The Motion Control Show, Ep 9: EMC Installation Preventing Electric Ghosts

Contact Valin today for more information at (855) 737-4716, or fill out our online form.

The Motion Control Show

Today we're taking a break from all the mechanical topics we've been talking about and let's talk about an electrical one: electrical noise. I've talked to people who don't believe in electrical noise.  They say they're not having problems with it.  But I guarantee you that when you've had the problem, and you've fixed it, and found out it was, you will wish you had known this.  I'll also tell you, you're probably having electrical noise issues and you don't even know it.  But here's a couple of easy steps you can take in order to make sure that you don't…or at least 95% of the time.  I'm Corey foster of Valin Corporation.  Let's talk about this.

Electrical noise is caused by some source and we do know that servo drives, stepper drives, VFDspower supplies, SCRs, oscillators, etc, they are sources of electrical noise because they have the PWM switching frequencies created inside of them.  So that comes from in them.  And this is caused by the making of the square wave used to control the current to the motors.  If you look at this equation, it has a summation of a bunch of odd multiple sine waves and there's the odd harmonics of the square wave and you can see how many there are in there.  This animation shows successfully the sum of the first 200 terms the Fourier series of a square wave.  Let's watch it for a moment because it's kind of mesmerizing, I think.  You can see, as it adds up the next term, it becomes more and more like a square wave.  It still has those sharp corners, but you can see how we go from being a sinusoid into a square wave and that's why we end up with all these different frequencies.  Those different frequencies manifest themselves in different ways.  How so though?  Could be voltage spikes.  Could be noisy and bouncy I/O signals.  This is the most common one that people will see: sensors that are behaving erratically.  They'll get false triggers, false positives, and they think they have a bad sensor or something else. (Also can experience problems in encoders, interlocks, communications, etc.)  But it very easily could be electrical noise and it's pretty easy to solve just by installing the system correctly.  You can install it with band-aids afterwards once you know what it is also, but I prefer to fix it right up front.

The electrical noise is going to exist.  A lot of people think of them as ghosts.  I like to call them electrical ghosts.  But we can manage the electrical noise especially if we understand it and minimize its damage.  The noise of this kind creates a circuit and the most important part about managing the noise is to understand it and manage that path back.

Let's take a look at some examples.  Here is one manufacturer’s hardware installation guide, in the EMC installation guide.  You can see the drive here, and a controller, and they suggest having a mains filter and a power mains filter and here's the motor feedback and motor cable.  If we know that the electrical noise comes out of the drive to the motor and we also recognize that it goes from the motor back into the drive, then maybe we will take the right steps to make sure we manage that path.  You see here that there is a P-clip, or an R-clamp, that is suggested to be installed with this cable.  This cable has the insulation removed so you can get to the braid and then this clamp goes around it.  The electrical noise likes a low impedance path, a path with low resistance, and it likes surface area.  That's why this clamp goes all the way around the cable and it's right here at the drive.  Now it's right here at the drive because that is very close to the power block inside the drive.  There's a ferrite absorber here which can burn up some of the energy, but it doesn't take care of all of the energy by itself.  The real source of the problem is the noise that's going to go through this P-clip back into the drive.  But the problem is that there's no good path between this P-clip and the power block inside.  There's no capacitance there.  It's because if there were a capacitor there, then the drive wouldn't pass the UL hi-pot test.  Since we want it UL certified most of the time, then we can't put capacitors in there.  So, what does getting that noise back to this P-clip do us any good?  Well, the mains filter does have capacitors in here and it does allow for noise to go directly into the power block via these wires so the noise can travel through the case into the backplane into the mains filter and then back into the power block right there.  Now notice here that the backplane has to be uninsulated.  That's a topic that I've seen over and over again.  People get too lazy and they don't take off the paint off the back panel or they try to put just a wire from a lug nut onto the ground of the back panel.  But it goes through the drive, through the panel itself, into the mains filter, through the capacitors, straight into the power block by way of these wires.

There are usually some notes in the EMC installation guides that are critical, but they're fine print and people don't like to read so they overlook them.  Let's take a look at a couple examples.  “Remove all paint and other non-conductive surface coatings from the panel mounting surface.”  Why is that?  Well, again let's go back here.  It's because that gives us the path through the back of the drive to the filter that gets it into the power block.  Okay so let's take a look at another one.  “The only exception is for the motor braid which must return to the drives R-clamp located at the bottom of the drive.  Do not return the motor braid to any other location.  This function is to return high frequency chopping current back to the drive.”  So that's the P-clip or R-clamp that I was talking about at the bottom of the drive.  And here's another one: “route high-power cables, motors and mains, at right angles to low-power cables.  Mount filters close to the drive and keep the supply wiring as short as practical.”  Why is that?  Well, again, that's talking about this wire here, keeping it as short as possible so that that can go in and not be radiated or cross talk to any other wires.  So those are some good examples and you might think that's very particular to this manufacturer.  So, let's take another look at a different manufacturer.

Here's the project planning guide and this is showing filters and a whole bunch of other stuff in the drives.  This is a power supply that can power multiple drives.  Here's the motor.  But what?  There's no P-clip, no R-clamp!  Therefore, this must be a better drive manufacturer, right?  They must have better product.  Well when you look a little bit closer into their product planning guide, this shows they do have an accessory that you buy separately that you “connect the cable shield with the lowest possible impedance to the drive-controller.”  So, they have the very same idea: you want to connect that shielding as close as possible to the drive.  Okay, there's some fuses for safety, the transformer as the power supply, it's a step up or step down or maybe an isolation transformer.  We got some mains filters.  Let's see, there's noise in and out of it.  24-volt power supply, it's wired here on the dirty side of the mains filter which assumes it has its own filtering.  And then there's a mains choke which adds inductance and softens the edges of the frequencies.  So, the same path is available.  We have the electrical noise going to the motor and back, out to the filter, and back into the drive.  Now the electrical diagram of it here doesn't show the motor but notice the grounding.  It's dotted.  Why is that?  Well, there's a note in the manual that says “connect the bare metal back panel of the device in conductive form to the mounting surface in the control cabinet.”  That's a non-insulated backplane right?  Again, the fine print here and the concepts are very very important.  They're simple concepts, but most people overlook them.  There have been many times where I've walked in and solved electrical noise issues just by attaching that R-clamp or P-clamp of a motor cable to the drive.  Next step is to tell them to take the paint off the panel behind the drive so that this path can happen.

It brings up a question.  If you have multiple drives how do you deal with multiple drives?  Well, there's a couple possibilities.  One filter per drive could get a little expensive.  Another one is one filter for all of them wired in a star pattern.  You may end up running dirty input power near the clean outputs which cross contaminates it, so you have to be very careful about that.  Another way, I've never actually seen this done, but here's a good way in theory, is wiring it in a star pattern, but putting the filter behind the panel and running the clean power through a hole, so the panel acts as a shield and we keep the dirty power outside.

Some dangers to be aware of when you're making your installations is running the high-power next to the low-power, a very typical one I showed that to you in one of the installation guides earlier, keep a significant gap between them or run them perpendicularly.  Contaminating good clean power with dirty, common problem. 

Again, I've solved a lot of electrical noise issues just by doing a couple of these very simple things.  So, I hope that you do them before you have electrical problems!  I'm Corey Foster of Valin Corporation. I hope that helps! 

If you have any questions and are wanting to know more, reach out to us at (855) 737-4716, or fill out our online form.