3-phase load reactors

[Buzzen]

I know this isn't exactly PLC related, but the noise my VFD is causing is. I was wondering if anyone had any success using this type of solution: 3-phase load reactors on the load side of the drive to minimize noise on network? Network is Allen Bradley RIO using blue hose. Manufacturer and Model No. is included with PDF>drawing below. Thanks for any help..Drive is a Allen Bradley 1336 plus.

 

[jraef]

Load reactors can help, but they are not a cure-all. What a load reactor does is to slow down the rise time of the PWM pulses going out to the motor. That can mitigate some capacitive coupling issues between the conductors related to the length of the motor leads, which can be a source of common mode noise as well. But if your problem is inductive coupling of AC signals onto the RIO cables, that is not going to help. That is a function of shielding, proper grounding of the shields, and doing all you can to avoid parallel runs of comm cables with power cables.

Read this:

 

[danatomega]

MTE sells Line/Load reactors:
http://www.mtecorp.com/products/rl-line-and-load-reactors/selection-tables/

In my experience, proper wiring (shielded (or armored)) cable helps reduce the effects of the noise created while using VFD's, as well as dressing the wires away from other cabling or crossing at right angles. This is a case of proper installation preventing issues down the road.

 

[James Mcquade]

load reactors are used primarilly on 480 volt systems (line and load) to help reduce the feedback voltage on the drive and plant when the distances exceede a certain range ( depending on amperage).

if you have a lot of noise, then your best bet is shielded cable from the drive to the motor.

regards,
james

 

[Lemming]

How old is the VSD ?, We installed a couple of largish 1336+ drives at pump stations around 15 years ago. At the larger station (150Kw bore pump...), we ended up with the radio spectrum officers chasing us because a chap down the road was having his AM radio blitzed.. We also had a lot of noise on analogue loops etc...Went back to Rockwell and they supplied a line reactor for the supply side, and big ferrite chokes on the 3 x load tails. Turned out the 1336 had little to no input filtering, so all our problems were coming out of the front of the drive and radiating around the LT retic. May be applicable to your install..?

Have since binned both those drives with alternate brands, neither of which caused any issues out of the box...

 

[OkiePC]

+1 for shielded and double drained VFD cable rated for at least 1000 volts. Grounded at motor and at VFD.

For DH+ and RIO, the Blue hose needs to be away from all this stuff and that is the easiest cure for noise on those networks most of the time, replumb your blue hoses, validate your grounds, shields and terminating resistors. Don't be puttin no pinch valves (three strands stripped with a dull knife) on the connectors and do use heat shrink on the bare shields at the Blue hose inlets and outlets.

 

[Bit_Bucket_07]

How old is the VSD ?, We installed a couple of largish 1336+ drives at pump stations around 15 years ago. At the larger station (150Kw bore pump...), we ended up with the radio spectrum officers chasing us because a chap down the road was having his AM radio blitzed.. We also had a lot of noise on analogue loops etc...Went back to Rockwell and they supplied a line reactor for the supply side, and big ferrite chokes on the 3 x load tails. Turned out the 1336 had little to no input filtering, so all our problems were coming out of the front of the drive and radiating around the LT retic. May be applicable to your install..?

Have since binned both those drives with alternate brands, neither of which caused any issues out of the box...

I had that same problem with an old Toshiba drive once. We alleviated the problem with an isolation transformer for the drive, but load side reactors serve the same purpose. I assume that it goes without saying that blue hose is being used here and that the cable is properly shielded, but the cable must also be isolated from the drive's power wiring on both the line and load side. That might require routing the cable through steel conduit -- something that I would have done by default anyway.

 

[OkiePC]

I have seen them used (input reactors) to mitigate harmonics and help slow input rise times for banks of VFDs and one for one matched to any above 10 hp. They never really had any drive problems, occasionally we would battle changes to the Devicenet side of things that seem to take a fault or two to stabilize, but almost never lost a drive. I am a big believer in rigid conduit and shielded VFD cable, you have not lived until you have landed a shield on a VFD cable from a running 10hp motor and seen how much noise you can arc back to its source. I thought I was lighting a tig torch for a second. With that pollution vibrating your shield at 4-20kHz, it will induce garbage onto any adjacent conductors. Output chokes can knock down the noise, but shielded VFD cable catches it like a fishing net at the hatchery.

 

[Tim Ganz]

I am a big believer in rigid conduit and shielded VFD cable, .

Do you do both at the same time? I had a question here http://www.plctalk.net/qanda/showthread.php?t=85984 a while back and never got clarity on how to do it properly

 

[Buzzen]

Wow very interesting input on everyone's end. The boss may have decided to sub it out based on the issues that we are having to an engineering firm. I like the comment on like catching a fish. haha

 

[DickDV]

The shielded motor lead cable is great stuff and will drop the EMI and RFI radiation from the cable to nearly nothing. However, these high frequency pulses are now in your ground system. Unless you have a very low impedance ground network back to the drive's distribution transformer, you may find you just made your problems worse.

If you are unfortunate enough to have a floating delta 480V source or a high resistance grounded wye source, you are in even worse shape. A drive isolation transformer with the output wye center point tied solid to the copper grounds going to the motor lead shields, vfd frame, and motor frame will be your only solution.

Ground noise is extremely difficult to eliminate and shortcuts in my experience just don't work. The transformer is the only solution I can trust.

 

[OkiePC]

Do you do both at the same time? I had a question here http://www.plctalk.net/qanda/showthread.php?t=85984 a while back and never got clarity on how to do it properly

Tim, I have my preferences, not sure what current code would say exactly. I just oversized the pipe and try to go rigid threaded conduit to as near as possible. I have often had to knockout a hole size larger for the connection box on a small motor to make it easier to wire. Many applications were in carriages or widely adjustable positions, and some were just easier to access or install that way, even with SO cord, let alone shielded braided cable, we would go the last six feet just cable to the motor through a CGB.

Then, you have any fault of a shielded cable at that spot, will be phase to phase inside the braided shield, or phase to ground so you are concerned with the ground fault detection and fault tolerance of the drive. We almost always only ran motor cable in the open behind covers or guards or otherwise out of reach during normal operation. Most times for motors that did not move, we would put it in sealtite conduit sized big enough to pull through a 90 degree fitting, but only the last six feet was the rule we tried to stick by. The junction from pipe to the "motor pigtail" would often be a lockable local disconnect.

 

[Buzzen]

New Drive?

Okay here is the latest. Hooked up an oscillascope and found the noise was literally off the charts. The drive is 12+years old. We are thinking of replacing it with this drive:ABB ACS550. It has built in line/load reactors. The question is do you guys think I will be running into the same situation? I am not going to be running blue hose in the new installation(all hard-wired)control circuits from RIO rack in area. By the way thanks for all the replies, I always feel I get the best answers from all of you rather than the internet. By the way not sure if I informed ya all that we have a three-phase open ground Delta feeding this.

 

[DickDV]

Buzzen, check my post just above regarding ungrounded delta power. Also, the ACS550 has an input reactor but no load side reactor. That has to be sized based upon the motor lead length.

 

[Buzzen]

Does that only apply to longer distances? Our motor is right next to control cabinet housing the VFD. The length of cable is only around 25'. We are using THHN that could be the problem too.