Noise Problem - NI cRIO, Control Techniques Unidrive SP

ParetosFew

Member
Join Date
Feb 2017
Location
Pennsylvania
Posts
5
System:

  • Drive: Control Techniques - Unidrive SP (Panel Mount)
  • Controller: NI cRIO
  • Power: Whole system runs off of a 3 phase 480V plug. There is a transformer for 120V that powers a few accessories and a 24V supply. VFD is run directly of the 480.
  • May be relevant, entire system is somewhat portable. Panel is on a steel stand that sits next to the machine. All grounding must be through ground prong on 480V plug.

Problem: Severe "bouncing" of several different input signals (mV/V and 4-20mA) whenever the drive is active. Only when the drive is told to run, even if at 0 speed. As soon as drive is stopped (but still powered) signals return to acceptable levels.

Things I've tried:


  • Sent signal with process meter - worked fine.
  • Signals cables are shielded only on panel side.
  • Ran signal cable(4-20mA) completely outside of panel and trays, no change. Moved 24V supply out of panel, laid on floor temporarily, no change. These make me think this is not radiated EMI.
  • Because different types of signals are effected, I suspected noise at power supply. Measured huge (>10V) periodic spikes between +24V and 24VCOM.
  • No "direct" connection between VFD and 24V system. Only connection is 24V sourced from drive through a Guard Master safety relay contact back to the drive, and chassis ground.
  • Moved power supply away from VFD, isolated power supply, tried another power supply all with no change. Seems like power supply is OK.
  • Replaced drive enable through safety relay with a jumper (temporarily) so that only thing shared between the drive and the 24V system is chassis ground. Problem persisted.
  • This made be suspect noise on chassis ground. Measured between chassis ground and earth ground from a nearby 120V outlet. Spikes still visible, I think confirming that the noise is present on the entire chassis. I definitely think this severe noise on ground would interfere with the cRIO.
Since the only thing I did not yet isolate between the VFD and 24V system is chassis ground, my next idea is to try taking the panel mount drive completely out of the panel however this may be a bit of a project since it is mounted in a very tight space. Tried removing the chassis ground connection, however I assume it is still grounded through the panel. If this does eliminate the noise, I will need to determine if the drive is faulty or configured properly (I do have a call out to control techniques, have not heard back yet). If it doesn't, then I am still stumped.


Another thing I may try soon is configuring the drive to not source the 24V for the enable signal. Maybe something in the transformer and power supply internal to the drive is the cause? I've attached photos of the scope measurements although I'm not real confident on them. I think I should have been DC coupling, which I did since and is where I measured the much higher spike values (+/- 2.5 - 3 divs at 5V/div). Unfortunately don't have pictures of latest measurement.


Any Ideas? Also, glad to finally make a post here, I've had many problems solved through google searches that lead to threads here over the past few years in this field. Thanks in advance!

IMG_0647.jpg IMG_0648.jpg IMG_0649.jpg
 
Did you use VFD cable from drive to motor or just flexible cord? I had this problem once where my light curtain kept tripping falsely when my drive was running till I installed a VFD cable
 
It is just flexible cord for each of the three phases and a heavy ground wire. Now that you say that, I recall reading that advice on other issues related to VFD noise. The three phases and ground go through a dedicated piece of seal tight that is grounded, however there is maybe a foot between the VFD connection and where it enters the seal tight.
 
I had a similar issue with a 25HP CT drive. The HMI several feet away would go nuts. We switched to shielded cable from the drive to the motor and it fixed the issue.
 
Is your 480 supply floating delta? or grounded wye?

I hate to admit this but despite reading up on this and vaguely recalling these configurations from school, I'm still not sure how to tell which I have. From what I read I can't tell by the plug alone? It is a 4 pin plug, 3 phases and ground if it helps. I can talk to the electricians that did that service as well.

I had a similar issue with a 25HP CT drive. The HMI several feet away would go nuts. We switched to shielded cable from the drive to the motor and it fixed the issue.

Seeing this recommendation a lot on other VFD noise issue threads. I guess it does make sense that a heavily shielded cable would do much better than just a ground parallel to the 3 phases. I will likely add this to the list of things to change on this machine even if the immediate problem is resolved some other way. Thank you both.
 
I've minimized the motor-lead radiated noise enough a few times by braiding the motor leads. Otherwise, ensure the motor ground is solid and from there, the only molution is VFD cable (shielded).
 
I was able to resolve the problem (mostly) by overnighting a proper VFD cable from control techniques. Also when changing this over I found something odd: the ground that I thought traveled parallel to the motor conductors actually traveled around the machine and connected to some fixed speed pumps and then back to the panel in a different area not through the drive. Obviously I connected the ground inside the VFD cable directly to the drive. The noise on my 4-20mA signals went away, there is still some bounce all the time but it is indistinguishable whether or not the drive is on or not by watching the numbers. I expect some variation in these due to the large pressure range represented.

The mV/V load cells still show some noise (greatly reduced) but I think this is to be expected with a low voltage signal. There is also some suspect cable routing I can change to improve them.

The cable from control techniques is foil braid shielded, and has the three conductors and a ground twisted together. I may even look into high performance belden VFD cable that has "more copper" to help prevent noise leakage, some of them have three uninsulated ground wires that spiral around with the conductors and insulated ground. Thanks to all for their responses and advice.
 
Last edited:
The shield part of the VFD cable is just a "nice to have". The important thing is that the ground runs directly from the motor frame to the VFD frame within or directly alongside the drive cable. Everything else is secondary. There are other "non noise" considerations which might dictate the optimum method of running the ground cable so I would recommend following the drive manufacturers instructions but in more than 99% of VFD related noise issues I have had to deal with the culprit has been having a separate route for the ground and drive cables or terminating the ground cable on the panel rather than the drive. With the high carrier frequencies used by modern drives the capacitance of the cables and the motor allows significant leakage currents to flow and these must be cancelled out within the cable installation to avoid noise issues.
 

Similar Topics

I have a panel that was retrofitted with a Zelio PLC and a 100HP Parker 590 DC drive. The panel has been having trouble for months with the...
Replies
2
Views
1,707
Hello- i have 200Hp compressor that is operated with a FW Murphy controller. As the unit ramps to speed, the input lights start to flicker and the...
Replies
4
Views
1,474
I have a systwem that I set up that uses 2 paddle wheels npn devices and an encoder card. I have a digital meter hooked up that correctly counts...
Replies
22
Views
8,482
So I had a strange problem today that defied what I've learned about data communication. basic setup is: Scanner -> PLC via RS-232 using only...
Replies
7
Views
6,467
Hello, I am running a AMCI SD4840 Stepper Motor drive system with Allen Bradley CompactLogix. Every now and then when a 480VAC motor kicks in on...
Replies
1
Views
571
Back
Top Bottom