CGI on low voltages

[dave8976]

Here is an interesting situation that I am looking for any comments and suggestions. I understand that a GFCI receptical works by using a summing toroid. That is, the toroid sums the current in both wires (hot & neutral) and any difference in current will generate a voltage that will cause the GFCI receptical to open the circuit. I believe you cannot use this device directly on a low votage circuit (12v) as I expect there is a limited voltage range that the GFCI device can properly work within (100v - 130v). By placing the GFCI on the primary side of a step down transformer does nothing to protect the secondary side circuitry (12v) for any current leakages. If the secondary curcuitry is rectified to 12VDC an OP amp could be designed to act as an operational differentiator comparing the current leaving and receiving on each side of the 12V circuit. I thought of putting a 1 ohm resistor on each side of the 12 VDC to be use as the input to the comparator circuitry. The current demand of the 12V circuit would be approx 3.5 amps. Am I making this more complicated? Is there an easier approach? The applicaiton is to provide a safe 12 volt source that can be used outside.

[russlk]

I cannot think of any case where 12VDC would not be safe. Why do you think you need GFI?

[dave8976]

If the 12 volt load was an under ground light for a pond, I want to remove any possibility of preventing any current leakage.

[Gary]

If skin resistance is low enough, even low voltage is supposed to be potentially deadly to people. Am not sure of the exact voltage required to "do the job", the destructive thing to living things is amperage, not voltage. Salty or hard water can be a very good conductor. Electrical current is probably bad for fish in a pond also.

[bodgy]

15 - 20mA is the current required. So measure the resistance on your body, it will vary depending on the dryness of your skin and the part of the body you probe - then do the calculations. Though with a digital meter it is harder to get a reading than with an analog type.<p>bodgy

[HomeBrew]

With an isolation transformer or other un-grounded power supply, and a well insulated lighting fixture designed for submersion, you should not have any leakage in the water.

[dave8976]

I agree with the use of an isolated drop down transformer and some rectification to provide 12vDC to a water proof light fixture. But if I connect another lamp (a test lamp)to one side of the 12vdc line and the other end of the test lamp to a ground rod would expect the test lamp to light showing some current is flowing to ground? It is this current that if it exceeds above 5ma I would like to the 12VDc circuit to open. I have not tried this but I expect this is what would happen as a potential difference woud exist between one side of the 12vdc supply an ground.

[russlk]

Your analysis is correct, but you don't know which side of the line has leakage to ground, so there is a 50/50 chance of getting the test lignt on the right line. If one side of the line is connected to the ground rod, near the source, then you can measure the difference in line currents downstream from the ground rod to determine if there is leakage from the "hot" line.

[dave8976]

I should have realized and not made an assumption in my previous note. Basically with the use of an isolation transformer you in fact removed any reference to ground. I set up a test circuit using a 9 volt isolation transform that was connected to a bridge rectifier and filtering circuit. I found no voltage measured from either side of the DC voltage to ground. Thus, I will proceed to build a device that has a GCI receptical on the primary side of an isolation transform and one the secondary side conect the leads to a rectifier with filtering in order to product a 12v DC power source that is isolated from ground. I will also add a current limiter/breaker to handle any shorts that may occur if water gets behind the light fixture. case the