When is an earth no good ?

Discussion in 'General Electronics Chat' started by Man_in_UK, May 18, 2010.

  1. Man_in_UK

    Thread Starter Senior Member

    May 13, 2008
    132
    0
    (I have an understanding of component level electronics but not mains installations regulations)


    I have a machine at work that is supplied by 240vac, L,N & E. The machine is also separately bonded to earth.

    If I run a wire directly to the building earth point to use as a reference, I can measure (using a standard multimeter) a resistance of 120ohm between the machine earth wire & my reference wire(supply earth, not bonding earth).

    If I remove the earth wire from the machine and test it I get a good reading of less that 1 ohm to my reference wire.

    I can't understand that this 1ohm wire goes up to 120ohm just from connecting it to "another" good earth. I thought resistors in parallel made the total resistance go down , not up ?

    Help!
     
  2. kingdano

    Member

    Apr 14, 2010
    377
    19
    is one earth connection tied to a more resistive material than the other?


    dont know the answer - just tossing ideas out
     
  3. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    When man has lived on it for a few thousand years. ;)
     
  4. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    Earth ground and circuit ground or (-) are very different things.

    Your not using the "earth" as a return path, your using it as a safety. I doesn't have anything to do with the electronic operation of the machine. It there to run faults to the earth instead of through you.

    A human is a few million ohm, your copper wire is a few hundred. Electricity takes the path of LEAST resistance, so in a fault, it will go to the ground rod before going through you.
     
  5. mik3

    Senior Member

    Feb 4, 2008
    4,846
    63
    The multimeter might show a greater resistance due to a voltage drop across the wire.

    Can you provide a drawing to show what exactly you are doing?

    Also, provide more details about the machine.

    What do you mean the machine is also separately bonded to earth? Which earth?
     
  6. Markd77

    Senior Member

    Sep 7, 2009
    2,803
    594
    The problem might be with the measurement method. Multimeters output a voltage and measure the current to measure resistance. If there is any tiny DC difference between the 2 earths that will mess up the measurement.
     
  7. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    There are standard methods for measuring ground resistance and using a multimeter isn't one of them.
     
  8. someonesdad

    Senior Member

    Jul 7, 2009
    1,585
    141
    When something like this happens, try reversing your ohmmeter leads. If you see a different resistance (sometimes you'll get a negative resistance), then you probably have a nonzero voltage (thermoelectric, galvanic, whatever) on what you're measuring (as Markd77 suggested). Fancier meters (usually bench meters) have an "offset compensated" ohms feature to fix this problem (as long as the voltages aren't too large). These fancier meters measure the open circuit voltage, then supply a constant current and measure the voltage drop. The two voltages are subtracted before the resistance is calculated. Another way is to reverse the current direction, then average the results. Of course, the cost is it halves the reading rate. I've long wished that DMM manufacturers included this feature as an option, as it's met with in the field quite a bit.

    When the potential for this problem exists (no pun intended), a careful investigator will check the resistances with both measuring polarities -- this gives a bit more assurance that you're making a correct resistance measurement.

    On my Fluke DMM, a 1 mV offset voltage makes a 1% resistance error occur on a 100 kΩ resistance. On a 100 Ω resistance, a 1.3 mV offset voltage results in 89.5 Ω and -6.0 Ω.
     
  9. rjenkins

    AAC Fanatic!

    Nov 6, 2005
    1,015
    69
    If each individual earth wire tests OK, then the installation is probably OK - but to be safe, get a local electrician to do an earth loop test on it.

    There are specialised test meters for checking earth loop resistance which use mains power and a high current pulse for the test. Anyone who does electrical installation should have one.

    Examples of one make:
    http://www.protrol.co.uk/Robin Loop Testers.htm
     
  10. Man_in_UK

    Thread Starter Senior Member

    May 13, 2008
    132
    0
    Thanks for your replies.

    It would seem that my meter is not the best thing to use in this test. I did not think to even do a voltage test between my 2 earths to see how much dirt there is, I would never have thought to reverse the probes.

    My mind is at rest, I'm not mad. Not just yet.
     
  11. Man_in_UK

    Thread Starter Senior Member

    May 13, 2008
    132
    0
    OK another thing to put in the pot .........................

    The reason I started with this test is because of a build up of static.
    Recently some old wood parts of the machine were replaced with plastic(quite a lot of plastic) and immediatly after it started giving comm problems. I know these comm problems are static related as the installers fitted extra earth bonding when we last upgraded the controll modules.

    I was trying to find out if our earth bonds were good enough to eliminate static build up. An electrician came to site and made earth loop tests and told me that there is no earth fault.

    But, I still got comm problems !
     
  12. GetDeviceInfo

    Senior Member

    Jun 7, 2009
    1,571
    230
    for your L,N,E supply, 'E' is the bond. Not sure of your regulations, but in North America, conduit made up securely, is sufficient for bonding. You may find that your 'E' line is not continuous to ground as the conduit fullfills the requirement. As well, termination techniques may well have introduced your measured resistance at some point. In addition, your 'N' will be grounded at the supply, so with no energization, you should have no resistance to bond, or ground. Only a visual inspection of your 'E' bond in regards to it's continuity to Earth ground will answer your question.

    In regards to your comms. Make sure your shielding is not grounded on both ends. I will assume that you have isolation on the comms, but you should confirm.
     
  13. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    If it is the static produced by the new plastic parts, you can attach springs to move with the parts to a local ground through a 10Mohm resistor. That should keep the building of the charge at a minimum. You can even use weak springs like a slinky. Just something that will stay out of the way and not load up the moving parts.

    If this is an often replaced part, you should see if the company who supplied them can use ESD safe plastic. It is conductive and will not act as a huge static generator.

    I believe there are sprays that you can use to reduce the amount of static produced.

    If either of these eliminate the problem, have the ESD plastic used next time.
     
  14. Darren Holdstock

    Active Member

    Feb 10, 2009
    262
    11
    Yup, I've been designing earth test instrumentation for the past few months, and it certainly isn't simple, and I quite agree that a multimeter wouldn't suffice. The earth resistances under test are swimming with DC and AC currents that swamp the test signal, plus it takes 3 or 4 carefully placed test spikes driven into the ground just to get a single reading. And that's assuming the plant is disconnected - if it can't be unhooked then more measurements with clamp meters have to be introduced. It's been quite an enlightening project, plus I learned I've been saying "homogenous" all these years when I should have said "homogeneous". Megger's popular publication "Getting Down to Earth" (pdf, 1.5 MB) tells everything you need to know about testing ground connections.

    retched's anti-static suggestions should work a treat. It sounds like the combination of moving parts and new plastic has inadvertently created a van der graaf generator - easily done. My Dyson DC03 vacuum cleaner seems to generate megavolts from the cat fur whirring around inside the cylinder, to the point where I'm nervous about using it too close to PCs and hi-fi etc., plus if I touch the collection cylinder I get a massive belt, which I never enjoy. One day I'll get around to making it static-dissipative with some conductive paint and earthed copper tape and suchlike. It will look terrible, I'm sure.
     
  15. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    Earths are my favourite subject.

    The supply earth and the building earth are not necessarily physically the same, especially in commercial premises. This may well lead to a resistance and/or potential difference between them.

    So the first thing to understand is what you mean by the building earth and what type of earth is provided by the supply.
     
  16. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    You had to wonder why the spell checker was always wrong.

    I had a girlfriend in highschool that I convinced she was saying "ash-tray" wrong. I told her it was "afsh-tray". Even had a few friends say "Oh my god, you thought it was A-S-H ? How old are you?"

    And she was embarrassed. A few times that day, on TV, a show or commercial would say "ashtray" or "ash" and she heard it as "afsh" so she would say "My gosh, your right. Why did I think it was a-s-h all these years?"

    Apparently that night, she told her dad to empty the "afsh-tray" and when he corrected her, she told him he was wrong and grabbed a dictionary.

    She wasn't happy with me the next few days.
     
  17. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    One of the things that can cause problems with intentionally introduced earthing is that earth loops can be inadvertently created. Earth loops are a potential source of noise and interference. For instance, connecting the screen in a signal or comms cable to ground at both ends is a potential problem. Good practice in signal (rather than mains) earthing is (wherever possible / practicable) to tie the various signal grounds in a system to a common earth junction point.

    I'm not sure what your electrician actually tested for - rather than looking for earth loop problems they may have been checking the integrity of the mains earth bonding - which it appears meets the requirements of the relevant standard in your country.

    Caveat: It's absolutely necessary to earth bond any exposed metallic part or surface liable to make faulty contact with the active conductor(s) in mains operated equipment.
     
  18. rjenkins

    AAC Fanatic!

    Nov 6, 2005
    1,015
    69
    An earth loop could cause low level electrical noise but would not cause static build-up

    There are commercial static elimiators available. These typically use a high voltage power unit and an array of needle emitters (contained in a housing for safety) to ionise air and in effect make it conductive.

    They are common in the packaging industry to prevent fine product powder sticking to the inside if the plastic film and messing up the seals.


    Some examples:
    http://www.staticbars.com/
    http://www.electrostatics.com/page2.html
     
  19. someonesdad

    Senior Member

    Jul 7, 2009
    1,585
    141
    Thanks for the PDF link. A couple of months ago I was curious what the resistance of the grounding rod of my house was, so I did some reading and an experiment. I've attached my notes. Our house is pushing 40 years old and about 20 years ago I built an addition onto it. I had an electrician put in the new service and drive a new ground rod, so this was a good opportunity to see what kind of job he did (adequate, as far as I can tell).
     
  20. Man_in_UK

    Thread Starter Senior Member

    May 13, 2008
    132
    0
    I will have a think about creating a schematic layout of this setup but it might be a little involved and complicated.

    I don't have good enough equipment to analyze static or interference on the comm circuits, but I do know that it does not have shielding in both ends.
     
Loading...