Allen Bradley Current Loop

[Mickey]

For the number of cards you say have "blown up", you could have bought a bunch of them.

 

[Goodknight436]

Agreed I should have been smarter with my testing.

 

[GaryS]

There should never be a voltage difference between the 2 com points to work properly that's why the code calls for all power sources to have one leg bonded to the ground conductor, in the US the most common connection is the –DCV on the power supply. But be careful some power supplies are a switched type with no isolation transformer between the input supply and the output side of the power supply. With a floating power supply you can bet there is a voltage difference between the 2 com terminals that voltage can change over time and can be higher than the original supply voltage.
Isolators will eliminate problems with those voltage differences but they too must be applied correctly

 

[Goodknight436]

Update

Measured 100 mv from the common to ground on the input and 0mv from the common to ground on the output. That doesn't seem like a big enough difference to cause alarm. How big would you want to go on something like that before you put an isolator on?

Also tried the proper wiring setup and analog out worked out for the output. The output puts out what it needs. Cant say that much for the cell682 inputs though as when I fried the contacts it must have did it to the whole board because I cant get any readings of any type.

Leaving at that since the customer is installing their own cell682 and I cant get ahold of a replacement, we are shipping it tomorrow by decision of management. So if you guys want to know anything else let me know.

Thanks
Goodknight436

 

[GaryS]

100mv tells me that either they are not bonded together or you have some current flow between the 2 common. I would expect maybe 1 or 2 mv I would check the grounding bond an isolator would help but if as you say you burned out other outputs to I would definitely recheck all the grounding to be sure they are not floating high UL code requires one side of all power supplies to be bonded to the chassis it sounds like maybe they was missed someplace.

 

[Goodknight436]

The Cell682 come with its own AC to DC 2 prong adapter so its not grounded because its attached to a GFI wired inside to comply with UL code. The Cell682 however still has a terminal to allow grounding but its attached to the AC ground. While the Controller has its own internal power supply that's also grounded to the AC Grounded. So I would assume that means the Cell682 is Floating. Is there a way to get around that?

 

[GaryS]

if you have a floating ground on either side then you have big problems. There is just no way to get a accurate reading on anything that way. as I have said before both com's must be bonded together. if there is a question about connecting the com terms together than a signal isolator will be needed otherwise you never get a reliable signal.

if you have a question about the grounding contact your local inspection agency they can have somebody come out and advise you on all your options

 

[Goodknight436]

I am using a 4-20ma signal though. Can voltage really effect that? From my understanding current signals are more reliable for this reason; being able to reject foreign voltages created by other equipment. I understand basic electronic theory but when you get into anything beyond that I need explanation.

Ill try to understand and apply what you are saying though. Since I get 0mv from the Output to ground, the output is well grounded and not floating. The input reads 100mv proving that its not well grounded and is floating.

So your saying the 100mv would create a situation where the span of each instrument is thus different creating two different scales. Couldnt I just calibrate the input to the correct span?​

If that's not true or not applicable you are saying a floating ground is inherently unreliable reference point for the instrument and is prone to change causing the instrument to "float" out of calibration for whatever reason. This then leads me to ask what would a signal isolator does exactly? Does it create a good reference bounded ground that both instruments can use through their Com terminal? Thus that would ground the terminal for the input since its unable to.​

 

[GaryS]

Goodknght436
Absolutely a floating ground has caused more problems than any other single item.
I have seen electricians spend months trouble shooting a intermittent electrical problem only to discover a floating ground.
Even thought you are using a 4-20 ma signal they do by default reject short duration spikes. But they are not without problems.
A basic rule of electricity the total sum current or voltage in an electrical circuit must sum up to = 0
I know that may not make sense but you have to look it up or just rust me.
The return path for the ma signal is the grounds or circuit commons .
The ma source is of course the OF4 output card then the signal passes through the wire to the input of the sensaphone . The current can't just stop there it has to return to the source the best available path is through the commons. That's why the path must have as low a resistance as possible. If either of those are floating then the resistance to the return current is constantly changing. And what's worse the voltage potential between these 2 points is also changing. If it goes high enough it could damage the internal circuit of either device. Electrical current will find a path back to it's source if possible and it may not be the path you want.
But if the commons are insulated from each other then there is no path for the current to take = no ma signal, but that dos not mean there is no voltage potential in fact the voltage potential could in fact be very high into the hundreds of volts (think static charge)
That’s why we ground everything.

No you cannot calibrate around a floating ground problem what you setup today may not be the same tomorrow or even this afternoon.

However there are times when you just can't ground or bond all the different power supplies together for whatever reason that where a signal isolator comes in. In a signal isolator the power supply , input signal and the outputs are all isolated from each other either by an isolation transformer coupling or optically coupled. Each has their own [input – com], [output – com],power supply High side - Low side or common. This allows you to connect across the different circuits without current passing between them.
Go out and lookup the data sheets on a few isolators and see how they are connected.

There some people that will claim that floating ground are a good thing. But if that was the case then why does the NEC require everything to be bonded (grounded) to a single electrical point of the building service. Ground conductor pulled with everything even the panel doors need to be bonded to the panel structure (not using the hinge as a conductor) and the panels must be bonded to the building structure.
Some of the worst shocks I have personally received while working in the industry were from floating grounds on circuits that were turned off. capacitance and inductive coupling on long wire runs. I have even heard or people getting electrocuted on power wires that are off but not properly grounded. I could write a book on the problems I have found due to a floating ground. And it seems to be getting worse not better with everything going to 24VDC input and outputs. Then they use multiple 24VDC power supplies (due to loading) but they forget to bond the power supplies together so the I/O works intermittently. And they spend hours trouble shooting a problem of their own creation.
I have actually seen a motor start and run with the PLC output off due to floating ground caused the contactor to pull in.
Floating grounds should never be used unless you have given the whole system a lot of thought and considered every possible current path and how you are going to handle a failure.

 

[danw]

Measured 100 mv from the common to ground on the input and 0mv from the common to ground on the output.

What is the AC voltage from AO COM to each of the AI COMs?
What is the DC voltage from AO COM to each of the AI COMs?