PointIO Change From DeviceNet To Ethernet

I took his 'Modules" as the DeviceNet socket part, not the Input and Output modules plugged in them.

alan_505 said:

When plvlce says back pane he means the down stream modules of the adapter as you noted these are tongue and groove connections.

Click to expand...

Exactly, those connections form the 'backplane' such as it is (and the literature does indeed call it a backplane, which is why I do as well).

In my opinion it is this lack of a solid actual backplane that causes many of the issues with vibration. The only things holding the connectors together is some plastic that allows several mm of motion and the din rail.

EDIT:

I_Automation said:

I took his 'Modules" as the DeviceNet socket part, not the Input and Output modules plugged in them.

Click to expand...

I attempted to consistently use 'module' for what you are now calling 'card' to be consistent with the nomenclature you were using at the time (eg "Node 20 has the Network LED's flashing constantly on 3 modules)

I apologize for any confusion due to mixing terminology.

In conclusion, in case it isn't clear, I have seen vibration cause problems with the backplane (comprising terminal base units) and with the attached modules/cards -- quite possibly the latter problems resulting from the former. Based on my past experience with such issues as well as the status light behaviour here, I think this is the issue.

I would start by cleaning/reseating all modules *and* the backplane units (done or currently in progress from the sounds of things) and if problems persist go armed with spares for both. For troubleshooting purposes it may be helpful to note that Point I/O supports RIUP (Removal Insertion Under Power, ie no need to power down if just pulling the module/card). This means you can quickly swap modules to test. For example, on Node 20 it appears that cards 3 and 4 are good and have good backplane connection (solid green). Card 5 does not. If you remove both, then insert Card 5 into Card 3's slot, then if the light continues flashing after it boots then Card 5 is suspect. If the light goes solid instead, Card 5 is good, but the terminal base behind it is likely faulty. *label all cards clearly before removing any*

Well I was just invited to go for a 2 hour drive so talk to you later, and hopefully with good news

In the words of Robin: Holy C***, Batman.


Turns out 3/4 of the IO cards are the wrong revision and that has to be exact.
Node 20 HAD a secondary power supply only for card 10 - and online could not see it, plus nothing wired to it, so I removed it but have to reprogram Node 20 for that.

Also the ADN's were the wrong revision, but although Node 10 fixed the problems and took a new download and shows all green now, Node 20 refuses saying the processor is in Run Mode. The help on this just says take it out of run. No mode switch or option anywhere in RSNetworx to change the mode.
The DNB was also the wrong revision, and the one they just bought scrolls a buss power problem so it's bad.
Trying to download to the DNB gives another RunMode - I am now getting online with RSLogix5000 to shut that off, hoping that puts the nodes in Program too, or the new DNB program allows the mode change.


EDIT: Changing the DNB Run BOOL from 1 to 0 isn't working -- it changes right back to 1


EDIT2: Finally found the one rung turning that bit on, let me download to the DNB and Node 20 ADN and now all green and not going to be home before the bars close.

I had not remembered Point I/O being that finicky about revision, but one of the real disadvantages of DeviceNet imo is needing RSNetworx for even minor changes like that.

Surprised to hear that a separate power supply for outputs would cause an issue -- the FPD should be completely transparent to the adapter.

Glad to hear you were getting things sorted out though.

plvlce said:

I had not remembered Point I/O being that finicky about revision, but one of the real disadvantages of DeviceNet imo is needing RSNetworx for even minor changes like that.

Surprised to hear that a separate power supply for outputs would cause an issue -- the FPD should be completely transparent to the adapter.

Glad to hear you were getting things sorted out though.

Click to expand...

I think the separate power supply was dead, had to end up putti,g card 10 back in without it and the node saw the card then.

However, even though everything is showing green the customer just woke me with the news they can't clear faults related to inputs going to node 20

How did you go after you wake up phone call?

alan_505 said:

How did you go after you wake up phone call?

Click to expand...

To save the driving it took hours to get the customer to put Teamviewer on the laptop.


When I finally got online 2 output cards with different revisions were swapped and the dead secondary power supply only powering the last (totally unused) output card was put back in. I had them remove the dead power supply and I corrected the configuration of Node 20.



Could not remove module 10 on node 20 and download the new configuration - it still reports the processor was in Run mode and would not download. So I had to leave the unused output module installed. Even going in the PLC and shutting off the DeviceNet adapter so I could download to it did not take Node 20 out of Run. Downloading to Node 10 was allowed even though it was running. Nothing on Rockwell.com about this other than "take it out of run." No mode switch on the ADN or anything to click in RSNetworx I found.


It finally came up running but after a few seconds Node 20 showed red even though it was working. Digging down found output 10 giving 8 errors of nothing connected to the output.


At least this got the faults clear and the press running, but this system is not being used this job. I told the customer that if the "Nothing Attached" is a problem when they put the next job in to add 8 pilot lights to the 8 outputs, just they would never light.



I couldn't see how you would tell it that an individual output wasn't in use, but can't imaging whoever designed this figured that every output on every module was going to be used with none extra & empty.


The only thing is I had to find an older version of Studio 5000 that included RSNetworx, as the newest v32 and all v33 don't include it anymore. Then, even though I have Studio Full Edition, that is supposed to include everything, it doesn't license RSNetworx and had to run on a grace period license. I have an old EVRSI license for it but not the original floppies from back then so Rockwell support's official answer is Go ! Yerself and pay us for a new license. Plus a scheduled call-back from them for Monday hasn't been made as of Wednesday afternoon now.


Instead of converting this PointIO to Ethernet I am going to check into Automation Direct Click's that can be directly added to the Ethernet IO tree of the CLX PLC. When I powered it down I pulled the card and it is a 1756-ENBT/A Rev 4.003.


Plus for added capabilities I am going to change the L55 CPU to the L63 I have and upgrade the v16 project to v20.

I_Automation said:

Could not remove module 10 on node 20 and download the new configuration - it still reports the processor was in Run mode and would not download. So I had to leave the unused output module installed. Even going in the PLC and shutting off the DeviceNet adapter so I could download to it did not take Node 20 out of Run. Downloading to Node 10 was allowed even though it was running. Nothing on Rockwell.com about this other than "take it out of run." No mode switch on the ADN or anything to click in RSNetworx I found.

Click to expand...

No idea why you'd have a difference in behaviour between the two nodes.

It finally came up running but after a few seconds Node 20 showed red even though it was working. Digging down found output 10 giving 8 errors of nothing connected to the output.

At least this got the faults clear and the press running, but this system is not being used this job. I told the customer that if the "Nothing Attached" is a problem when they put the next job in to add 8 pilot lights to the 8 outputs, just they would never light.

I couldn't see how you would tell it that an individual output wasn't in use, but can't imaging whoever designed this figured that every output on every module was going to be used with none extra & empty.

Click to expand...

In the subnet, open the Properties for the individual output card and under the Configuration tab look for a series of parameters labeled "Enable No Load" (one parameter for each output). Set them to disabled.

The only thing is I had to find an older version of Studio 5000 that included RSNetworx, as the newest v32 and all v33 don't include it anymore. Then, even though I have Studio Full Edition, that is supposed to include everything, it doesn't license RSNetworx and had to run on a grace period license.

Click to expand...

Indeed, 'Full' does not include RSNetworx...

I also told the customer to tell EVERYONE that the cards, adapters and interface on the entire DeviceNET system CAN NOT be swapped.



The IO cards would have to be the exact revision and the adapters and interface reprogrammed every time.

Obviously some one did not listen!

alan_505 said:

Obviously some one did not listen!

Click to expand...

Hoping it took, but I told them that during the call - not when I was there