TwoBlocked
Senior Member
- Location
- Bradford County, PA
- Occupation
- Industrial Electrician
We've been experiencing problems with the ABB XIO controllers. My thinking is that there is undesirable DC current in the negative busses due to the chassis and all the negative terminals on the I/O cards being bonded to together through the DIN rail. The power terminals on the controller do not have a separate ground, just a positive and a negative, with the negative internally bonded to the chassis.
I see this as the same as the AC definition of undesirable current when the Equipment Grounding Conductor and the Neutral are bonded together at the field devices. Negative DC current anywhere in the system can flow through the grounded negative busses in the controller and the I/O cards.
The problem that is being experienced is the analog input cards lose their "calibration" sometimes for no apparent reason, and sometimes related to a DC breaker opening. When "calibrating" an AI point on these, using the programming software, 4mA and then 20mA is sourced to the card and the chosen Engineering Units are assigned to these values. I would call it scaling, not calibration... So what happens is that instead of the program using the calibrated/scaled units, unexpectedly, other values are used. Of course everything goes whacko, we get called in, and "re-calibrate."
The local ABB Guru was consulted and he said the problem is with the inductive kick back from solenoids, both those operating valves and those operating relays. His solution is to install 6 amp diodes (Six Amps !!!) across all the wet and all the dry contacts of circuits controlling the solenoids. This includes the open collector outputs on the DI/DO cards going to din mounted wafer relays. (Really ???) I was told this was a dirty little secret that ABB knows about but won't publish. (Hmmm, Colonel Mustard, in the drawing room, with the candlestick?)
OK, I haven't personally talked to the Guru about this (a great, very knowledgeable guy) but plan to tomorrow. The solution is what was explained to me by someone else. (Yeah, one of the Fine Fellows involved in the single phase motor VFD fiasco...) I am very familiar with protecting contacts on control devices (especially wet ones) that feed inductive loads with various types of "snubbers." Doing so won't "hurt" anything, although it introduces another potential point of failure. But it will still send the inductive kick back current from the solenoid to the chassis ground where it can still flow through the negative buss of the controller and I/O cards, which I believe is the real problem.
My thought is to electrically isolate the controller and I/O cards from the backplane, thereby producing a system with an isolated ground; the negative to the controller would be the only ground connection. I often see this in quality installations. Will also need to ensure that the negative return from field devices go to the specific I/O terminal point, and not to a convenient negative terminal somewhere else.
So I am asking you fine folks for two things:
1. Are my thought processes on the right track?
2. Can you point out me to some good documentation about installing controllers with isolated grounding? I've looked for stuff, but it all seems to be either about grounding just one end of instrumentation shielding or about receptacles with isolated grounding. I want something independent to show those that make these decisions.
It's already on this week's schedule for the techs to go out and start installing 6A diodes all over the place. If this is a bad idea, like I am convinced it is, I want to "head it off at the pass." (Yeah, I like westerns.
)
I see this as the same as the AC definition of undesirable current when the Equipment Grounding Conductor and the Neutral are bonded together at the field devices. Negative DC current anywhere in the system can flow through the grounded negative busses in the controller and the I/O cards.
The problem that is being experienced is the analog input cards lose their "calibration" sometimes for no apparent reason, and sometimes related to a DC breaker opening. When "calibrating" an AI point on these, using the programming software, 4mA and then 20mA is sourced to the card and the chosen Engineering Units are assigned to these values. I would call it scaling, not calibration... So what happens is that instead of the program using the calibrated/scaled units, unexpectedly, other values are used. Of course everything goes whacko, we get called in, and "re-calibrate."
The local ABB Guru was consulted and he said the problem is with the inductive kick back from solenoids, both those operating valves and those operating relays. His solution is to install 6 amp diodes (Six Amps !!!) across all the wet and all the dry contacts of circuits controlling the solenoids. This includes the open collector outputs on the DI/DO cards going to din mounted wafer relays. (Really ???) I was told this was a dirty little secret that ABB knows about but won't publish. (Hmmm, Colonel Mustard, in the drawing room, with the candlestick?)
OK, I haven't personally talked to the Guru about this (a great, very knowledgeable guy) but plan to tomorrow. The solution is what was explained to me by someone else. (Yeah, one of the Fine Fellows involved in the single phase motor VFD fiasco...) I am very familiar with protecting contacts on control devices (especially wet ones) that feed inductive loads with various types of "snubbers." Doing so won't "hurt" anything, although it introduces another potential point of failure. But it will still send the inductive kick back current from the solenoid to the chassis ground where it can still flow through the negative buss of the controller and I/O cards, which I believe is the real problem.
My thought is to electrically isolate the controller and I/O cards from the backplane, thereby producing a system with an isolated ground; the negative to the controller would be the only ground connection. I often see this in quality installations. Will also need to ensure that the negative return from field devices go to the specific I/O terminal point, and not to a convenient negative terminal somewhere else.
So I am asking you fine folks for two things:
1. Are my thought processes on the right track?
2. Can you point out me to some good documentation about installing controllers with isolated grounding? I've looked for stuff, but it all seems to be either about grounding just one end of instrumentation shielding or about receptacles with isolated grounding. I want something independent to show those that make these decisions.
It's already on this week's schedule for the techs to go out and start installing 6A diodes all over the place. If this is a bad idea, like I am convinced it is, I want to "head it off at the pass." (Yeah, I like westerns.
)
