU cards don't have the problem (and the Coupler cards work find when swapped), then you are looking at an EtherCAT issue.

If the CPU cards have the same problem, then I really don't know, but I would be tempted to swap around the PD1000 power cards just in case one of them is faulty. I have no idea if a faulty PD1000 could cause such issues but my understanding is that they do provide the bus that powers the I/O cards. I am just throwing ideas out there at this point.

I have never had the exact problem you currently have, but I have had similar errors when I had 2 racks of I/O, 1x EtherCAT coupler and 1x CPU. Faults just randomly appeared a few months after commissioning, and could not be reset without a power cycle, after which it worked fine anywhere from 2 minutes to a couple of days. In the end it turned out to be a faulty EtherCAT junction slave (hub) in my EtherCAT network. This installation has worked fine ever since.

Edited by BE

Share this post


Link to post
Share on other sites

Thanks to drforsythe and BE for their replies.

I have 2 racks of IO. One is direct to CPU and other is on EtherCAT coupler. CPU rack photo is attached here. CPU rack does have issues with AD3204 modules going into error. Most recently, AD3204 in slot 1 was failing when prodded, so swapped it to last position of the 8 where it behaved flawlessly as did module now in slot number 1. The EtherCAT rack also has issues with TS3101 modules going into fault when prodded and behaviour is similar to that described above where a module that seems unreliable can be relocated to a different position in the rack and then behave fine as will module now in position of previously failing module. 

20201023_081350.thumb.jpg.3ff15454b28065

Share this post


Link to post
Share on other sites

So taking the CPU rack for example, you said when you prodded the AD3204 in slot 1 that it failed. When you say failed, did the I/O for that one card drop out or did everything drop out? It really does sound like there are bad contacts somewhere that are disconnecting with slight movement, but to be honest I have never had that problem. And if the CPU cards are dropping out, then I can't see it being an EtherCAT issue.

On a slightly different note, it doesn't appear that you have a PF0730 card in slot 1 on the CPU rack. I only raise this as every time I order an NX102 CPU, my Omron tech tells me I need to order one for the I/O supply (for the CPU only, the Coupler doesn't need one), but that Sysmac Studio doesn't show it as being needed. So I take their advice and order one. Now I am starting to wonder if I actually need them.......:-0. I wouldn't think that would be the problem, as it has worked fine for ages, but it was just something I noticed.

Sorry I can't be of any more help, maybe someone else has some ideas too or chime in if they have seen similar behaviour before.

Share this post


Link to post
Share on other sites

When inducing failure by prodding cards, the end result is TS lamp on card being prodded and all cards to right flashing red. CPU then cannot “see” these cards until power cycled.

I use PF0630 to feed IO power to cards. From memory, I think the PF0730 is just a higher current version. The AD3204 cards immediately to the CPU right of don’t need IO power so didn’t place the PF0630 in the rack until just before the ID5442 cards that require the IO power supply. The options available with the NX bus for feeding and taking off IO power is one of the things I like about the NX system so would like to keep using it if I can work out why it is behaving so unreliably.

Share this post


Link to post
Share on other sites

Did you try to replace your ECAT cable to shielded one (more industrial build, not for offices like you have)? Let's try to debug from beginning.

Edited by scotty134

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now