<NOT Solved> Flyback power glitch

[papa]

First. I had a nice Fathers' Day & also wish that to other male parental units out there.

I tried another AC light switch node with a miniwireless & had some success with replacing a light switch with a node & solid state relay (SSR):  The node toggles the light off & on nicely.  However, at least one problem:  The box also has a fan switch daisy-chained from where the light switch was.  The fan's switch still turns it on.  However, now when the fan is switched off, power (flyback??) flashes the light & seems to disable the Arduino functioning a while.  The latter is not permanent so far, but I'm afraid it might get there so I've disconnected the fan from its switch for the time being.

Perhaps the following pic will show the original wiring & the captions will explain how I wired in the node.  (Click on pic for larger view)

     

I'd appreciate advice:  how can I wire this differently or additionally (some kind of diode ??)  to eliminate the flyback power & its potential damage.

[greginkansas]

I would go for the power first-
1 big cap on power, after the supply before the node
2 use a 9 volt supply then a 5V regulator then the node
3 move the node away from the box with a longer 110V wire 3 feet to check for EMF

[papa]

Thanks for responding, Greg. Let me try to clarify / respond to your suggestions. Thanks in advance for your patience with my question / concerns.

1) Do you mean putting a big capacitor between the DC positive & negative coming from the adapter before they connects to the node to power it?

2) I had had a 9 volt power supplied node (Buono + Shield + Solid State Relay in a box) on the bathroom counter that toggled a small lamp. I'm trying to address the appearance / safety of that by containing most everything (except the DHT sensor) inside the switch box. I suppose a 9 volt power supply & a 5v regulator could be used as a test, but would be difficult to add INSIDE the switch box.

3) Again I suppose this could be a test to see what might need to be addressed. Otherwise, it's like the node I had on the counter before.

Also if the problem is a collapsing magnetic field (in the motor when its power goes off) that is feeding back power to the light & the miniwireless Arduino's power adapter, I need help understanding how your suggestions address that.

PS I consulted with a couple people plus a lot of internet research before I posted my need for help.

[ragoth]

Hi Papa, I suggest you to connect snubber circuit across the Fan switch. Ie.,a 39 Ohms 0.5 watt resistor in series with a 10nf 400 volts capacitor and the other ends connected across the Fan switch. This will suppress the spark produced by the fan switch when it is open/ Off and in turn suppress RFI produced by arcing. This RF interference can disturb the circuits in close vicinity. Always keep all wires, especially the Analog and Digital input wires as short as possible to avoid stray pick up of this RFI.

[papa]

Thanks, ragoth, your response along with Greg's starts to help me understand better.

In the updated diagram below, I have labeled Switch B's (for the fan) connecting points with letters F, G, & H. You say "connect a snubber circuit across the Fan switch," but for me to be sure, could you please use those diagram references to say where to connect the resistor & the capacitor?  Much appreciated.

     
           ^^ Updated switch box diagram (esp. more labels on fan switch).  Click on pic for larger view.

You also said, "This RF interference can disturb the circuits in close vicinity. Always keep all wires, especially the Analog and Digital input wires as short as possible to avoid stray pick up of this RFI." This may have been part of what Greg was talking about. For the time being (while I was seeing how to position things in the box's tight space), the Solid State Relay's AC terminals are connected to fairly long pieces of unshielded wires. I was probably going to shorten those wires, but this possibility adds to the likelihood.

PS ragoth, do you mean something as simple as the pic below that I got from here?
     

[papa]

More info that might help: I looked into the ceiling box that holds the fan. I cannot see the wiring connecting to the switch, just a short appliance type cord that connects the fan to an outlet in the fan housing AND another short appliance type cord that connects a light bulb socket (that I did not know was there) to another outlet in the fan housing.

The above information plus both black & red wires relating to the fan (wire bundle #2 in the diagram) makes me wonder if this is a variation of wiring a ceiling fan. From what I've read about wiring ceiling fans, I'm getting hints that my exhaust fan may not have been wired right in the first place.  Maybe #2's black & red wire & perhaps wire C at the bottom of the diagram are not used properly.

What do others think about this?

[greginkansas]

Jun 20, 2016 14:59:11 GMT papa said:

1) Do you mean putting a big capacitor between the DC positive & negative coming from the adapter before they connects to the node to power it?

Yes this might do it

2) I had had a 9 volt power supplied node (Buono + Shield + Solid State Relay in a box) on the bathroom counter that toggled a small lamp. I'm trying to address the appearance / safety of that by containing most everything (except the DHT sensor) inside the switch box. I suppose a 9 volt power supply & a 5v regulator could be used as a test, but would be difficult to add INSIDE the switch box.

This might be a clue - was this powered by the same circuit as the new node?


This confuses me, "appliance type cord " is this in the fan?

More info that might help: I looked into the ceiling box that holds the fan. I cannot see the wiring connecting to the switch, just a short appliance type cord that connects the fan to an outlet in the fan housing AND another short appliance type cord that connects a light bulb socket (that I did not know was there) to another outlet in the fan housing.

[ragoth]

Hi Papa you got it right. You may see switch's connecting terminals just behind the switch. Connect at it as close as possible.

[papa]

Thanks, ragoth, I've ordered the resistor & capacitor to give it a try. Til then will leave the fan disconnected from switch.

Greg asked, "was this powered by the same circuit as the new node?" Uhh, depends on what you mean by "powered by the same circuit." The previous node (now removed) was completely outside the switch box that now contains the second node. However, the previous node's 9 volt adapter was plugged into a nearby outlet that shares AC power with the switch box.

Greg asked, " 'appliance type cord' " is this in the fan?"  In the ceiling box that houses the fan, I found the fan & a previously unknown light bulb socket, both connected to AC outlets in that ceiling box housing via short appliance type cords.  The point I was trying to make:  Having both red & black wires relating to the fan plus having a (another) light next to the fan reminded me of a typical ceiling fan (with the long blades) with a light kit.  I was trying to ask, "If the wiring in my switch box is supposed to be similar to wiring a switch to a ceiling fan (again the kind with long blades), is the original wiring of my switch box correct or should it be corrected in some way?"

[papa]

Today, I received the 39 ohm resistor & 10nf 400 volts capacitor for the snubber circuit.

I connected resistor & capacitor in series across the two switch connecting points.

(Thanks, ragoth.  The snubber circuit works pretty well: When the light is off via the node, switching off the fan does not usually flash the light, especially if the fan has been run more than a few seconds. & this is true even though I did not yet shorten the wires related to the miniwireless & the solid state relay on a chip. If the fan is turned on & then quickly off, it does flash the light but that seems less pronounced before. Another way avoid the lights being flashed when the fan is switched off is the have the lights on & then turn off the fan.

I guess a spark can happen in the switch when the fan is switched off & the spark can briefly trigger AC flowing thru the solid state relay & flashing the lights that are supposed to be off. I'm surmising that the snubber circuit discourages the spark & the unwanted flash of light.

[papa]

The snubber circuit worked fairly well for a while then went back to the same problems. I tried higher value resistors & capacitors, but got no improvement. For now, I don't know how to fix this.