What the heck am I measuring?

Discussion in 'General Electronics Chat' started by JimG, Aug 4, 2010.

  1. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    I started some troubleshooting tonight. An ADC that works fine when powered by a 9V battery (or laptop USB cable) gives bad readings when powered by a simple 9VDC wall wart.

    The ADC is reading thermocouples which are grounded to a brass boiler in a coffee machine. The boiler is very well connected to the earth ground of my electrical system. (Yes, I must use grounded thermocouples for this application).

    I suspect the problem has to do with a ground loop, so I measured the voltage between the (-) output from the wall wart, and the earth ground of my electrical system.

    To my surprise, my DMM is showing a 50VAC potential between the (-) output and earth ground. So I checked a different wall wart, and read a very similar number.

    I'm having a hard time accepting that the output of the wall wart is really parked 50V above ground. So what the heck do my readings mean?

    If there really is a 50V difference, then do I need to add an earth ground to the ADC circuit (which is now floating)?

    Jim
     
  2. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    That is probably some capacitive effect. The voltage reads high because of the meter's impedance. Try reading current, starting at the highest range (just in case). Each range should show very close to 0 milliamps, which would tend to confirm a capacitive cause (stray leakage?).
     
  3. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    Makes sense. As you expected, there is no measurable current.

    Does the lack of grounding of my ADC circuit, coupled with the high apparent voltage across a high impedance, explain the flaky behavior of the ADC when using this wall wart?

    Should I ground the ADC circuit to the coffee machine?

    I do a lot of work with thermocouples, but I still don't have a complete understanding of ground loop problems.

    Jim
     
  4. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Ground loops are almost always caused by more than one path to ground.

    Here is one example. In an AC circuit the outlet has both ground and neutral. Neutral is connected to ground where the current is tapped off the power company at the power pole. However, Neutral and Hot have the same current flowing through the wires. Whatever the resistance of the wire will show up as a voltage due to Ohm's Law (resistance of the wire divided by the current through the wire). You can see up to several volts on Neutral at the outlet due to this resistance, even though in theory it is the same electrical point as ground.

    If you have several paths to ground you can have an imbalance of voltages, or unexpected current paths between what is treated as the same points in the schematic (they aren't really). Hence the name ground loop.

    Since these currents are not planned for they can occasionally cause some very unexpected results.
     
  5. someonesdad

    Senior Member

    Jul 7, 2009
    1,585
    141
    You might want to take a look at an HP app note on measurements, ground loops, and guarding. It might help you with ideas and show how to make a schematic of what's really going on.

    Perhaps connecting some e.g. 100 kΩ or so resistors from both sides of the wall wart's output to ground can shunt off these common-mode voltages sufficiently -- or at least provide an avenue for more experimentation.
     
  6. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Isn't adding the earth ground to the ADC as simple as connecting your ADC supply's negative terminal to the earth'ed side of your thermocouples? That should short out any AC which has been coupled across the transformer in the wall wart. I can't see how it could cause any other problem.
     
  7. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    I think you have to be careful doing that with thermocouples that are also grounded at the tip. Since you would be forcing both ends of the negative tc wire to be at ground potential, you end up shorting out the emf generated along the negative tc wire that results from the temperature difference you want to measure.

    My current plan is to tie the ADC's negative supply (i.e. the wall wart's negative output terminal) to ground, as you suggested, but leave both tc wires ungrounded at the "cold" end. There's going to be a potential difference between the grounded tip of the tc and the ADC's negative supply, but hopefully it will be small enough to cause no problems.

    To all -- thanks for the helpful replies.

    Jim

    Jim
     
  8. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    This is what I meant. It doesn't short out the thermocouple. Does it?:confused:
     
    Last edited: Aug 6, 2010
  9. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    No, it doesn't. Sorry, but I initially misunderstood your idea!

    FWIW, grounding the negative supply (but not changing the thermocouple connections) seems to have fixed the problem. I tried that before I saw your diagram. I think what you've shown will work, too, but right now it ain't broke ....

    Jim
     
Loading...