illegal word operand access Vision 570 Stop Mode Fatal Error

[Joe Tauser]

Thanks for the pictures.  That installation reminds me of how I used to do things 20 years ago.  I rarely use Snap IO anymore and my enclosures have gotten bigger.  System designers (myself included) are notorious for forgetting that you have to leave room for the wire to connect things together.  And God forbid you leave a couple of empty inches on the DIN rail in case you need more terminals.

I have coined a saying about machine and pump skid builders- "Electricity is the nasty thing required to make our wonderful machine work".  So they go to great lengths to put together an awesome machine and then hang the electrical panel in too small of a space underneath or on the side.  "That's all the room we have left". 

I do understand about the cards you're dealt.  I keep an old reliable Fluke 123 in my car for the problem you're running into.

Joe T.

[Ausman]

2 hours ago, PaulK said:

Also, apparently the Stop Error happens several minutes after the VFD/Motor is running, which would lean towards vibration?

It would perhaps be a good exercise to log the speed of the VFD somehow.  With the plc getting the fatal error I don't know whether doing it into MIs will be held for perusal on cranking it up again, though. But my thinking is that perhaps a particular speed is causing Flex's vibration, or the other thoughts about noise.  I routinely wiggle-then-re-gaffer-tape snapins I use, just to keep contacts working ok.  Screws sound good.

Lots of possibilities here, look forward to the solutions!  You're going to have soooo muuuuch funnnnn!

cheers,

Aus

[PaulK]

16 hours ago, Ausman said:

It would perhaps be a good exercise to log the speed of the VFD somehow.  With the plc getting the fatal error I don't know whether doing it into MIs will be held for perusal on cranking it up again, though. But my thinking is that perhaps a particular speed is causing Flex's vibration, or the other thoughts about noise.  I routinely wiggle-then-re-gaffer-tape snapins I use, just to keep contacts working ok.  Screws sound good.

Lots of possibilities here, look forward to the solutions!  You're going to have soooo muuuuch funnnnn!

cheers,

Aus

Thanks Aus, I really do love a good challenge.

[kratmel]

Please read some rules from this two documents:

http://literature.rockwellautomation.com/idc/groups/literature/documents/in/1770-in041_-en-p.pdf
http://literature.rockwellautomation.com/idc/groups/literature/documents/in/drives-in001_-en-p.pdf

I think that problem came with PLC power supply grounding.

Please try to temporary separate OPLC power supply  . Power them from battery backed UPS (Disconnect grounding of PS)
Maybe you will see where is problem - electromagnetic  noise or vibration.

Also check the VFD  specification - is it embedded noise filter present inside the VFD.

Generaly this type of problem is in incorrect grounding of vfd + motor + equipment chasis.

 

P.S. Please check position of noise filter enable  switch inside VFD box.  It present in all VFD.

[PaulK]

Hello,

Going to start troubleshooting tomorrow morning.

One more piece of information. Before the stop error, V570 starts beeping before it faults out.

Going to check grounding and cable routing and probe for noise.

[PaulK]

I want to thank everyone who responded... Thank you!

I fixed the issue within 90 minutes.

1. The digital IO to and from the drive (Run Command and Drive Faulted) were tie-wrapped to the 575 VFD Line Supply.
2. They had installed Line Reactor  beside the 24 VDC power supply (yes, un-shielded motor leads up and out of the control panel)
3. Shields on twisted pair cabled had been snipped off, so there was no drain connection to ground.

• I ran new digital IO wired and connected the drains at one end to ground
• The new digital IO wired crossed the 575 at 90 degrees, and I made sure the IO was spaced away from the 575 and the motor leads
• I removed the Line Reactor

Problem solved...I hope.

The system has been running for 48 hours, previous to this it would only stay up for 15 minutes...looks promising.

This was not following wiring fundamentals 101.

Best regards,

 

Paul

[Ausman]

That's great Paul,

Hope it stays up.  I also hope you hammered those fundamentals into the client's brainstrust with a mighty big stick. 

So.................now for the sharing of the fee.............!!!!

Cheers,

Aus

[PaulK]

Well, it is back after 14 days of running.

The PLC/HMI starts to beep slowly and then Black Screens.

Unitronics tech support is suggesting OS is corrupted?

Is it possible unit has been degraded?

I am considering replacing the V570 as well as the Snap IO board.

 

[Flex727]

I'm still of the opinion that this could be a vibration issue. Did you tape or screw down that Snap-In module? Vibration can also wear the contacts, so if you have already taken steps to lock down that module, the 570 and/or the I/O module may need to be replaced and then locked together.

[PaulK]

Thanks Flex727 - no, I didn't add any tape or screw it down.

I rerouted some IO away from 575 and motor leads, after doing that the results were dramatic, I went from a fault every 15 minutes to one fault after 14 days.

I am ordering a new Snap IO Module and V570, was talking to Micheal from Unitronics just now, and the thought was that the unit could be degraded from all the previous noise exposure.

It is probably a long shot, but something we are going to do.

I am also going to be probing for noise as well at checking all grounding.

One last item to try was to float the 24VDC power supply from earth ground.

With the new unit, I will tape where I can to keep the Snap IO fixed to the V570. The current snap IO module is very secure and snug with the V570, but I will try what you recommend.

Problem is, if it is only happening every two weeks, and I change more than one thing at a time, then I will not know with certainty what corrected this issue.

[Ausman]

Going on the previous photos, I would also consider putting a suitable filter onto the mains supplying all the control gear power sources.

And I still don't like that contactor being so close to the power supply and associated wiring.  Separate that all out as much as possible, even put the contactor elsewhere.

If they are happy to be buying new gear, then get metal enclosures as well!  Nowhere near the cost of the plc, but admittedly a bit of mucking around rerouting all the wiring and possibly the component layout.  But gee...plastic enclosures and a vfd?  Arggh!  Get 2 x s/s with one for power gear and one for all controls.  I routinely have NO mains in the PLC enclosure at all, only control voltages, with the 24vdc supply coming in to the fuse distribution totally separated from control lines.  Probably overkill, but makes it far safer as well as reducing noise issues...which I've never had.  Hmmmmm!

Finally, it looks like some of the conduit is in Flextite or similar.  I'd be earthing one end of this as well, if it is indeed metal spiral.  Everything that can induce noise but can be earthed down as a shield should be done.

As for the degraded theory....perhaps....but I would have thought it would play up from the word go.  Have you reinstalled everything into the plc?

cheers,

Aus

[PaulK]

Hello - So the noise is coming from a 24VDC valve solenoid coil. The coil is being energized directly off of the Snap IO output (I didn't design this).

Question - I am not an electronics type, more software.

So options...

Opto Relay?

Solid State Relay?

Diode Snubber Circuit?

I would like to test with a Diode, has anyone used this method, and if you did, which type of diode was selected?

Thank you!

 

Paul

 

p.s. - I am sick to my stomach over this. Work, make mistakes and learn,

 

[PaulK]

We are going to try Diodes on each solenoid valve, across the Coil, reverse bias.

From reading on the interwebs, it would appear the 1N4004 is the go to diode.

 

I will report back.

[Ausman]

Hi Paul,

Yep, a solenoid valve coil turning off will definitely do a big spike.  Have a read of this topic for a lot of tips, particularly Joe's post at the end re diode type. 

Important.......I'm assuming that when you say the noise is coming from coil, you are actually saying it is when it turns off.  If it is happening all the time there is an issue with the coil itself.  There should be no noise if it is in either state......only on transition from on to off.  If there is noise during on, then it may also be the valve itself not opening correctly, but is most likely a faulty connection there somewhere.  Check connections etc and try a new coil if possible.  Also make sure the code is not written incorrectly and rapidly cycling the output, giving the impression it is always on but it really is going on an off.  Put the scope on the line direct during on to check this.  Very unlikely but possible.

As for how to drive it, for a solenoid valve I would never be driving it direct from the snap-in.  The only things I drive direct are very low power things without any inductive qualities at all, like LEDs.  Every other instance goes through a standard relay or SSR, depending on the load being driven and frequency of switching.

Lastly, my Confucius bit of the post....we all learn something new every day.  Sometimes it annoys immensely, especially when it is fixing work by others who should know better.  In this case it sounds like the entire machine was built without any thought about noise at all.

cheers,

Aus

 

[PaulK]

Hi Aus, thanks for the support.

Yes, when the coil is de-energized,  a short time after (instant to 10 minutes) the PLC goes "Roy Munson" (that is a Kingpin movie reference).

Correct, this panel was designed and built without regard to best practices, standards, noise, guidelines, solar flares or silk underwear. 

I have been doing software exclusively for a while (how the cards were dealt), and I know a tiny bit about electrical from just working in this field for 25 years.  I can not count all of the things I would have done differently in this panel...some of it is just common sense.

Thanks again, I will update when I have more feedback.