KHarms
Oct 30, 2009, 11:02 AM
One of the most frustrating problems to solve is a device that is not going on or off as expected. It is hard to know why. It could be HCA. It could be the power line interface. It could be a signal propagation problem. It could be noise. It could be a problem with the device itself.
This problem is even more frustrating when the computer can't seem to turn the device on and off, but you can from simple plugin keypad or timer.
For these problems to be resolved you must be a detective. A detective works in a logical ordered manner. Starting with a whole list of what could be wrong and working methodically to remove one possible explanation at a time. With effort, you will be able to eliminate many things that are not problems and, hopefully, be left with a single cause that you could fix.
Let's start by focusing on the software, computer, and power line interface.
In order for the computer to control the device many things need to happen. The HCA software has to instruct the communications port on the computer to send a signal to the power line interface. The power line interface must receive the signal and place the command on to the power line. Then the device must receive it.
It is hard to know where the problem is because there are so many layers. Does HCA send the right stuff to the communications port? Does the power line interface receive it from the computer? Does it put the signal on the power line? Once we know that the signal is on the power line then we can eliminate HCA, the computer and the power line interface. Later on we can see why it doesn't reach the device.
The best tool for this part of the investigation is a signal meter. The fanciest ones can see the signal, measure it, then decode and display it. Problem is they cost a lot. Less expensive ones measure the signal but doesn't decode it. Take a look at your favorite automation store site and see what you can find.
By plugging the signal meter into the power line some place close to the power line interface attached to the computer you can see when you use the HCA user interface to control a device if a signal is being put on the power line. When using the HCA user interface, if you never see signals being placed on the power line for any devices then the problem lies in the software, the computer or in the power line interface. HCA Technical support can help with this.
This is the most IMPORTANT point of your testing. Until you know if HCA is placing a signal on to the powerline at the time you expect it to, you will not know if it what is wrong is on the computer / HCA / power line interface side, or after the power line interface and the signal is being lost in your home some place.
Let's assume for this discussion, you have shown that you do see a signal on the power line when you expect HCA is sending it.
Next we have to deal with why the signal doesn't control the device. We know it is being placed on the power line. The simplest reason could be that the address you are sending the signal to is different than the address of the device. Sounds obvious! But mistakes like these happen all the time. If it is an X10 device and the device in HCA is set to, for example, C1 and the switch in the wall is set to D1. If it is an Insteon device, check that you entered into HCA the three part address correctly. There are usually stickers on the device that show the Insteon address. Check the most obvious things first.
Second, the device could be broken and no signals will activate it. The way to determine this is to use another transmitter besides the computer to send a signal to the device. If neither transmitter can activate the device, then it is a good bet that the device is broken. A good tool for this is a plugin keypad. This can be plugged in close to the device so that you can know that the signal generated by the keypad gets to the device. Again, check your favorite automation store site for plugin keypads. They are made for X10 and for Insteon.
If the device works then we have are left with the hardest problem of all. Why, when you use a keypad to control the device, does it not work when you try and control it from the computer?
There are four possible problems that could be causing this.
First, it could be a signal propagation problem. That is, the signal that the power line interface generates doesn't reach the device at a sufficient level to control the device. This can be very common. If you can control some devices but not others this is a likely cause.
Second, noise on the line could corrupt the signal so that when it reaches the device it doesn't recognize it.
Third, the signal could collide with other signals being transmitted and become corrupted.
And fourth, it could be a combination of the other three possibilities. Unfortunately, it usually is this. So in our discussion we will deal with all three at once.
The primary problem is that no two powerline signals are alike. What you get out of a keypad is not the same as what you get out of the power line interface connected to the computer. Why? Well, they have different implementations of their transmit circuits so they generate slightly different signals at different power levels.
Also, no two signals placed on the power line at two different points are the same even if generated by the same transmitter. The two points you choose may be on different branch circuits or, at the very least have different wire lengths to the target device.
If your home doesn't have a coupler or amplifier / repeater connecting both sides of your power panel, you will have problems. You may not know this, but all of your branch circuits in your home are divided into two groups. Signals sent on a circuit in one group have a really long path to get to a circuit in the other group. This is what a coupler or amplifier / repeater does - it shortens the path. Repeaters are made for X10 and Insteon. With Insteon you may have a number of repeaters and you can move them around your home and plug them into different locations. This may improve signals propagation.
Next we have to think about what happens to the signal as it moves down the wire. One branch circuit may have many devices - each one absorbs a bit of the signal strength - and another branch circuit may have few or no devices. So if you place a signal on the heavily loaded branch circuit it may not make it to the target device. If that signal was placed onto a little used circuit it would make it to the device.
The heavier you load the line (circuit) the more signal absorption you will get. This is not unique to automation devices, it's the same for all loads on the line. Each device you put on a circuit will introduce two things, first a higher current draw on the circuit, and second more resistance. Every automation device you put on the line does indeed attenuate the signal in the rest of the circuit. This results in most cases increased current in the circuit and more resistance, both of which makes it more difficult for automation devices to decode the signal. The problem is worse with each device added. Also, some devices themselves generate noise.
If, for example, you were to go to that circuit and plug in another device and you just done all of the above - increased the current draw, add more resistance, and attenuate the signals. If you're near the thresh-hold on that circuit the new device may or may not work. Let me say that again: Just adding a new device to a circuit may cause everything to stop working.
And often the worst circuit in the home is the one the Computer interface is plugged into! That is because that circuit has the computer on it with probably a surge suppressor someplace - a big signal sink. If you are like most folks you have lots of other electronic gear plugged in to that circuit which adds lots of noise and signal attenuation. Home Entertainment Centers are the worst. If noise is a problem, one way to solve it is to put a 10A filter into the outlet then use power strips for the things to plug in. This should be done with any electronic equipment because even though it appears to be off, it is really still on and still generating noise on the line.
Generally you'd have to have a circuit really loaded with a large number of devices before you started pulling down the signal but some devices are indeed worse than others. Your power line interface should be on a circuit that is as free from other electronic gear as possible. For example, your power line interface should not be on the same circuit as your Entertainment Center even if it's filtered.
Again, the best tool for to help with signal propagation problems is the signal meter.
A problem with signal meters is that what you want to do is to place it at the exact point where the signal is to be received. This is easy to do when it is a plug in module you are having problems with and harder to do when it is a switch. You will need to do some detective work to see if there is a receptacle outlet on the same branch circuit as your light switch so you can get the signal meter close enough to the destination so the test is meaningful. Then generate the signal from the computer and take note of the signal strength. Now use a keypad controller and see what the signal strength is received at that same point. They probably will be different. Is the computer signal so low that the switch doesn't respond? That's what we need to know.
Before we look to solutions there is one type of problem we haven't talked about: Collisions of signals on the power line. Because of the way powerline signals work, transmitters don't know if the signals they transmitted were messed up by another transmission on the power line. This can happen in odd ways. For example, you turn a light on in a hallway. The hallway has a motion sensor. Many of those motion sensors transmit when it gets light and when it gets dark. If it is at night - or a dark day or dark part of the house - each time you send a signal to turn the light on, the motion sensor transmits. If you are doing a series of commands on the power line, subsequent commands after the hallway light may not be received correctly by their target devices since the motion sensor transmission messed them up.
Finally, remember that your problems can be intermittent. Turning things on in the morning can work fine and in the evening things don't work at all. How could that be? Because at various times in your home different things are on and different things are off. One home we worked with had compact fluorescent lights that when on in the evening totally destroyed the powerline signals. HCA was able to control everything in the home until dusk when the lights went on and then it all just stopped working. This was, as you can imagine, very hard to find.
So, what's the answer? First, locate as many obvious noise sources as possible: TV, CD, DVD, computers etc. and isolate them with filters. Second, and most people won't like this, use as few 2 way devices as possible to reduce collisions.
Be prepared to buy lots of filters. Chances are you'll spend more on filters than you did on your power line interface. Some folks have them all over the house. One tip is to not forget the microwave. You need a 15A filter on those as they draw 11 to 15 amps. When the microwave is on, it generates both power line noise and RF noise and is generally the biggest noise maker in your home.
When you think about it what we are really trying to do is to run a communications network over the existing power lines which of course they were never designed to do. Nor were most of the devices plugged into the power line designed to be noise free. This leaves us with a very noisy network in which devices can become intermittent or not work at all.
The only truly reliable device for finding and fixing noise on the power line is an oscilloscope so you can see the noise and the powerline signals and essentially "tune" your power line for max through-put and minimum noise. This is obviously beyond the scope of most folks but in some homes a professional automation installer will do this.
Getting powerline communication to work can be a challenge. The important thing to remember in your testing is to remember that all signals on the power line are different if produced by different transmitters at different locations. Work slowly and carefully. Eliminate things so you know that they are not the problem.
This problem is even more frustrating when the computer can't seem to turn the device on and off, but you can from simple plugin keypad or timer.
For these problems to be resolved you must be a detective. A detective works in a logical ordered manner. Starting with a whole list of what could be wrong and working methodically to remove one possible explanation at a time. With effort, you will be able to eliminate many things that are not problems and, hopefully, be left with a single cause that you could fix.
Let's start by focusing on the software, computer, and power line interface.
In order for the computer to control the device many things need to happen. The HCA software has to instruct the communications port on the computer to send a signal to the power line interface. The power line interface must receive the signal and place the command on to the power line. Then the device must receive it.
It is hard to know where the problem is because there are so many layers. Does HCA send the right stuff to the communications port? Does the power line interface receive it from the computer? Does it put the signal on the power line? Once we know that the signal is on the power line then we can eliminate HCA, the computer and the power line interface. Later on we can see why it doesn't reach the device.
The best tool for this part of the investigation is a signal meter. The fanciest ones can see the signal, measure it, then decode and display it. Problem is they cost a lot. Less expensive ones measure the signal but doesn't decode it. Take a look at your favorite automation store site and see what you can find.
By plugging the signal meter into the power line some place close to the power line interface attached to the computer you can see when you use the HCA user interface to control a device if a signal is being put on the power line. When using the HCA user interface, if you never see signals being placed on the power line for any devices then the problem lies in the software, the computer or in the power line interface. HCA Technical support can help with this.
This is the most IMPORTANT point of your testing. Until you know if HCA is placing a signal on to the powerline at the time you expect it to, you will not know if it what is wrong is on the computer / HCA / power line interface side, or after the power line interface and the signal is being lost in your home some place.
Let's assume for this discussion, you have shown that you do see a signal on the power line when you expect HCA is sending it.
Next we have to deal with why the signal doesn't control the device. We know it is being placed on the power line. The simplest reason could be that the address you are sending the signal to is different than the address of the device. Sounds obvious! But mistakes like these happen all the time. If it is an X10 device and the device in HCA is set to, for example, C1 and the switch in the wall is set to D1. If it is an Insteon device, check that you entered into HCA the three part address correctly. There are usually stickers on the device that show the Insteon address. Check the most obvious things first.
Second, the device could be broken and no signals will activate it. The way to determine this is to use another transmitter besides the computer to send a signal to the device. If neither transmitter can activate the device, then it is a good bet that the device is broken. A good tool for this is a plugin keypad. This can be plugged in close to the device so that you can know that the signal generated by the keypad gets to the device. Again, check your favorite automation store site for plugin keypads. They are made for X10 and for Insteon.
If the device works then we have are left with the hardest problem of all. Why, when you use a keypad to control the device, does it not work when you try and control it from the computer?
There are four possible problems that could be causing this.
First, it could be a signal propagation problem. That is, the signal that the power line interface generates doesn't reach the device at a sufficient level to control the device. This can be very common. If you can control some devices but not others this is a likely cause.
Second, noise on the line could corrupt the signal so that when it reaches the device it doesn't recognize it.
Third, the signal could collide with other signals being transmitted and become corrupted.
And fourth, it could be a combination of the other three possibilities. Unfortunately, it usually is this. So in our discussion we will deal with all three at once.
The primary problem is that no two powerline signals are alike. What you get out of a keypad is not the same as what you get out of the power line interface connected to the computer. Why? Well, they have different implementations of their transmit circuits so they generate slightly different signals at different power levels.
Also, no two signals placed on the power line at two different points are the same even if generated by the same transmitter. The two points you choose may be on different branch circuits or, at the very least have different wire lengths to the target device.
If your home doesn't have a coupler or amplifier / repeater connecting both sides of your power panel, you will have problems. You may not know this, but all of your branch circuits in your home are divided into two groups. Signals sent on a circuit in one group have a really long path to get to a circuit in the other group. This is what a coupler or amplifier / repeater does - it shortens the path. Repeaters are made for X10 and Insteon. With Insteon you may have a number of repeaters and you can move them around your home and plug them into different locations. This may improve signals propagation.
Next we have to think about what happens to the signal as it moves down the wire. One branch circuit may have many devices - each one absorbs a bit of the signal strength - and another branch circuit may have few or no devices. So if you place a signal on the heavily loaded branch circuit it may not make it to the target device. If that signal was placed onto a little used circuit it would make it to the device.
The heavier you load the line (circuit) the more signal absorption you will get. This is not unique to automation devices, it's the same for all loads on the line. Each device you put on a circuit will introduce two things, first a higher current draw on the circuit, and second more resistance. Every automation device you put on the line does indeed attenuate the signal in the rest of the circuit. This results in most cases increased current in the circuit and more resistance, both of which makes it more difficult for automation devices to decode the signal. The problem is worse with each device added. Also, some devices themselves generate noise.
If, for example, you were to go to that circuit and plug in another device and you just done all of the above - increased the current draw, add more resistance, and attenuate the signals. If you're near the thresh-hold on that circuit the new device may or may not work. Let me say that again: Just adding a new device to a circuit may cause everything to stop working.
And often the worst circuit in the home is the one the Computer interface is plugged into! That is because that circuit has the computer on it with probably a surge suppressor someplace - a big signal sink. If you are like most folks you have lots of other electronic gear plugged in to that circuit which adds lots of noise and signal attenuation. Home Entertainment Centers are the worst. If noise is a problem, one way to solve it is to put a 10A filter into the outlet then use power strips for the things to plug in. This should be done with any electronic equipment because even though it appears to be off, it is really still on and still generating noise on the line.
Generally you'd have to have a circuit really loaded with a large number of devices before you started pulling down the signal but some devices are indeed worse than others. Your power line interface should be on a circuit that is as free from other electronic gear as possible. For example, your power line interface should not be on the same circuit as your Entertainment Center even if it's filtered.
Again, the best tool for to help with signal propagation problems is the signal meter.
A problem with signal meters is that what you want to do is to place it at the exact point where the signal is to be received. This is easy to do when it is a plug in module you are having problems with and harder to do when it is a switch. You will need to do some detective work to see if there is a receptacle outlet on the same branch circuit as your light switch so you can get the signal meter close enough to the destination so the test is meaningful. Then generate the signal from the computer and take note of the signal strength. Now use a keypad controller and see what the signal strength is received at that same point. They probably will be different. Is the computer signal so low that the switch doesn't respond? That's what we need to know.
Before we look to solutions there is one type of problem we haven't talked about: Collisions of signals on the power line. Because of the way powerline signals work, transmitters don't know if the signals they transmitted were messed up by another transmission on the power line. This can happen in odd ways. For example, you turn a light on in a hallway. The hallway has a motion sensor. Many of those motion sensors transmit when it gets light and when it gets dark. If it is at night - or a dark day or dark part of the house - each time you send a signal to turn the light on, the motion sensor transmits. If you are doing a series of commands on the power line, subsequent commands after the hallway light may not be received correctly by their target devices since the motion sensor transmission messed them up.
Finally, remember that your problems can be intermittent. Turning things on in the morning can work fine and in the evening things don't work at all. How could that be? Because at various times in your home different things are on and different things are off. One home we worked with had compact fluorescent lights that when on in the evening totally destroyed the powerline signals. HCA was able to control everything in the home until dusk when the lights went on and then it all just stopped working. This was, as you can imagine, very hard to find.
So, what's the answer? First, locate as many obvious noise sources as possible: TV, CD, DVD, computers etc. and isolate them with filters. Second, and most people won't like this, use as few 2 way devices as possible to reduce collisions.
Be prepared to buy lots of filters. Chances are you'll spend more on filters than you did on your power line interface. Some folks have them all over the house. One tip is to not forget the microwave. You need a 15A filter on those as they draw 11 to 15 amps. When the microwave is on, it generates both power line noise and RF noise and is generally the biggest noise maker in your home.
When you think about it what we are really trying to do is to run a communications network over the existing power lines which of course they were never designed to do. Nor were most of the devices plugged into the power line designed to be noise free. This leaves us with a very noisy network in which devices can become intermittent or not work at all.
The only truly reliable device for finding and fixing noise on the power line is an oscilloscope so you can see the noise and the powerline signals and essentially "tune" your power line for max through-put and minimum noise. This is obviously beyond the scope of most folks but in some homes a professional automation installer will do this.
Getting powerline communication to work can be a challenge. The important thing to remember in your testing is to remember that all signals on the power line are different if produced by different transmitters at different locations. Work slowly and carefully. Eliminate things so you know that they are not the problem.