Difference between isolated and non-isolated CompactLogix AI modules

Is there any way to get a non-isolated IF8 CompactLogix I/O module to behave like one of the isolated IF4I modules? I'm wondering if I can leave the ANLG COM floating and that will work.

The issue I have is I have mixed 2-wire (loop powered) and 4-wire (externally powered) devices and I bought an IF8 card. I'm hoping there is some way I can get the card to work for both types of inputs. My 4-wire devices are powered by remote external power supplies and I can't link the commons like the IF8 documentation says (my inputs are coming from a Micromotion and another flowmeter, both 120VAC powered with their own internal power supplies). I'd like to wire the card so it behaves just like a simple 250 ohm resistor, I have the loop put the 4-20mA across it and it reads the 1-5VDC. Is that configuration possible with an IF8 or do I have to get an IF4I?

Thanks in advance.

Greg

Uhh, just use a signal isolator in front of the input. Seriously.

If the inputs WERE isolated, it would be sold as an isolated input module, and be more expensive. Isolated inputs need, well isolated power supplies and A/D converters for each channel.

You can try wiring multiple devices to a single ended AI card. It'll work until the common mode limit is exceeded. So don't be surprised when all the signals drive offscale when a 2nd or 3rd or 4th input is added. You just might get by with 2 devices.

But like rdrast says, if you do need isolation, you either need to replace the card or use stand-alone isolators.

You WILL notice a rythmic jitter on the "non-isolated" modules for more sensitive AI circuits. You may need a scope to see this but the non-isolateds strobe through using a single A to D converter for all channels rather than a separate one for each which would allow constant power. So be aware that you can blame some jitter (rythmic) on less exopensive, non-isolated modules...niot usually a concern but can be misinterpreted as "resolvable" noise" but in fact is just a characteristic of the beast.

Going to use isolators

Thanks for the feedback. I ordered a couple of the 931 series Passive isolators from AB (the 931(S/H)-A1A1N-IP series)...those appear to be cheaper than buying the isolated card for the PLC and I'm hoping those will do the trick.

I've seen folks blow channels on their AB I/O cards from that exact scenario. The I/I isolates the current but not the voltage. If you exceed 30 volts the channel is gone

How would I exceed 30 volts?

JoeFin,

I'm not sure I understand your concern. The way I understand the I/I isolator is it is basically a device that takes a current reading on it's input side and then throttles the current on the output (like a 2-wire transmitter). My AI card is on the output side of the isolator and that side of the isolator is being driven from the 24VDC supply that is powering my PLC, so I don't see how there could ever be more than 30V on the output side of the isolator...

-Greg

radfahrer said:

JoeFin,

I'm not sure I understand your concern. The way I understand the I/I isolator is it is basically a device that takes a current reading on it's input side and then throttles the current on the output (like a 2-wire transmitter). My AI card is on the output side of the isolator and that side of the isolator is being driven from the 24VDC supply that is powering my PLC, so I don't see how there could ever be more than 30V on the output side of the isolator...

-Greg

Click to expand...

Been there / Done that

Basically blew an input on the card just to prove to the dimwitt in charge its a voltage thing.

But I didn't even feel bad - the AB Start Up Engineer (pretty fancy sounding title) blew a 755 Drive mother board, 3ea 755 Drive I/O Cards, and 2 or 3 input channels on the AB I/O card. I think it was about $6-7000 in cards

As for you I/I it uses the power from the foreign source to create and regulate the 4-20ma output. The current is isolated but not the voltage

And yes I looked at all the applicable connection methods before making the recommendation to go with the Isolated I/O cards

The problem was once they started using Foreign sourced power to provide 4-20ma inputs it elevated the ground potential - But Oh !! - every thing is floating in Power Plants !!

The 24v 4-20ma signal goes into the input card, drops approx 12v in the process and then is tied down to DC com. - but it still has 12v !!

The difference between that 12v and the 24v supply is 36v. and that is what you'll measure too. Only problem is AB cards let the "Magic Smoke Out" at 30v

I had to re-read JoeFin's post twice to catch what he was saying. My initial reply would have been radfahrer's reply.

from A-B's spec sheet:
Passive Isolator, 1 Channel
931H-A1A1N-IP can be used for the galvanic isolation of standard 0(4) ... 20 mA signals.
They are supplied by the measured signal and require no additional auxiliary supply.
The measured signal is transmitted in a ratio of 1:1.
http://literature.rockwellautomation.com/idc/groups/literature/documents/in/931-in001_-mu-e.pdf


---------------------
The graphic above does say it is current (I)/current (I) isolated.

The graphic does not include U (voltage) like the ones I've used:



How something could be "galvanically isolated", but not have the input voltage isolated from the output voltage is something I don't understand.

I'm not sure what the 931 device is . . .

I'll try to clear all this up.

In the Non-isolated Input cards the A/D is tied to Ground and if in any way you exceed 30Vdc it's gone

So unless DC com at both sources has a common reference to ground there is no way you can ensure that.

Isolators are working fine. And I'm not seeing JoeFin's voltages

JoeFin,

I installed the 931 Isolators and the behavior I'm seeing doesn't match what you were talking about. These isolators are basically mini power supplies which are driven by the input...the 931H models don't have a cover on one side so you can see the circuitry which is basically a big toroid with windings, so the isolation is galvanic and not optical (the windings are over the same toroid but independent). There is total isolation between the input and output (measure open circuit with a meter). And I measure exactly what I expect on the output of the isolator when its connected to the IF8 card...5V at the IF8 AD input relative to my PLC common when the input to the isolator is 20mA.

By having the (-) of the isolator output connected to PLC DC supply common, that ensures a common reference and the isolator is designed to produce 20mA max across its output, which going into the 250 ohm input impedance on the AI card gives 5 volts.
The voltage on the input of the isolator is irrelevant (as long as it's not enough to break the isolation, ie 100's of volts)...all it's going to do is drop by something a bit more than 5V at 20mA (conservation of energy). ie, it could be going from 45V to 38V relative to my PLC common, but it doesn't matter since the output of the isolator is what matters and that has the (-) tied to PLC common.

That's what I observed, I figured I'd report it in in case others plan to try this. If you have a integrator agreement with your local AB rep, you can get these isolators are ~$30 a channel, I think that makes them cheaper than getting the IFCI card. I agree JoeFin that the fully isolated card is probably the best design (less hardware, optical isolation), but if cost is an issue, these worked fine for me.

Regards,

Greg

Greg I'm glad the application worked for you

I guess I should have prefaced my concerns with more application specific details.

Where I ran into trouble using an I/I in leiu of Isolated AB cards was in a 700 Mwatt facility with individual 480v transformers equiped with "High Impedence Grounds" supplying the various buildings. As with some power generation facilities they elected to leave the DC Common "Floating". Not some thing I endorse having spent most of my time in Oil Refining where every thing including the DC Common is grounded.

The problem arose with the AB 755 drives where as per manufactures instructions in relation to "Drives supplied by High Impedence Ground Networks" the bonding jumper was removed. At this point the DC common is so sufficiently unrelated to ground it could and would float where ever it felt like it.

At one point I presented the IEEE and ISA standards pertaining to isolated grounding to the electrical erngineering staff and sat patiently by my phone awaiting a conference call which never happened - why because some one would have to admitt fault and assume financial respocibility for the Water Treatment facility which was now running in LDs.

I much more prefer what I consider the correct method of grounding like I am experiencing now on a 450 Mwatt facility I engaged in

My point being - The Non-optically Isolated I/I WILL NOT limit voltage. If the voltage inducing current across the bridge is 36 volts above the DC Common ground referece (or lack there of) of the PLC the voltage at the card will be 36 volts. And if your dealing with AB you all know what that means.

From the perspective I was first introduced into this profession in of producing control systems and shipping equipment to the end user, if you couldn't guarantee the voltage wouldn't float to an intolerable level then you would need to take measures to insure damage couldn't occure from it.

Like I said - I'm glad every thing worked out well in your application. If your refering to a skid mounted equipment/control systems, then obviously your much less likely to encounter conditions such as I've outlined. Take GE Frame 7 generators / Mark 6 control systems for instance - every thing is optically isolated.

But if it was me - I would sure feel better having qualified start up techs checking voltages before pushing in fuses