I have developed one instrument and for safety purpose. I want to ensure that machine start only when earthing terminal properly connected to machine and voltage between earth and neutral terminal are less than 10V.

I have search on internet but I can't find any method or circuit that detect that earthing terminal connected with machine and is proper.

If any one have idea or circuit than its great help

  • \$\begingroup\$ You could connect a large resistor from hot to earth and check the voltage. But doing so might introduce a greater hazard than using an un-earthed machine. And if the resistor is too small, you might trip a GFI (aka GFCI or RCD) breaker. \$\endgroup\$ – mkeith Jun 12 '17 at 4:17
  • \$\begingroup\$ but phase and neutral terminals are not fix, it will change so i cant define any one as hot . \$\endgroup\$ – unobtanium Jun 12 '17 at 5:39
  • \$\begingroup\$ You could try both ways. \$\endgroup\$ – mkeith Jun 12 '17 at 7:30

What you are asking for is similar to an earth-loop impedance tester. These are not simple to incorporate into your equipment.

enter image description here

Figure 1. Robin earth-loop impedance tester.

I purchased one of these many years ago for industrial use. The Robin representative on the trade-show stand understood the principle of operation but was very coy about disclosure. One of the selling points is that it can test the loop impedance without tripping the RCD/GFPD. I surmised that it must be running a DC current on the earth line and returning it somehow to the instrument through the building / supply earth-neutral bonding point and L / N wiring. His response made me think I was getting close.

A simple but non fail-safe option would be to check for earth presence in the same manner as the mains testers. Marc's pages has a sample using neons.

enter image description here

Figure 2. Marc's mains tester. All lights on indicates wiring is correct.

If you could couple each neon to an LDR or photo-transistor you could use a micro or combinational logic to determine the earth connection status if live and neutral are connected either way around. Again, this is not fail-safe and gives no measurement of the quality of the earth when present. i.e., A low quality earth may not pull adequate fault current to trip the breaker or blow a fuse.

Neon opto-isolator

enter image description here

Figure 3. A neon opto-coupler. The outline of the neon lamp is visible through the heatshrink tubing. The LDR (light dependent resistor) is probably behind it and facing into the side of the lamp for maximum light capture. A similar arrangement could be used to couple a neon and photo-transistor.


simulate this circuit – Schematic created using CircuitLab

Figure 4. Possible configurations.

You'd have to do some testing on these circuits. The LDR might be simpler in that its slow response time might even out the flicker of the neon due to the AC signal. Swap the LDR1 and R2 positions to invert the logic, etc.

  • \$\begingroup\$ i want to ensure that earth terminal is connected and quality of earth is good. but at this time from other two terminals i don't know which one is phase and which is neutral. \$\endgroup\$ – unobtanium Jun 12 '17 at 11:16
  • \$\begingroup\$ I understood both of those points and addressed phase reversal in my final paragraph. Please use capital letters where they are supposed to be. \$\endgroup\$ – Transistor Jun 12 '17 at 11:25
  • \$\begingroup\$ please attach circuit that how can i use LDR and photo transistor . \$\endgroup\$ – unobtanium Jun 13 '17 at 6:05
  • \$\begingroup\$ See "Neon opto-isolator" onwards. \$\endgroup\$ – Transistor Jun 13 '17 at 6:30

If you understand the impedance of Neutral to ground and that distribution losses at rated current for service ( e.g. 100A or 200A) allows <5% drop then you would expect 5% of 120V max or 6Vrms or 8.5Vp of line frequency plus harmonics and possibly more with motor start surges and load dumps depending on bandwidth of measurement.

Thus is you expect 6V drop on 100A then you estimate the neutral wire resistance is 60 mOhm + X(f) where line length and L/m can be estimated for inductive reactance for Line and ground loop.

I would suspect the easiest way to test ground continuity and not trip the RCD/GFPD is to use a pulse of sufficient short duration to not give the mA-s duration needed to trip it. I dont know what this threshold is but like any relay can be estimated from power (mW) and transition time and converted to mJ.

Thus the best test for me would be open circuit voltage with 100Hz LPF and AC induced pulse and detect current flow. Let's take a 10kHz burst at 15V with a coupling cap with an impedance of 15 ohms at 10KHz using 1uF non-polar with a ground side current sense R of 1 ohm , we could expect to see 1Amp and 1V. If you drive with a 15Vpp square wave and use a unity gain 10kHz BPF I would expect 1.27*15Vpp from Fourier series for the fundamental ( where the odd harmonics reduce the apparent Vpp amplitude for the square response.) Thus I would expect you could make a comparator trigger a latch or better use an envelope detector require 2 pulses of 1A to reach a pre-determined threshold of about X to validate the Neutral ground connection.

So your self test would consist of if Voc (@line f) < 10Vpp then test with switched 10kHz and detect loop current > xxx mA which should take about 50ms max.

You can choose any other parameters to validate loop impedance considering loop inductance and resistance of both wires without creating any EMC disturbance.


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.