Lathe VFD 3: How to Fix Noise Problems

[Music] welcome back to cloud for tea - I'm James we are in the middle of a variable-speed upgrade for my Geo 602 lathe last time we got the motor and the VFD set up on the bench everything wired got the VFD programmed and everything is running beautifully but there's one problem that you couldn't see in the video and that's EMI electromagnetic interference now in many home shops this might not be a problem but for me it is the I have a computer running my CNC mill and the EMI can interfere with that but more importantly I shoot video here in my shop obviously and the EMI being emitted by the variable frequency drive is causing problems with the video gear let's go over to the bench take a look at the problem and then we'll work through how to solve it this is the same setup we had last time this is the tico Westinghouse VFD hooked up to a 1 horsepower three-phase motor and this is just where we left it last time wired up programmed and ready to run now I pulled a couple of extra things into the shot today I've got an oscilloscope that I have wired up to some transformer current sensing coils and we will look at those in a minute but I've also brought in my production monitor now I normally shoot with three cameras here in the shop just so I can have a variety of shots and so I can work around my big hands and my big head that keep getting in front of the camera and I stream the video from all three of those cameras to this monitor so I can keep an eye on what's happening as I'm working and I can get an idea of what's being captured by the cameras and I can again move my big hands in my big head out of the way now watch what happens to the monitor when I turn on the VFD I just got this set around 20 Hertz you can see the monitor goes completely blank now depending exactly where the monitor is in the shop and which circuit it's plugged into sometimes it just gets a jittery and I lose horizontal sync but sometimes like right here it just shuts off completely now what's going on here what's happening is the VFD is generating electrical noise and the reason this is happening is because the VFD has some big transistors in it IG be T's insulated gate bipolar transistors and they are switching the power to the motor coils on and off at 16,000 times per second this is the 16 kilohertz pulse width modulation and they do this to be able to control the amount of power going to each coil now the igbts have to switch very very rapidly from off to on and on and off because any time that they spend in the kind of partially conducting mode they're dissipating heat and so to limit the heat that's being dissipated from the VFD you have to switch them very rapidly but when you switch rapidly like that you generate very sharp edges that have a lot of high frequency noise content and that high frequency noise escapes from the VFD into nearby electronics there's basically two ways this happens first is radiated noise and that's primarily what's happening here on the motor lead because the motor lead doesn't go anywhere except to the motor it can't really carry conducted noise to any other attached equipment because there isn't any attached equipment but it does set up noise currents in the lead that then radiate electromagnetic waves out into space and those can interfere with nearby equipment but probably the more important source of noise here is conducted noise and that's noise currents to travel back up the line cord into the power supply into in this case a plug strip that I have over here on the left and anything that's plugged into it including this monitor the computer on the CNC machine and any other equipment that I have here in the shop or in the house before we start trying to fix the problem let's see if we can measure it see if I can illustrate what's going on here so what I've done is both on the motor lead and on the line cord I have formed a current transformer now this is just a coil of wire in this case it's just a test lead that's wrapped around the line cord eight times same over here with the motor lead and why eight times because that's the length of the test lead wire over here on the motor lead that came out perfectly wrapping at eight times and I wrapped it the same number of times on the line cord now the current flowing through the line cord here produces an electromagnetic field and this coil of wire wrapped around it forms a transformer that induces a current in that coil from the magnetic field that's in this cable now because the hot and neutral lines are both inside the cable we're only measuring the net magnetic field and the net current the power that's actually going to the motor the sixty Hertz AC power is traveling in one direction on the hot line and at that same moment traveling in the other direction on the neutral line and that cancels out so we shouldn't see that current in the current transformer we should only see the noise current or the common mode current that's being transmitted back up this cable now we can't measure a current directly we need to measure a voltage with the oscilloscope so I have the coil connected to a resistor this is a 1k resistor in this configuration it's commonly referred to as a burden resistor and 1k is probably way too high we're gonna see more than just current we're also going to see some electrostatic coupling in addition to the electromagnetic coupling but that's okay again I'm not really using the right equipment I should be using a spectrum analyzer and all I have is low end oscilloscope I've probably got you know - a few turns on the coil I probably got too large of a burden resistor but we're gonna roll with it because of just trying to illustrate and get some relative measurements as we try to do something about the noise let's give it a shot and see what happens so I've got my oscilloscope connected across the resistor so that we can measure the voltage across the resistor caused by the current that siddhu's induced in this makeshift trance former coil and I've got exactly the same thing over here on the on the motor lead okay now looking over here at the oscilloscope you can see I've got two traces on the screen this is for the two probes the yellow trace is the line cord transformer and the blue trace is the motor lead transformer and this is set to 200 millivolts per division on both of these and you can see that you know aside from just the sampling noise that causes these little flickers every now and then there's basically nothing going on here now the power is on to the VFD you can see the lights are flashing it is drawing some small amount of current but you don't see anything reflected here and the reason is because those currents are cancelling out and the coil isn't picking anything up now let's take a look at what happens when I turn on the motor now you can see all kinds of noise and junk being picked up by the the yellow trace again is on the on the power lead the line cord and the blue is on the motor lead and if I kind of try to zero in on just sort of one of these pulses and I can mess with my trigger so that I can catch kind of only the highest ones I've got the scope measuring the peak-to-peak voltage and so on the being noise it's all over the place but on the line cord I'm seeing peak voltages as high as say about 1.2 volts and on the motor lead I'm seeing it up to as high as 860 so let's write that down and on the motor [Music] okay so this is this is our starting point we're getting these spikes and again what is 1.2 volts mean it doesn't really mean anything because this thing is completely uncalibrated I just wrap some wire around this grabbed a resistor out of my junk drawer hooked it up and I'm getting 1.2 volts across this configuration what's the actual voltage or the actual current noise in the wire what's the magnitude of that I'm not even gonna try to figure that out if you know put it down in the comments but we're just going to use this as a starting point so as we start to try to do something about this we can see the effect that we're having I think our biggest problem here is probably the conducted noise as it usually is traveling back up the line and interfering with the monitor so let's start by trying to deal with that the easiest way to deal with conductive noise is just to use a filter and this is a filter that I bought I got this one on Amazon there'd be a link down below if you're interested and this is a common line noise filter so it's got inputs for a line in neutral I just scribbled on here there there actually it's balanced so it doesn't matter which side you put the line in neutral on but I went ahead and labeled them how I'm gonna hook it up just so that I don't get confused later now what's actually in this box are a couple of common mode transformers are common mode chokes that the line and the neutral both go through and what's happening here is when you create when you have an inductor it has higher impedance for higher frequencies so what's going on is we want to pass the 60 Hertz power through this essentially unimpeded but the higher frequency noise content we'd like to block and so the way this works is it runs through these two series inductors and the inductors look like a higher impedance which is exactly what it sounds like like a higher resistance or it weakens the strength weakens the current of that noise energy that's going through and then it's hard to see on here I'll pop up a circuit diagram on the screen there are then capacitors both between the line and the new trol and between the line and neutral to ground and a capacitor looks like a low impedance to high frequencies and so what happens is in layman's terms as the energy as the power goes through the coils the coils weaken the high frequencies and then the capacitors drain that high frequency energy off to the ground and this one's actually two-stage you can get these with a single coil this has got two coils in series so you just hook this up got to have a ground connected and hook the line in neutral to one side and your load which in this case is the VFD to the other side so let me unplug the VFD hook this up and let's see what that does to the noise being conducted back up the line I gotta wait for the capacitors in here to drain because I don't want a nasty surprise and they're dead okay that's the the line neutral and ground of the filter and the line and the neutral sorry I didn't have a white wire connected up to the input of the VFD last thing is I have to ground the VFD now when I ultimately put this in a box these wires are the right links so I went ahead and made them up the ground wires gonna have to go off to a terminal in the box somewhere so for now I'm just gonna use a test lead now this is going to be probably higher higher impedance than I would really like so probably reduce the effectiveness but I think it's still going to be fine okay we double check the connections here okay I think we're good let me plug this back in and there was no loud bang and blue smoke so I think we're good so I guess let's just try this again let me flip it on and let's watch the monitor and lo and behold nothing happens let's take a look at the oscilloscope and you can see we're not triggering anymore and that's because the peaks are not high enough to hit where I had the trigger set let me bring that down and let's take a look at the values it looks like the peak to peak voltage I'm watching this on the line cord and I'm seeing I think about 760 s the highest number I've seen and over here on the on the motor lead I'm seeing I'm seeing maybe 900 950 so you can see we've cut almost in half the conducted noise that's going back up the line to the to the the monitor now this monitor is fairly sensitive to a lot of things I've had problems in the shop where I touch my cameras and it goes off for a second but nothing like the just complete loss of picture that I was getting from the VFD so what's going on with these numbers it looks like the line filter in terms of conducted noise back up the line again even dealing with the low efficiency of this ground test lead I have we've just about cut the amplitude of those noise currents in half now on the motor side we haven't really done much in fact it may have gone up a little bit this could just be noise or random variation or it could be because of the capacitors in the input filter we've actually stiffened up the power supply that's not really a technical term and we're actually getting sharper current spikes to the mode so let's take a look at what we can do on this side now again just this filter has actually solved my problem with the monitor the monitor stays on now I'm good with that but let's take a look at what we can do on the motor lead as well if we're the motor lead there are there are filters you can buy three-phase filters line reactors all kinds of very expensive gear that you can you can buy to filter noise but again we're primarily not concerned with conducted noise on this line we're primarily concerned with radiated noise and that radiated energy is going to be at relatively high frequencies perhaps above 35 megahertz which is where the FCC classifies EMI and in order to deal with that all we really have to do is round off the square edges of the waveforms reducing the high frequency energy and there's a couple of ways to do that but the easiest way is with steroids this is a ferrite this is just a composite material made with powdered iron and some binders and then covered in a ceramic coating and you these are sold primarily these are transformer fair rights they're sold for actually winding wire around them to make small transformers but if you put these around the conductors on the lead they will actually absorb or reflect high frequency energy that's going down the cable and have the result of squaring off and taking the the high frequency edges and hence the radiated noise off of the signal that's going into the cable yeah I have a whole bunch of these and we're going to go ahead and put them on this motor lead we're gonna put some of them on around all of the wires together and some of them on the individual wires we're not going to bother trying to wrap them we're just going to run them straight through let me turn off the power here and we do the same thing we're gonna wait for the capacitors to discharge because I don't like unpleasant surprises okay good all three these out how many of these am I gonna put on well I bought a package of tin for about $8 so we just throw them all on here are they all necessary no probably not but they're cheap and regardless of whether there's a real issue here or not this is good practice and they will help control radiated energy so I put two coils around all three wires two coils on each wire and because I had a ten pack I got two left so I'll just go ahead and throw those around all three this is all gonna be sitting in a Hammond box it won't matter at all and I can use you use zip ties and try to organize these and make them look nice but there's my little ungodly pile of fair rights on there let me hook the wires back up okay tan plug it back in okay we got everything back up and running again with the fair rights on here let's power up the motor and see if we can see any difference on the oscilloscope and in fact as you can see we're not even triggering wow this has worked a lot better than I expected let me bring this all the way down and let's take a look at what we've got so if we take a look at the highest peaks that peak to peak on the line cord is down I think I saw 576 there call that and on the motor we're down to it was 544 568 I think 576 let's call it 576 so you can see with those fair rights we have cut at least a third maybe a little more than a third of the high frequency noise out of the motor lead and that has actually further then dropped the noise that's being transmitted back through the filter up the line cord and the reason it's it's affected the line is because we've taken the sharp edges off of this so that that power that went that had those sharp transitions and those sharp square edges going to the motor was being drawn from the source you know through a bunch of filtering that's in the VFD but it was being drawn from the source so by slowing down those edges we've slowed down the rise in the current draw back to the line and so all in all we've cut about more than half of the noise out of the line cord and we've cut at least a third of the radiated of the noise that we're measuring in the motor lead and we've probably cut quite a bit more than that from the radiated noise that I don't really have the the right equipment to measure that's all for today I had to walk through the process of troubleshooting this noise problem and coming up with a solution and I thought it might be useful to some other people so I decided to show it here on the channel the next thing we need to do is stuff all the electronics into a metal Hammond enclosure and then stuff it all into the lathe and we will work on that next time if you enjoyed this video found it useful please give me a thumbs up also feel free to subscribe to the channel and leave me a comment I'd like to know what you think thank you for watching [Music]

Channel: Clough42

Views: 87,483

Rating: 4.9699354 out of 5

Keywords: VFD, 3-Phase Motor, EMI, EMC, VFD Noise, Noise, Filter, Power Filter, AC Line Filter, Ferrites, Oscilloscope

Id: DJfiOqaeFDg

Length: 21min 35sec (1295 seconds)

Published: Thu Jun 14 2018