Floating measurements with a 9215

[RyanBattelle]

Hello,

I am trying to take 4 measurements using a 9215.  The three voltages I am looking at are the following:

3 out of current sense resistors.  Each connection to an AC outlet (line, neutral, and ground) have one of these resistors .05ohm resistors inline.  I am trying to read the voltage across each.

The 4th measurement is across a smaller resistor in a divider network placed between line and ground, this is to measure the actual line voltage.

What I am seeing is a very noisy signal on each.  I have referred to the "Field Wiring and Noise Considerations for Analog Signals" article, and it states that the use of bias resistors is important.  I tied each of the first 3 inputs (+ & -) to COM with a 68k resistor.  My thoughts are to try a larger one....

However, when we get to that fourth measurement and the application of grounding, I become wary.

As I have to keep things floating in order to not violate individual channel maximums (isolation), I am troubled as to how to connect to a reference point.

Any suggestions would be welcome.  A beginning schematic is attached with the initial approach.  Blue wires have happened since then.  (Note that J7 is a jumper block.  Currently the jumper is installed that puts R4 to the ground line from the AC source.)

Thanks

[Paul_M.]

Hi RyanBattelle,

Thanks for the detailed info, it really helps. I'll take a look at your schematic and see if there are any potential issues.

Also, is the noise apparent when only one signal is being measured or is it only when all 4 are connected?

What type of environment are you in?

Paul M

[RyanBattelle]

The noise is on each signal, but in different manners.  In some cases, I wouldn't call it noise.  For instance, the signal on the Line current measurement displays the sine-wave (60Hz) starting to develop problems at the peaks.  The peaks voltage starts to go down, you would say an inversion of the peak voltage heading back to the 0 line.  The line voltage does the same.

The neutral current measurement actually maxed out at +10/-10V.  The ground current measurement looks like a 60Hz sine wave with the same problem that line current has.

Im confident that it is my grounding methodology.  I added a 68k resistor between each signal and common, trying to take care of bias current.  But the "noise" is the same.  

I think the difficulty is the ground scheme...where do I look for a proper reference.

If by environment you mean location of work....benchtop lab space.  So you could have 60 cycle hum from different places (lights, etc).

[RyanBattelle]

An update for you.  I changed the schematic slightly...Instead of connecting COM with AI3-, AI3- is still through the 1M resistor to the ground line and COM is connected right to the ground line.

The input on the AC side is 110Vrms from a wall outlet, the load is a Dell Optiplex computer.  Attached are the waveforms that I get.  I believe the ground waveform...not sure about Line and Neutral.  Neutral maybe, but I would expect no current on that line.  Line should have something in the positive range?  I am trying to find a suitable "pure" resistive load to test with.

[Paul_M.]

Hi Ryan,

Thanks for the additional updates. I'm going to take a look at the schematic with some colleagues to determine whether or not what you're experiencing is expected. As for the Line signal being negative, are you sure you have the leads connected properly and not reversed?

Paul M

[RyanBattelle]

Hey,

Yes, I checked that first thing because it occurred to me I might have swapped them.  But, it is connected correctly.

I put a different load on, a light bulb.  This is a decent resistive load.  The attached screen shot shows the new data.  I believe the neutral as it looks like what I would expect.  And, using the measured line voltage and calculating a power based on the current measured I get .8 * 114 = 91.2 watts.  The bulb is stated to be 100 watts.  That's close enough for me at this point.

The line current, or voltage across that line sense resistor, is still wrong.  I think I am looking at a problem of the potential I am applying at that input.  On both sides of the input (+ and -) the potential is around 115Vrms AC.  Is this leading to the detector being overdriven?  What other options do I have for looking at that voltage differential and measuring the current?

[RyanBattelle]

Sorry, forgot the screen capture on the last one.

[Paul_M.]

After seeing the response with the light bulb as the load, I think things are looking as expected. The neutral current with the computer as the load is going to be variable because the computer is not going to draw the same amount of current continuously like the light bulb does. The ground current is essentially zero so that is looking good as well. From your schematic, you have applied your bias resistors correctly.

Also, looking at your code I would recommend using some of the other DAQmx VIs rather than just the DAQmx read within the while loop. If you just have the read within the while loop it's not going to be as efficient as it will have to reload the configuration settings with each iteration. I would recommend taking a look at one of our example VIs to get an idea of what an ideal DAQmx program would look like.

Paul M

[RyanBattelle]

Im not sure I agree that things are as expected....what about the Line current?  Its still a square wave going negative.  In the case of a resistive load, you should have a sine wave corresponding to the current draw (voltage across the sense resistor).  In regards to applying the bias resistors...there is only one and it is applied differently that should be expeted.  Its on AI3- going to ground, which is the reference for the Line voltage measurement.  While that is giving the correct line voltage, I dont see how its a bias resistor applied correctly.

As far as programming...thats not on my list of issues right now.  If I look at MAX and just go channel by channel, I get the same result.

So my issue still stands, unless you can explain why the square wave going negative is correct?

[RyanBattelle]

Ok..in looking at this, I think I now realize my error.  The channel to earth max is 250Vrms, however the channel to COM is only +/-30V.  As I am putting a line voltage on the input, its overdriving it.  Now I just have to figure out a way to sense it without that problem.

Thanks all.