Minor Arcing

[ProjectDelta]

Friday at work I was working on a large VFD swap out. Before powering it down, I measured the distance between the line feed pipe (1 1/2” rigid) and the load side conduit (also 1 1/2” rigid) to match the existing knockouts on the unit which both pipes penetrate the bottom. When I put my tape measure across the two conduits, there was a very tiny arc occurring every time I touched the conduits with the tape measure. Anyone have any theories on why this is? They ground the hell out of everything where I’m working so this is baffling to me.

 

[paulengr]

Friday at work I was working on a large VFD swap out. Before powering it down, I measured the distance between the line feed pipe (1 1/2” rigid) and the load side conduit (also 1 1/2” rigid) to match the existing knockouts on the unit which both pipes penetrate the bottom. When I put my tape measure across the two conduits, there was a very tiny arc occurring every time I touched the conduits with the tape measure. Anyone have any theories on why this is? They ground the hell out of everything where I’m working so this is baffling to me.

A VFD requires good stranded supplemental ground from the motor frame to the VFD ground. All VFDs generate significant common mode current which flows to the stator coils and is returned via the motor grounding. VFD cable foil shields aren’t enough. Steel conduit has issues with high frequencies. If this cable is missing or not landed properly such as just tying it to a grounding bushing which is pretty common or the steel has a lower impedance you get a current on the conduit and most likely you were creating a second path which reduced impedance and allowed current to flow. I have measured common mode currents on a 900 HP VFD as high as 90 A peak. These peaks are where it induced currents through the bearings and since it exceeded the considerable bearing maximum current of 10 A, it was arcing through the bearing causing fluting and bearing failures every 6 weeks. This is not your typical circulating current and the usual methods (brushes and insulating bearings) did very little for it.

Don’t forget too that VFDs generate considerable RFI that radiates onto nearby metal objects and causes current flows. Grounding keeps the voltage at or near zero but you have currents flowing on neutral wires that measure zero volts.

Also on larger systems a common installation error with multiple conductors per phase is to install them say AA, BB, CC, in separate conduits instead of ABC, ABC. This causes inductive currents to flow between conduits. In steel conduits terminated in a steel enclosure it can actually melt the enclosure. Need a ground cable in every conduit.

So a statement like “good grounding” isn’t necessarily true unless you follow the VFD manual to the letter. Grounding back to the transformer isn’t nearly as critical as the connection between the VFD and the motor. And don’t think you can cheat by using VFD cable grounds, foul shields, or conduit bonding. And don’t park that supplemental ground on the backplane, grounding bushing, or anywhere else except the designated location on the drive itself. This is THE source of many, many VFD issues if not a major contributor.

 

[ProjectDelta]

Wow, thank you for taking the time to post this excellent reply. I think I’ve learned more about common mode noise from this than I have in the hours I’ve spent trying to understand the concept.

 

[ptonsparky]

Talking to another EC about that designated ground on the VFDs and he asked why weren't those on drives in MCCs. I

 

[LarryFine]

Anyone have any theories on why this is?

I would expect the two conduits to be solidly bonded.