EMF problem?

Hello All. In an effort to avoid water damage from in-house supply, I have developed a system to disable power to our water pump in the absence of human activity within our house. There is a PIR on each floor, the status of each is sent via RF transmitters to a receiver. If no 'movement detected' is received from any of the transmitters for 30 minutes, the N.C. contacts of a large relay (located between the circuit breaker and the water pump circuitry) open, depriving the the 220VAC needed for pump operation. The 115VAC coil of that large relay is controlled by a small 5VDC relay (commonly found on ebay, etc.), driven by the Nano.

It seems to work fine up until the time the large relay enabling the 220VAC operates. When it does, the Nano appears to freeze. The Nano is embedded in a PCB and all the electronics is in this box: |373x500

In the image below, the PCB with the Nano is in the upper left. The U shaped object the the right of that is a place where I originally had the 5VDC relay mounted. To the left of the PCB is the 5VDC power supply. That large open area in the middle of the box is where I originally had the large relay located. |373x500

The image below shows where I now have the large relay and the 5VDC relay located, in the circuit breaker panel. |373x500

On the front door of the electronics box are 6 led indicators. One of which is a 1HZ heartbeat. This 'heartbeat' (along with all the other circuitry) ceases to function when the large relay is energized. I must point out that even though the large relay is being energized, there is most likely no current flowing between contacts since the pump circuitry is only calling for water for short periods of time, intermittently within a day. The point I wish to make is that any EMF problem there may be is doubtfully caused by any current flow through the large relay contacts.

I initially thought the 'freeze' problem was due to the close proximity of the large relay to the Nano so I decided to move it to it's current location, inside the circuit breaker panel. I also moved the 5VDC relay there too, enabling me to run a low voltage cable to the circuit breaker panel from the electronics box. That cable is stranded 3 conductor, 22 GA, unshielded. The 115VAC for the coil of the large relay is fused, If I remove the fuse, disabling the large relay, the 5VDC relay still operates and does not cause the 'freeze'.

The total current draw with the 5VDC relay energized is 142ma @ 5.2VDC.

Schematic below: |500x286

Does anyone have any suspicions as to what is causing the 'freeze'? Have any ideas of things to try in an effort to troubleshoot or perhaps fix this problem?

  • Scotty

EMF = ElectroMotive Force, aka voltage.

Did you mean EMI aka ElectroMagnetic Interference or RFI aka Radio Frequency Interference?

You say "to-may-to" I say "to-mah-to"... In this case EMF is causing EMI of whatever frequency such that the Nano "freezes".

Moving the relays was a good move. The EMF(I) produced when the 115V coil is energized is getting back to the Nano via the 5V relay wire. As an experiment, can you move the 115V relay farther away from the 5V one?

Is a solid state relay an option for you?

The usual terminology is EMI/EMC. Electromagnetic Interference http://en.wikipedia.org/wiki/Electromagnetic_interference and Electromagnetic Compatibility. http://en.wikipedia.org/wiki/MIL-STD-461

How your design does with other's emissions, how much your design emits.

Is the box itself grounded?

I don't see a diode on your 5V relay. What about the inductive kick when it turns off?

Thanks All, for your responses.

Is a solid state relay an option for you?

Not for the large relay. I need a double throw relay so I can make use of the N.C. contacts to power the pump circuitry in the event of deenergized control circuitry.

Is the box itself grounded?


I don't see a diode on your 5V relay.

Wow! Good eyesight! Yes there is a diode across the relay.

I'm gonna dig around in my parts to try to find an opto coupler and put it in line with the cable going to the control relay. Doesn't seem necessary since the relay module already has one but I can't think of another thing to try.

  • Scotty

I can't see your power supply for this system. {The photos are a little too narrow to see the whole configuration.}

Where is it?

Here’s a photo of the supply. It’s just the guts from a walwart rated at either 1000 or 2000ma. Can’t remember which. - Scotty

If the box is grounded, I assume it is to an AC ground. Your DC supply should not also be grounded there. The DC ground should be maintained separately, isolated from the AC ground.


All the electronics is isolated from the box. - Scotty

if you already have a flyback diode (which you said you do), hook an oscilloscope to the Vin pin of your arduino and run the program (to the point where it normally fails). I suspect you are seeing a spike on the vin pin, causing your arduino to freak out.

Is the power supply powering the arduino powering anything else in the circuit?

Is the power supply powering the arduino powering anything else in the circuit?

Yes, the control relay module and the LCD display. - Scotty

If you have access to a oscilloscope, pin out Vin like i mentioned and see what the voltage looks like through out the course of the program. I doubt you would catch it on a multimeter.