1
\$\begingroup\$

I have a function generator and I noticed that its DC GND is internally wired to its earth pin.

Similarly I have an oscilloscope and I noticed that its DC GND is also internally wired to its earth pin.

So now If I use this function generator with this oscilloscope(with 3-prong power cords) wouldn't there be a ground loop?

If so, should I use 2-prong power cable for one of the instrument to eliminate it as a solution?

EDIT: An illustration for the question:

enter image description here

On the left side the box represents the function generator, on the right side the box represents the oscilloscope.

Circles represent the BNC connectors.

Number 1 and 3 are analog grounds of the function generator and the scope.

Number 2 is the GND shield of the coax cable, where the coax cable is connected from function generator to the scope (i.e. function generator and scope GND are connected)

Number 4 is the earth wire

L, N, E are the line, neutral and earth pin of the 3-prong power cable.

\$\endgroup\$
6
  • 1
    \$\begingroup\$ Does it cause a problem when you hook the two together? \$\endgroup\$
    – W5VO
    May 24 '16 at 14:29
  • \$\begingroup\$ I dont know but I thought there must be a principle \$\endgroup\$
    – user16307
    May 24 '16 at 14:36
  • \$\begingroup\$ If two sytems are earth grounded I mean \$\endgroup\$
    – user16307
    May 24 '16 at 14:36
  • \$\begingroup\$ Currents tend to travel the path of least resistance. \$\endgroup\$
    – PlasmaHH
    May 24 '16 at 14:44
  • \$\begingroup\$ How bout ground loops? Wouldnt they effect the measurement? \$\endgroup\$
    – user16307
    May 24 '16 at 14:57
2
\$\begingroup\$

To measure high-voltage signals up to 40KV, we used 2 channels of the scope and 40KV probes connected to the device being tested. The scope inputs were grounded to earth gnd. In this case the signal ground was NOT used, as the voltages would have arced inside the oscilloscope and destroyed it.

The 40KV probes divided the signal by 1,000 so it was within the limits the scope could read. The 2 channels were combined with software to yield a differential result, seeing only the signal across the 2 probes.

The same setup can be used to measure lower voltage waveforms without any ground-loop issues. Most scopes have a 50 volt signal limit but the probes that come with the scope normally have a x10 switch built in, allowing a maximum signal level of 500 volts without using special HV probes.

If the device being tested has a floating power supply, or you use a isolation transformer, then issues with ground loops go away.

\$\endgroup\$
7
  • \$\begingroup\$ I edited my question and added an illustration to make it more clelar. Do you agree there is a ground loop in this setup? Through 1 - 2 - 3 -4 \$\endgroup\$
    – user16307
    May 24 '16 at 21:36
  • \$\begingroup\$ @user16307. As you have drawn it, yes there is a loop, but the loop has no effect unless you induce a strong signal (high-current) into it. If each instrument has a gnd post, run a short piece of 14 awg wire between the 2 to minimize loops. Remember there is no current in this loop unless you inject it from a strong signal or power source. Your function generator by itself puts out only milliamps of current at most. \$\endgroup\$
    – user105652
    May 24 '16 at 21:47
  • \$\begingroup\$ Im not sure what you mean. In data acquisition systems these loops are serious issue. \$\endgroup\$
    – user16307
    May 24 '16 at 21:50
  • \$\begingroup\$ I would consider buying a small isolation transformer for just the scope. I have used them myself for the reasons you just gave. Using an ungrounded plug can create a shock hazard unless you use a isolation transformer. \$\endgroup\$
    – user105652
    May 24 '16 at 21:55
  • \$\begingroup\$ Using a scope or a function generator with ungrounded plugs can create a shock hazard. I can understand that it is not very safe. But are there legal obligations to use them with earth grounded considerations in work places? \$\endgroup\$
    – user16307
    May 24 '16 at 21:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.