How reduce spikes when switching AC inductive load

[JackofallTrades]

Hello.

I have a system in which a PC switches on an off various loads .

Small resistive loads work fine, as do DC inductive loads with the proper flyback diode circuit.

For one large AC inductive load however (a small household box fan), I am getting system malfunctions presumably from AC switching noise. The AC fan is controlled through a contactor and the PC will periodically hang upon turning the fan off.

For the contactor the input coil has a flyback diode which is working correctly to limit voltage back to the PC. Therefore I presume the noise is the energy from the fans coil collapsing back into the AC line. To get a measure of the fans steady state current, I plugged a "Kill-a-Watt" inline with the switched output to the fan and the problem improved greatly. This also leads me to believing the lockup is caused by noise on the AC side of the system.

I do not have differential probes nor a battery operated scope to look at the AC line properly.

Is there a circuit, corresponding to the flyback diode, to sink current caused by the collapsing field for AC loads?

 

[Orson Cart]

When you de-energise the contactor to the fan, any current flowing in the fan motor wants to keep going and the volts fly up (surprisingly high) across the contactor contacts and radiate an RFI noise spike with oscillation that gets into your PC as a near field radio wave.
The common cure is a snubber across the contactor fan wires, say 220nF 275Vac rated and 100 ohms, 5 W resistor (in series), other cures are high voltage back to back zeners, a bidirectional TVS (540V say), or a MOV rated for the AC line.

 

[JackofallTrades]

Using a solid state relay for an inductive load.

Hello. This is a followup to a previous question listed here.

https://www.edaboard.com/threads/333952/

I am switching on and off a standard AC home box fan via PC control thru a contactor. Switching on is fine. Switching off is generating terrible back EMF that is causing noise spikes. As this is an AC load, I cannot use a flyback diode and my snubber circuits while having some effect are not enabling flawless performance with some PC lockups still encountered, presumably thru the back EMF noise on the AC line, or EMI at the contactor.

I am considering switching to an Solid State Relay. In general how are these at switching inductive loads? How robust do the snubber circuits tend to be?

Do you envision much noise getting back into the AC line driving this load.

https://www.walmart.com/ip/16503230?wmlspartner=wlpa&adid=22222222227009507530&wl0=&wl1=g&wl2=c&wl3=40871930912&wl4=&wl5=pla&wl6=78764700152&veh=sem

 

[dick_freebird]

I imagine this may be involving a ground loop of some sort,
what are you doing to isolate the switched currents from
the control logic ground? Ditto the logic supply?

 

[Vbase]

Solid State Relays (SSR) have integrated zero crossing and that significantly reduces the spikes. Some SSR include snubbers and some don't.

 

[JackofallTrades]

I imagine this may be involving a ground loop of some sort,
what are you doing to isolate the switched currents from
the control logic ground? Ditto the logic supply?

The control logic ground is connected to earth ground through a switching power supply.

The switched load is isolated through the contactor from the logic.

 

[JackofallTrades]

Thanks, however in my reading online it is generally recommended NOT to use a zero crossing SSR when switching an inductive load.