EMC Testing

[fifty60]

I have a device that is failing conducted emissions. At this time, an input filter is the remedy. This equipment consists of motors, fridge compressors, heaters, solenoids, electrical-mechanical relays, SSR's, and PID controllers, 24VDC power supply.

The heaters are time proportioned over a 2 second cycle from the PID controllers and SSR, the compressors are time proportioned over a 6 second cycle from the PID using a E-M relay. Everything inside the equipment is powered from the same 115 Volt 20A wall outlet.

There really isn't any way to isolate my circuits inside the equipment to reduce noise is there since they are all powered from the same circuit. For example, I could try to isolate all my heater wiring and heater control wiring from all the other conductors, but the noise from the heaters circuits will still be put out on to the supply right?

How about separating input power to electronic components (like my 24VDC power supply, PID controllers controllers, SSR's)? Will this have any benefit for me? I've heard the phrase "Separate inputs from outputs" but does this mean separate the control circuit from the power circuits via a shielded transformer?

Would the conducted emissions be primarily from the power supplies in my components (like the PID controller), or do the cycling of the SSR's have a mojor contribution?

 

[GoldDigger]

Do you have snubbers of some sort on the EM relay contacts? Opening an inductive circuit like a motor can cause high transient voltages at high frequencies which will be hard to filter just because of the magnitude.
I am not sure I understand your "time proportioning". Switching a compressor off and on at six second intervals is a great way to burn out the motor.

 

[fifty60]

GoldDigger, it is not the compressor that is cycling, but the relays control the refrigerant flow to the compressors.

I am failing at 438kHz and 514.5Khz. This could only be a power supply right? What can I do with wiring to reduce this? I don't think there is anything that will substantially reduce this.

However, when I take the components out of my equipment, and hook them together, they pass. This would be the 24V power supply, the HMI screen, the pid controllers. when I place them in the equipment (with all the other circuits) the equipment fails. The equipment is small, so routing the wires is a problem.

Would trying to isolate all the power supply, hmi, PID power input wiring from everything else make a difference?

 

[GoldDigger]

Is the emission constant or is the test equipment capturing peak EMI values?
You might be able to find the source of the RF using a scope probe. Once you have identified it you can look at how it is getting to the external wires.
I suppose that it is possible that something is oscillating when packed together but not when spread out. Or you might be getting magnetic coupling from a poorly shielded transformer or inductor.
Look for loops of wire.

 

[fifty60]

It is the constant average peak values that are failing. The conducted emissions would come from the input of my power supplies right? For example, I have a 110V to 24VDC power supply. The 110V input is the side that is generating the conducted emissions?

I know it is the process of switching that 110V to 24VDC that is creating the noise, but do the conducted emissions come out the input and output? Does noise on the output affect the conducted emissions on the input?

 

[fifty60]

I currently use 14AWG MTW wiring. Would there be any benefit in switching to a 14/2 twisted and shielded cable for the inputs of my electronic components, for example, my PID controllers, HMI, 110VAC/24VDC power supply?

Or would this only eliminate radiated and emmisions, and improve common mode radiated immunity? It is all connected to some degree, right?

Conducted is what I am failing, but my electronics outside of the equipment enclosure pass. Inside the enclosure they fail.

 

[GoldDigger]

Possibly there is a resonance condition inside the enclosure. But more likely, IMHO, is that some of the interconnect wiring is being placed in a position where it is picking up non- conducted RF from some part of the equipment.
Have you tried packing the components together in the same orientation, just without the conductive enclosure?

Twisting wires will not affect conducted RF, but it could affect whether the wires pick up radiated energy and conduct it outside.