No-Ground detection circuit

Hi,

One problem we have with our machines is that customers will plug them into circuits that don't have the ground connected to earth ground, or just be lazy/stupid and not connect it to earth ground. Either way, this causes us problems.

90% of our support is phone support, so I can't look at the wiring that the customer has. What I want to do is put a light on the machine that indicates a proper ground. I know that a GCFI works by comparing the voltage on N to the voltage on ground, and trips if they approach each other. In this case, however the ground would not be connected at all.

I have seen music products that have an LED for when the ground is properly connected. How does one detect when the ground is connected to earth ground vs. to nothing?

Thanks! Tom

Reply to
TBass
Loading thread data ...

If you have a capacitive filter on the line in you could measure current in the ground leg.

Or - if it causes problems - try to detect the problem.

Thomas

Reply to
Zak

connect a neon indicator between live and earth.

Bye. Jasen

Reply to
Jasen Betts

Why not putting similar construction as used on those mains outlet testers to your machine. Anyway you need to make sure that the test circuit does not cause dangerous leakage to the equipment case if it is not grounded... Something like those commercial devices:

formatting link

Your knowledge is not correct!

GFCIs are implemented usually in such way that phase (live) andC neutral wirespass through a sensor coil. If currents are equal there is no net magnetic field in the coil. There should be 0A differential between hot and neutral if there is no leakage to ground. If live ant neutral currents are uneqal because some current is leaking to earth, a voltageis induced in the coil and that activates a circuit breaker. Simplest RCDs have just a toroidal transformer, the L and N being monitored being fed throughthe middle, with the secondary feeding a trip solenoid that trip the switching element. Some more complicated ones have electronics in them to process the input signal.

One example circuit seen on web for grounding testing is at

formatting link
here the "Earth checker". This circuit has one problem if attached to the equipment, this can cause over 1 mA leakage current to equipment case if the equipment is connected to non-grounded outlet.

--
Tomi Engdahl (http://www.iki.fi/then/)
Take a look at my electronics web links and documents at 
 Click to see the full signature
Reply to
Tomi Holger Engdahl

Careful! If you do that and the equipment is not properly grounded, then the current lighting the neon bulb will be going through the person touching the machine.

Reply to
Thomas Veik

This 'trick' requires a series resistor in the 500 koms - 1 Meg range. That limits any current through both the neon and the person to a safe value.

Keep in mind though, that that resistor should be fit to cope with the peak mains voltage. (Some 175 volts in the US, 350 Volt in Europe. And your common resistor won't match the latter !)

--
Kind regards,
Gerard Bok
Reply to
Gerard Bok

yeah... same as those screwdrivers with the neon in the handle, the current isn't enough to cause any problems.

--
Bye.
   Jasen
 Click to see the full signature
Reply to
Jasen Betts

Why a neon bulb as opposed to a incadescent or LED?

What I'd like to do is get a 24VDC digital output if the ground fails so that my controller can alert the interface of a ground problem and alert the operator. I suppose I could use a comparator, and the relay the 24VDC from an external power supply.

[snip] This 'trick' requires a series resistor in the 500 koms - 1 Meg range. [/snip]

How did you make that calculation?

Thanks everyone!

Reply to
TBass

Because a neon bulb glows on a much lower current than a LED. (Well, at least that used to be true.)

I didn't. I just read the color bands :-)

Actually: if you buy a panel mount indicator over here, you can find a 220Kohms upwards series resistor mounted.

The screwdriver type indicators rely on passing current through your body and they typically use a much higher resistor. Somewhere in the 500 Kohms - 1 Meg range. At 230 volts, with a 65 volt drop over the neon, you end up with

0,3 mA or less.

For comparison: the now common two-lead mains testers use leds, but require the return wire to be connected to either neutral or earth. And a ground current protector typically trips in the latter case :-)

--
Kind regards,
Gerard Bok
Reply to
Gerard Bok

reasonable light output for very low current

Reply to
Jasen Betts

That's true. You can convert this kind of sensitive neon bulb signal to digital signal with help of some sensitive optical sensor. Pack the well insulated neon bulb (like that panel mount indicator) and the sensor of that optical sensor inside some some suitable black plastic box or heat shrinkin black insulating tube.. Idea is that only the light form the bulb can get to the sensor. Select sensor a that can generate needed 24V signal.

Other idea: Take a sensitive optoisolator (AC input type or normal LED input type with external protecting diode). Put the optoisolator input in series with the neon bulb + resistor circuit. You wil get that current that goes though that also through optoisolator sensor. Have suitable circuit on optoisolator output to amplify circuit. If suitable optosolator and amplifying circuit is selected, I think that you shoudl be able to detect fraction of mA current that passes through...

--
Tomi Engdahl (http://www.iki.fi/then/)
Take a look at my electronics web links and documents at 
 Click to see the full signature
Reply to
Tomi Holger Engdahl

I have an LCD based mains tester. No battery AFAICT. The display shows

12 V, 36V, 55V, 110V and 220V.

There are two return/touh contacts; one is for teh voltage indicator; the onther completes a more sensitive circuit that allows passive tracing of the phase. You can figure out at which side of a round cable is neutral and which side is phase.

formatting link
- available on ebay for less than 2 dollars. Mine came in a set of 2 other items, one a neon bulb indicator thad the other a battery powered that lights when you touch the rear and the blade. No idea what it is used for.

Thomas

Reply to
Zak

Continuity testing :-) Pick up a fuse (or a bulb, switch, whatever). Hold the rear, touch the blade with the fuse :-)

--
Kind regards,
Gerard Bok
Reply to
Gerard Bok

can do the same thing, but also for testing lightbulbs etc...

Bye. Jasen

Reply to
Jasen Betts

seems like a simple solution but I wonder though if the machine is large it may have enough capacitance to light the globe even though it is not earthed. Also, since the current is tiny this wouldn;t detect a 'bad' earth.

David

Jasen Betts wrote:

Reply to
quietguy

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.