Bridge Rectifier question

Thread Starter


Joined Jun 18, 2009

I am trying to understand how this bridge rectifier works that is labeled below as MDA 208... This circuit is supposed to be a GFI that is designed to trip whenever it senses a water leak... However, it is constantly "sensing" a water leak and therefore always tripping.

I posted the entire circuit in hopes that it may help solve the problem.

Please let me know any ideas that you may have.




Joined Apr 20, 2004
The only way the circuit can function is for the "leak sense" line to complete the circuit.

The line coming off pin 4 of the optoisolator must be the shutdown function.

That is not the same as a GFI. They look for balanced currents in coils in the AC hot and AC neutral lines.


Joined Jul 5, 2006
That is a truly bizarre leak detector circuit. Where did you find it? Is it a commercial product? Are you trying to detect water on a concrete floor or in a boat or is this for a more exotic application? Are the two high voltages shown supposed to be applied simultaneously or are they two options? If they are two options, why are there two separate but identical inputs instead of one?

As shown, if both voltages are applied simultaneously, the voltage at the upper AC input of the bridge will be zero and there would be no output from the bridge except for that due to imbalance of resistor values or voltages, assuming that the leak is to ground at zero volts relative to the power supplies.

Why use such high voltage for leak detection? If you have a sensing plate out on insulators on the floor you have a potentially very dangerous situation unless the plate is very well shielded. 150 volts behind 3.5K X 2 can supply about 20 ma, more than enough to kill a careless or curious person touching the plate while standing on a damp floor or touching a grounded object with the other hand.

I'd recommend asking for a suggestion for a safer, more practical leak detector circuit, rather than trying to optimize this one.

If you are using rectified line voltage for the 150 VDC supply, your problem could be stray capacitance to ground in the sensing leg. Are you using shielded cable to the sensor with a grounded shield?