Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Differential vs. Single Ended AIs

Status
Not open for further replies.

timebone

Civil/Environmental
Apr 27, 2014
9
Hey guys,

I've been wracking my brain over this for a while now and I'm having a heck of a time trying to identify when to use differential wiring vs. single ended. Basically all our control loops are set as 4-20mA and usually go into a single ended ai card. However, for whatever reason the engineers who set up anything that is EXTERNALLY powered, is set as a differential input. So devices that our 24V supply feeds the transmitter is a single end, and anything with its own external supply is differential.

However, I really don't see the justification here. It seems like we're using more channels than necessary to accomplish things like isolated commons. However does this really matter here? The thing is, all of our external supplies are actually the local 24V DC supply. So essentially, all the commons are the same here.

Ive read that difference in commons could be one reason to use differential inputs. But if we have a 4-20mA loop, does that difference in common voltage really manifest itself as a large amount of error? I could understand that if we were measuring DC voltage... but not current.

Any insight into this? I also know that noise is another common reason to use differential wiring. However, some of these devices are like 3 feet from the cabinet... I can't imagine there's that much noise present.

Thanks a lot guys.
 
Replies continue below

Recommended for you

>we were measuring DC voltage... but not current.
4-20mA always IR drops across a resistor so the AI 'sees' a voltage, and that voltage can include ground loop current drop along with the signal current drop.

As you mentioned, differential inputs can handle common mode noise.

But a differential input is no guarantee of isolation which is what is needed when there's a high level of common mode. The graphic shows common mode Vcm on a differential input.

10wij9j.gif


As long as the field side is floating and isolated from ground, single ended AIs seem to work OK. Most 2 wire, loop powered instruments are floating and isolated from ground so they're happy being powered from your controller's power supply and being wired to a single ended input.

Where you'll see a problem with single ended or poorly isolated differential AI's is when you connect externally AC powered (4 wire) devices, particularly devices like analyzers which seem to always connect the 4-20mA negative to ground.

When the ground potential between the field and the AI exceeds the common mode limit of the AI, the input sees the common mode voltage as an offset, drives the signal off scale or smokes the input, depending on the Vcm magnitude.

There's no free lunch - isolation costs something and well isolated inputs cost the most.

You might be missing the pain of isolation and ground loop problems by using the differential inputs for externally powered field devices, and just not being aware that the isolation is working for you. Can you move some diff inputs to single ended and see what the result is?
 
That's a great response and it actually got me thinking. You're right that gound loop current isn't really any different than ground loop voltage.

But here's my question: In a single ended mode, is the potential measured between the two terminals, or from the + terminal to the card ground? I know they refer to this as RSE and URSE I think (referenced vs. single ended). I'm just asking what is typically done. I'm working with a 1756 CLX AI module. The reason I'm thinking about this is because lets say the transmitter is referenced to ground, but the analog input module is just measuring the potential between the two terminals with no connection to ground at the PLC... wouldn't this inhibit any sort of ground loop issue? So maybe I'm thinking that these cards have some sort of internal connection to the supply ground, or maybe the return wire is actually part of the DC return.

Here's my follow up. I'm assuming that differential cards must have some sort of reading between the output and ground... or else how would the common mode voltage matter here? So my transmitter is connected to ground, and my card will read referenced to ground. This would then eliminate any common mode voltages here.

Like ultimately the differential has to be measuring wrt to something, no? I could understand how single ended could do that, since it's just like a voltmeter probing between two points... but what about differential? It has to be measuring that difference relative to the ground of the module supply, no?

I'm sorry about all these questions, but there doesn't really seem to be any detail on this in the manual. I'm kind of trying to figure it out myself.
 
Keep in mind that single-ended inputs are connected to passive transmitters. The two-wire/loop-powered transmitter acts as a variable resistance, adjusting the current output from the DC supply to the proper mA value proportional to the sensed input parameter. The current goes from the DC supply, through the transmitter, into the card channel input terminal, and then out of the card's common connection back to the DC supply. Since all power originates from a single DC supply, there is a complete path for current to flow for all channels.

Differential inputs are connected to active transmitters. The four-wire/externally-powered transmitter acts as a current source (it's not a variable resistance), adjusting it's output current to the proper mA value proportional to the sensed input parameter. The current goes from the transmitter output terminal, into the card channel input terminal, out of the card channel output terminal, and back to the transmitter's common terminal. If you try to connect multiple externally-powered transmitters to a single-ended card, the current leaving the card's single common connection would be the sum of all of the currents entering the channel inputs. How would this current return to each of the transmitter's common terminals? Hence, differential inputs are needed.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.
 
What do you mean how would it return? The external transmitter has the + and - wire. You're saying the current can't come back through the - wire?
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor