3 VFDs Analog Interference

Hello guys,

I need your help. The client has 3 vfds. My colleague noticed that when they ran them simultaneously, the vfds were having analog signal problem and won't run.

They tried to run them individually and had no problem at all.

I just wanted to be clear if the problem is due to noise interference. I haven't check the vfds at site so I am not sure if the cables were properly shielded. Possible that I would go there tomorrow and I want to make sure that I have all the knowledges needed.\

Thank you!

you failed to mention if the signals were input feedback or output control wire to the vfd's. voltage or current outputs.

for the analog input signal i would use twisted pair wire that is shielded with drain wire. ground the wire at the analog input card. the other end needs to have 5 wraps of the drain wire around the outside of the cable with shrink tubing.
for the analog output, i would use 4-20 ma signal, same type of wire.
if it's 0-10 output, still use the same type wire.
you also failed to mention the voltage.
is the incoming power wires to the vfd's shielded?
are the output wires shielded?
are the power wires rated for vfd use?
what is the distance from the vfd's to the motors?
if 480 volt power, are there line & load reactors?

this is what i had to consider
regards,
james

Thanks James for the quick response.

Will get back to this thread once I visited the site.

Thanks,
Marlon

i am no expert, but this is what i used to design panels and controls.
james

I appreciate your answer on this.

BTW, one of my superior mentioned that the analog output of the vfd going to the input of the plc is the problem.

I need to go the actual site to gather more data for this.

you didn't say but my guess would be that the analog signa is 0-10V feeding common to all 3 vfd's
it should be 4-20ma 1 output to 1 vfd no common feeds
there are other options you can try but we would need more information on this system to be much help

Isolation is your friend. With noise, isolation usually beats suppression.

The manufacturer that I use has 16-point common-return analog input cards and 8-point individual isolated cards - same price so the isolated points cost twice as much per point. The key for me is to not tell anybody there is a choice and don't ask - just use the isolated cards.

I'm for proper shielding but it cannot solve isolation problems.

James McQuade,
I am interested in your response about grounding one end of the drain wire and the 5 wraps around the outside of the other end.

Is that a personal preference or is there some specification somewhere that dictates that ?

Just curious. I've never done it that way.

I'm all in on minimizing noise interference.

Thanks

Jim G. said:

James McQuade,
I am interested in your response about grounding one end of the drain wire and the 5 wraps around the outside of the other end.

Is that a personal preference or is there some specification somewhere that dictates that ?

Just curious. I've never done it that way.

I'm all in on minimizing noise interference.

Thanks

Click to expand...

Not speaking for James, but I will wrap the unterminated shield from a multiconductor cable around the other conductors and cover it with tape for the sole reason that I might decide later I need to land it. Wrapping it that way keeps the full length of it available and out of the way.

With shielded cables, I follow the general practice of draining one end to ground, preferably the end nearest the power supply generating the signal. There have been cases where I have had better results by not draining the shield anywhere or by landing it to ground at both ends.

Because of the rare cases where I have needed access to that conductor, I don't cut it short, but preserve it.

OkiePC,

OK, I get that. There have been times where I cut the unlanded end off and later wished I hadn't.

I was more curious about the specific 5 wraps.

But I understand the end result.

Thanks

I do apologize for not getting back to you earlier.

the 5 wraps of the drain wire was an engineering requirement of a well known internet site. every wire is checked.
i'm sorry, but i am not allowed to state the name of the company.
i will state that every facility is built and wired the same, no exceptions.

james

You should also check the grounding of the VFDs themselves - both the main power section and the control card. I once solved a similar problem when I discovered the ground wire between the control card and the chassis wasn't connected.

It's been my experience that shielded cable alone will not adequately block EMI that radiates from nearby VFD motor leads. Especially so if the VFDs are not using shielded motor lead cable. Running the analog wiring in steel conduit is a more effective solution to minimize interference.

I was commissioning a system that had been installed by another contractor once and whenever we sped up a variable speed feeder onto a conveyor belt the belt scale reading would actually decrease rather than increase as the material flow increased. Needless to say, this rendered my feed rate PID control non-functional.

It turned out that they had run the VFD motor leads and the shielded cable from the belt scale in PVC pipe in a concrete duct bank. They wound up having to run separate galvanized steel conduit for the analog signal in order to correct the problem.