Okay, where are we. We talked about “noise”, or what I prefer to call “electrical pollution” in the last chapter. I gave you the gospel according to Uncle Phil as far as what I think of companies who pollute the electrical environment, and finally, I showed you some electrical pollution on a couple of oscilloscope displays. Now we are up to the part that you were hoping for two months ago. This installment will discuss actual filters, what they are, what they do and how they differ from other types of filters.
But before we get to that, I want to welcome all of you who have fallen into our little web trap here at HTI. In a creative bit of futuristic programming, we have now taken control of your computer. But don’t worry, we are doing this for your own good. First we will look over your hard drive and…yes, it looks like you are interested in Home Automation (we can tell by looking at the other HA programs you have your drive, plus we can see that you have visited some other HA web sitesâ€¦and, uh oh, you go to THOSE kind of sites too?!?! Shame on you).
Anyway, it looks like you have sufficient interest and qualifications to be allowed to continue. As a matter of fact, you have even qualified as an “advanced student” in HA and so we will allow you access to all the other articles here at HTI. Some of the better ones are the previous 9 articles that precede this one (but I am somewhat biased, I’ll admit). Just click here to go a listing of subjects covered in parts 1 through 9.
This article is officially titled “Which One Should I Use, Part X” but don’t be like the guy who refused to see the movie “Malcolm X” because he didn’t see the 9 movies that came before it. In this case, while there are 9 other chapters that came before this one, you don’t have to start with #1. I would like it if you did but it’s not absolutely necessary. You should know, however, that this one is the second in a sub-series on “Noise and Filtering”. If you are interested in this subject, I strongly suggest that you go back and read Part IX first. That’s okay, we’ll wait on you….
Oh, you’re back. Are you ready to continue? Good! Now we are up to WOSIU#10, which will be forever archived under its official title:
Which One Should I Use, Part X
(Noise and Filtering, So Let’s Look at Filters.)
As I began writing part IX, I knew it was going to be a 3-parter. In the first part, we discussed electrical pollution in general. In part IX, you were also part of my experiment to create a little movie that shows something that is very difficult to explain with just words. This time we get to dive head-long into our first line of defense: Filters! Plus this one will have three short mpeg files.
A word about these mpeg files before we go too far. I wanted to offer them because they show you in 20 seconds what it would be nearly impossible to show otherwise. Please bear in mind, however, that I am doing this with limited equipment. To those of you who have attended one of my classes, you know that there is no substitute to seeing it with your own eyes. Having said that, these little mpeg files are better than a still picture. Remember, if you have any problems, please let me know. The first one (HTI June 98) was pretty well received but some people did have problems. My new computer at home (P2, Win95-C, 56K modem) was able to download and display the mpeg file just fine. My older slower office computer (Pentium, Win95-A) didn’t fare as well. After downloading the file, it refused to play. I had to rename it to get it to work. It didn’t matter what I named it, as long as it was different than the name it had when I downloaded it. Why? Only God and Bill Gates know. Anyway, this time we will have three files for you. They are not intended to be dazzling cinema, instead they just show you how filters work. (Whatdaya want, I ain’t Spielberg, ya knowâ€¦)(Editors Note: When you click on an mpeg it will download in a separate window. That way, you can go back to the original window and continue to read the article while the file downloads … then go over and watch it when it’s ready. By the way, don’t forget you need a browser plugin to watch mpeg videos.)
Alright, on with the show.
Filters! Just the word brings to mind all kinds of common filters. Coffee filters, air filters, water filters, fuel filters, oil filters and vacuum cleaner filters. All of these fit the classic dictionary definition of a filter. “Any porous material for trapping solids from a fluid or the air”. If we think of the noise on a circuit as the “solid” and we think of the 60Hz power (50Hz in most of the rest of the world) as the fluid, then we can see the correlation.
I was going to draw a diagram that shows your car with the gas tank in the back and the gas line going forward to the engine, but my drawing of an engine looked pretty pitiful so you’re going to just have to imagine it. If you buy gas from a station with a rusty storage tank, you may get all kinds of contaminants in your gas. To help prevent your engine from being damaged by those bits of crud and corruption, your gas line has an in-line filter. That filter is what we would call a “blocking filter”. Its job is to block the contaminants from getting to your engine while letting the gas flow through it.
hti811.gif (12499 bytes)Let’s take this to the next step. Figure 1 (I can draw this) shows a greatly simplified breaker panel with emphasis on one particular circuit. On that circuit is one particular load that is generating a lot of electrical interference that will (among other things) give your X-10 based HA system a lot of problems. Since I used a CFL (Compact Fluorescent Light) in the last chapter to illustrate the amount of “noise” one load can produce, I will continue on that vein. (There are lots of other loads other than CFL’s that can generate electrical pollution. Plus there are many brands of CFL’s that don’t pollute your electrical lines, so take this only as one possible example.)
See that little box in the middle? Without that little box, the noise from that CFL will flow right back onto the breaker panel. It has the potential of polluting the entire house (electrically speaking). Of course, it is rare that one device can mess up an entire house but I have seen it happen.
That little box is a 5 amp plug-in filter. It is a blocking filter that acts just like the fuel filter in your car (except it can’t get clogged up). ACT sells them as our part number AF100 (A = accessory, F = filter, 1 = 120v, etc. etc. and yes, Leviton and X-10 Pro also sells them, but since I have house payments to make, I would appreciate it if you buy ours.) There is a common misconception that this filter “reduces” the noise. That’s not entirely correct. It simply “blocks” it. It may not seem like much of a difference to you, but to the noise it is a big difference. You see, all that noise is still there, it’s just trapped on the load side of that filter. Sometimes I imagine that I can see the lamp cord bulging out with all that trapped noise. If you would like to use an oscilloscope and measure the levels in that lamp cord, you would see that the noise is even larger than it was before. It’s just that now it doesn’t have anywhere to go.
To see what that filter can do, click here and download a little 21 second movie that shows a noisy load (our old friend the CFL) and how a simple plug-in filter can block that noise.
Since I presume that you are continuing to read while you download that hunk-o-file, I will continue writing. Okay, so if you find yourself in a situation where you are stuck with a noisy load (television, computer, fax machine, low-voltage lighting power supply, etc.) and it draws less than 5 amps, this filter should help you out.
But what if you know what is causing the noise but it draws more than 5 amps. Well, we also have 20 amp filters. There is a 20 amp version of the AF100. The part number is AF300. It is rated for 20 amps on voltages of 277v and down (so it will also work on 120v). It is a hard-wired device but we are working on a 15 amp plug in version.
Both the AF100 and the AF300 are “series blocking filters” which means that the current must flow through them. Both of them are rated for about 40:1 blocking ratio. That means that if the noise or signal (yes they block signal, too) is at a level of 1 volt (peak-to-peak) then only about 25mv should leak through. There is no such thing as a perfect filter so if the noise level is really high, enough of it may leak through to still do nasty things.
Did you notice a couple of sentences back, that I said that these filters block noise and signal? These filters are tuned to block those frequencies that do the most damage to X-10 communications. That means that they are made to block everything between 80kHz and 150kHz (plus or minus a hertz or two). X-10 signals are at 120kHz. That means that the AF100 (or AF300) would not do you any good if you were trying to send X-10 command through that filter.
Figure 2 shows a block diagram of a simple circuit. hti812.gif (8132 bytes)Placing the filter between the X-10 device and the noisy load, blocks the noise and lets X-10 signal flow between the X-10 device and the other X-10 devices to the right. If the filter is installed on the line side of the X-10 device, it will be trapped with all that noise and it will not be able to communicate with all its X-10 brothers and sisters on the other side of the filter.
The fact that X-10 signals can’t get through these filters can be a great advantage under some circumstances. Some loads create problems for your X-10 system that isn’t electrical pollution. Sometimes the loads create little black holes that suck up all your signal. These are usually highly capacitive loads. Line capacitors are used on computer power supplies and some televisions plus other electronic devices. Interestingly, those capacitors are usually put in as “filters” but instead of “blocking” noise, they try to “absorb” the noise. Unfortunately, they also absorb your X-10 signal.
hti813.gif (8376 bytes)In this situation, you can use a blocking filter to help keep your signal strength high. It works like this. Instead of blocking noise (figure 3) they block the X-10 signal. I have another CFL that I use when I give classes. This one doesn’t generate any noise, but it sure sucks up X-10 signal. If you click right here â€¦you can see a 14 second example of that.
I know that by this time you are getting a few things out of sequence. You may still be downloading the first little movie and I am already talking about the second one. Sorry. Anyway, we have things to do so we can’t slow down. You may want to take a break and go raid the refrigerator while you download so you don’t get any more behind.
Now that you know what a blocking filter does, you are ready to learn about another filter. This one is called a “shunt” filter. Where a blocking type filter is like a “porous material for trapping solidsâ€¦” a shunt filter is more like fishing for noise with a fishing pole and bait. A shunt filter is very similar to the “capacitive load” that we talked about a couple of paragraphs back. It “sucks up noise” just like a big capacitor, but since this shunt filter designed for X-10 use, it is tuned so that it only sucks up noise that is significantly above or significantly below the 120Khz X-10 frequency.
Oh, very good. Some of you are already ahead of me. You can see the danger here. Some people buy shunt filters (they are usually less expensive than buying a bunch of blocking filters) but if the noise is at or near 120kHz, the shunt filter will have very little effect. ACT sells this shunt filter as part number AF310 (and yes, Leviton sells something very similar but remember my house payments). I have had more than a few angry people call me and complain, “I installed that there shunt filter and it didn’t do a dang thang!”. Of course, when I ask them the amplitude and frequency allocation of the noise, they have no idea what I’m talking about.
I don’t want to give you a misleading idea about the shunt filter. In many cases it does a great job. I have seen situations where several hundred electronic ballasts were all creating noise that was roughly 90kHz and down. The noise was close enough to the X-10 signal frequency to screw up the receivers but far enough away that three shunt filters (one for the A leg, one for the B leg and a third one for the C leg) cut the noise in an entire panel from 500mv to less than 30mv. That’s better than trying to install 300 blocking filters.
hti814.gif (9639 bytes)Where a blocking filter must have a current rating (5, 15 or 20 amps), the AF310 has no such rating. That is because no current flows through it. It only has two wires and is installed between the line and neutral (or sometimes, ground). The shunt filter has the job of trying to attract the noise to it and then “shunting” it to ground. The noise can be upstream or downstream (see figure 4).
Guess what? I even have another little 20 second mpeg movie that shows you how effective a shunt filter can be. Click here â€¦ to see that little 14 second masterpiece. While you are downloading that, let me tell you a little about what you will see. The first two movies showed me creating the problem, then unplugging the problem device and plugging it into a series blocking filter. In this last one, I don’t have to “unplug” anything. That is because a shunt filter is not intended to be installed “between” the problem load and the rest of the circuit. A shunt filter can just be nearby and God willin’ and the creeks don’t rise (and if the noise is significantly above or significantly below the 120Khz), it will suck it up like a Bounty paper towel.
Well now you have seen the effects of two of the most basic kinds of X-10 filters. You also have an idea why those TV interference filters that you can buy at Radio Shack do nothing to help your X-10 system. You also have just enough information to make you dangerous. If you can hold out for a while longer, we will try to finish up this 3-parter. On the first of October, I will have the last chapter in this sub-series. In that one, we will cover a little on troubleshooting (you have to know how to find that noise), a little on test equipment (you gotta have the stuff to do that troubleshooting) and then a little explanation on how I got my oscilloscope to show all that noise so impressively.
There is also one more thing I want to cram into the next chapter. What do you say, we talk about whole panel filters. Okay? Then, it’s a deal.
As usual, I am writing this at the last second (it’s about 10 minutes till midnight on Friday July 31st) and I’m sure that Bob at HTI is checking his email every 30 seconds to see if I’ve sent him this article yet. Well Bob, here it is. (Editors Note: I stayed up til 1:30 AM waiting for this puppy to download so we could bring on home … hope ya like it )
What will become of Captain Coupling?
What’s this we hear about the new ACT coupler/repeater that does extended code? What about some other new devices from ACT?
Stay tuned, children! Same Bat Time, same Bat Channel!