Ok to use single ended instead of differential when common is not referenced to earth ground?

[davey31415]

I am trying to intuitively learn the differences between single ended and differential analog signal inputs.  I've looked at Table 1 here several times in the past but still find myself questioning the differences.

When a lead is not connected to earth ground the signal is considered "floating," correct?  Are there any other cases where a signal would be considered floating?  I guess I could instead ask if there are any times other than when a signal is referenced to earth ground that a signal is not floating?

It is okay to use a single-ended input when the leads are connected to ground by resistors, correct?  (top left of the Table 1 diagram).

If the common for all my signals is the same, what are the advantages in using differential? 

Are there any times when single-ended inputs would be superior to differential (other than cases of using only a single wire and using half the inputs on the device) ?

Most of the time I am acquiring signals from DC power supplies (some isolated, some not).  If all my signals are originating from the DC supply, it is not clear to me if I can use single ended inputs for both isolated and non-isolated supplies.  Table 1 clears things up but the answers still do not come to me intuitively.  A list of advantages and disadvantages of using the various modes in different situations would be nice to have.

[Pnt]

My English are not fluently but..

"When a lead is not connected to earth ground the signal is considered "floating," correct?"

The signal is consider floating when it does not have the same ground with your device. Earth has nothing to do with it. Earth is just another ground.

"Are there any times when single-ended inputs would be superior to differential (other than cases of using only a single wire and using half the inputs on the device) ? "

As far as i know it is more easy to measure a sinle-ended input, so there is better precision, higher sample rates....(assuming you have similar devices)

"I am acquiring signals from DC power supplies ..."

You can, if your measurment system has the same ground with the DC power supply.

Always use single-ended, if you can.

[smercurio_fc]

---

Pnt wrote:

As far as i know it is more easy to measure a sinle-ended input, so there is better precision, higher sample rates....(assuming you have similar devices)

---

What do you mean by "more easy"? As for better precision, this is based on what? In fact, the whole point of differential measurements is to allow the amplifier to remove common-mode noise and common-mode voltage, thereby giving you a more accurate measurement. Also, what does sample rate have to do with single-ended vs differential?

---

Pnt wrote:

Always use single-ended, if you can.

---

I would say the exact opposite.

[Coq_Rouge]

differential measurements work best then you are dealing with true differential signals. One way to accomplish this, is to use the SSM 2142 circuit from Analog Devices (or equivalent). The best ting is to do some testing, and select the method that works best on that particular application. Also emphasise on the cabling. In your case I think a shielded twisted pair, with the shield connected to shield ground in only one end may work best

[Pnt]

smercurio

I design my own circuits, so it is easier to make a sinle-ended than a differential. Differential measurments can be tricky.

For those who use  ready hardware, as a NI's device, which have very high quality (and price, as an effect) i think you are right.

I was talking for.... myself.

So yes, what you wrote is "more" right.

[Will_S.]

Floating is the general term for "not grounded." It is typically best to have your signal and measurement device tied to the same ground. 

Differential is great in the fact that it literally takes the difference between two signals. This greatly reduces noise and tends to provide you with a more accurate signal.

This article gives a thorough description of different field wiring techniques and wire diagram comparison:

http://zone.ni.com/devzone/cda/tut/p/id/3344