Electromagnetic Interference problem

[Cyanogen281]

I am performing vibration analysis on a motor that is around 50 feet away from the control panel.

The output of the data translation module I am using has a USB output, therefore I had to install A USB over CAT5 converter to run it out to the control panel (since the limitation on USB is 16.4 feet)

The only problem is that the contractor ran the cat 5 converter box right under a soft starter in the control room and now my program freezes every time I run the motor while connected inside the control room.

I need to leave a PC inside the control room for a few months and be able to remote into it for recording purposes.

I believe that the soft starter is causing Electromagnetic Interference with the USB to CAT5 converter and in turn causing my analysis software on my computer to freeze.

Would a Faraday cage solve this problem?

Are there any kind of EMI shielded USB cables out there? Or am I right about the USB to CAT5 Converter's circuitry receiving a massive spike from the soft starter?

 

[IRstuff]

That would depend on your software, but it's certainly doable in the general sense:

http://www.usb-over-ethernet.com/,

but you probably need a software driver as marketed by that site, since normal Ethernet drivers only understand TCP/IP, which would not be what's coming out of your ethernet cable.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[DRWeig]

If I have this right in my head:

You have a vibration analyzer on the motor, and its output is USB. Your PC is in the control room. So out by the motor is a USB-to-ethernet converter and in the control room is an ethernet-to-USB converter for plugging into the pc?

Why not try a long USB extension and keep the active (repeater) portion outside of the room?

http://www.newegg.com/Product/Product.aspx?Item=N82E16812228657

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[VEBill]

I agree with Gunnar when he said that ground loops are a common (mode, LOL) problem; but you've probably eliminated that possibility with your hand held laptop experiment.

How about just getting some very large (2-inch wide) braid and shielding the entire cable including the dongle?


Beware one problem doesn't exclude the other. Be prepared to repeat the hand held laptop experiment in case it's both ground loop and EMI.

 

[djs]

There is such a thing as shielded Cat 5/6 cable. I don't think that going from Cat 5 to Cat 6 cable will make any difference. Instead try going to a shielded type of cable. You will also need special shielded RJ-45 connectors. I still don't know why you can change the location of the converter within the control room. Mount it as far as possible from the soft-starter and as close to your computer as possible.

 

[itsmoked]

How about unbolting the converter from the wall and temporarily setting somewhere else like on the floor, just as a troubleshooting method.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[zappedagain]

Your USB to Ethernet adapter is going to be a microprocessor/microcontroller with a USB port and an Ethernet port. The firmware initializes both ports and then passes the data through. You can easily do this in your controller if it has the proper ports.

Z

 

[hacksaw]

you've stated nothing regarding how your

1."data translation" hardware is grounded or isolated from ground
2. how the cable is routed (tray, conduit), or it's degree of isolation from adacent power wiring.

sounds like a ground loop issue, there is test equipment for isolating gound loops, have youused such hardware?

 

[Cyanogen281]

Our data translation hardware is grounded next to the motor that we are measuring the vibration from.

When we move the computer (Laptop for this application) away from the motor and begin the vibration analysis everything works as it should.

The CAT5 cable that is run into the control room is, unfortunately, ran in the same conduit as the power wiring.

We have isolated the problem to the USB to Ethernet converter and it only seems to freeze the program when the USB/CAT5 converter is inside the control room.

Is a Faraday cage completely out of the question at this point? Wouldn't that shield the USB/CAT5 converter? Perhaps some copper tape wrapped around the device, then grounding the tape?

 

[MikeHalloran]

The CAT5 cable that is run into the control room is, unfortunately, ran in the same conduit as the power wiring.

You've now lost whatever money _that_ saved, just in your time.

A separate conduit for the signal cable basically amounts to a Faraday cage, without the expense of trying to pull a foil-wrapped signal cable in with the unshielded one.



Mike Halloran
Pembroke Pines, FL, USA

 

[ScottyUK]

An optical fibre doesn't know or care about EMI, regardless of how bad it gets. Run your USB data over fibre.

 

[DRWeig]

You can get a good sheet metal guy to Faraday-ize your USB/Ethernet converter, or just wrap the heck out of it (and also the cables next to it) with aluminum foil and see if it really works.

I like ScottyUK's suggestion the best.

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[Cyanogen281]

An Optical Fiber cable converter is around $800 for the entire setup for my particular application.

Being as the (CAT5) data cable has already been run in the conduit with the power wires, I am thinking of simply replacing the UTP (Unshielded Twisted Pair) CAT5 cable with a shielded version of it and hopefully that will fix the problem. If not I don't see the harm in simply wrapping the USB/CAT5 Converter in something like aluminum foil essentially creating a Faraday cage for testing purposes.

Any other advice or suggestions would be greatly appreciated.

Thanks for all the help guys.

 

[MikeHalloran]

Foil, whether aluminum or copper, is pretty much a waste of time, Mu-Metal slightly less so.

A simple steel box will be superior to any of them, every time.



Mike Halloran
Pembroke Pines, FL, USA

 

[ScottyUK]

That does depend whether your problem is capacitive (E-field) or magnetic (B-field) coupling. Any earthed metal will knock capacitive coupling right down. Magnetic coupling is best dealt with by increasing separation, followed by mu-metal and the like.

 

[Skogsgurra]

Are you still discussing this? What is this - an academic exercise?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[mikekilroy]

The CAT5 cable that is run into the control room is, unfortunately, ran in the same conduit as the power wiring.

move it

The CAT5 cable is UTP (Unshielded Twisted Pair)

go spend $2.60 and buy shielded cat5 cable

http://www.KilroyWasHere.com

 

[MikeHalloran]

Another reason to move the signal cable to a separate conduit, on a planet where the NEC applies, is that the Cat5 cable probably isn't rated for voltages as high as the motor voltage.



Mike Halloran
Pembroke Pines, FL, USA