Speech:Summer 2015 Eric Hutchins


 * Home
 * Semesters
 * Summer 2015

Note to Self: Create excel sheet to display relevant information here.

Week Ending June 17th, 2015
Currently I am working on two different aspects of my project both involving collecting as much information as I can before I actually begin working on any kind of implementation.

The first section is gathering data on power consumption on the micro and macro levels. I need to identify how much a server pulls for example, both in the on and off states. That alone would not be worth much so I am also getting information from Bob Beecher and Don Clark about how much power the building uses and how much power we gain from the solar panels respectively. This is going to be a long process from what I can tell, but hopefully I'll get that information sooner than later.

The second part that I am currently taking my time with is the X10 research. I need to work with it more involving turning objects on and off not only with the remote control but through commands in Windows and eventually Linux. If I can get it to work wireless through Linux that would be the best case scenario.I don't have a dedicated Linux machine unfortunately so I'll need to work with a VM and I can only hope I won't run into any weird problems in that route.

These two main steps are going to be the majority of my time for the next few weeks. A slow and steady process, but I have time on my side currently.

Week Ending June 24th, 2015
This week was uneventful. I am still trying to contact Don Clark and Bob Beecher but I'm not having much success in obtaining the relevant information from them. Otherwise I have not had the time to experiment with the X10 firecracker hardware or software, but plan on doing so next week. Next week I will also be starting to measure the energy draw from various sources in the server room. I will do two to three tests to show the differences between a piece of hardware in the off state, on state, and the on state with a heavy load (such as if the server is running a train or something similar.) I'm not depending on the third option, but the extra data point would be worth while.

FireCracker
This week I have been experimenting with the FireCracker hardware and software through a Windows machine. The main goal is to do this for a Linux Redhat installation, but it is important to get a baseline to make sure everything works as expected.

The software for Windows can be located here, at the bottom of the page: http://www.x10.com/software The guide for using it is located here: http://kbase.x10.com/wiki/FIRECRACKER_V2.0.0.33

In order to use the FireCracker device you need a USB to Serial adapter for your computer, unless for some reason you already have a Serial port. Since that is highly unlikely you'll need to purchase one for this model. The main receiver unit, the one covered in some kind of hard plastic, receives signals from the transmitter. It then sends X10 signals along the electrical wiring of which individual X10 modules can pick up and react to accordingly. Each module can be given two different sets of frequencies to listen to, 1-16 and A-P. The main receiver, the plastic one, is permanently set to 'M'. For all testing purposes only use M if you expect a result.

Using the manual transmitter I can send signals to the receiver which sends signals out to the modules. All of those components work as expected. The only thing not working is the Wireless FireCracker transmitter that you can hook up to the PC and send out commands through a GUI. The FireCracker is hooked up via a USB to Serial adapter, but does not seem to be responding. I've installed drivers for everything involved and I'm running it all on Windows 7, but no luck as of yet.

This coming week I need to do more testing to figure out what the issue is, whether it is a driver, whether it is the device itself, etc.

Energy Bill Information
In regards to getting information from Bob Beecher on the power consumption rates of the UNH building I have had success. The full reports for the past several months can be found here:

https://drive.google.com/open?id=0By-K58vIObCrfjBKQmhTaTgwNGYxMkxrMmlnWWF1djYyaEdYdmI5djAtbnRZamxZa0N2Ym8&authuser=0

It appears more complicated than it actually is. If you are looking over it yourself note that each line beginning with 'Detail for Service at:' is a different meter. It appears that each floor as at least one, sometimes two, meters.

Using the time period of May 26th to June 24th (29 days) we can see the following:

The entire energy consumption for that time period is 24,440 KWH.

The energy rate is $0.105600

The total cost is therefore $2580.864.

Looking at the summarized bill above, that seems close enough. There may be some credits on the account which would explain the small difference.

To summarize these findings, that means our energy costs are around $2,600 a month and our energy expenditure is around 25,000KWH per month. If we project that out it means we are spending approximately $31,200 and 300,000 KWH per year.

Next week I will try to find more time to go into greater detail with each period and begin showing what the servers are projected to cost if kept on/off all year long. This Wednesday I will also be doing some double checking on measurements to make sure I get the numbers I need to get.

Week Ending July 8th, 2015
The majority of this week was spent attempting to get the FireCracker to transmit. I had no luck regardless of operating system and regardless of Serial to USB adapter I used. Marcel should be bringing in his radio card that can receive all sorts of signals, so he'll be able to tell me if this is transmitting anything or not.

Other than this, I have not done anything else of significance this week.

FireCracker Update
This is going to be another short entry for the week. The main reason is that Marcel has the FireCracker and is attempting to get it to work via some Serial port switching. Of note is that the device does function, but it only functions on Professor Jonas' old computer in his office. We almost didn't find out as Jonas originally plugged the device into a computer that was off instead of the one he was using, but at least we know the device works.

In truth we'll need to get something better than a Serial device anyway. The point of this project is to save energy, not use old pieces of technology. Unless Marcel can come up with a solution that allows me to plug it into any of the PCs that I own, or if we can find an adapter that actually translates the information properly, we'll need to purchase a new one.

Potential FireCracker Replacements
If and when I do choose to replace the FireCracker with something a bit more modern I need to make sure I get something that is cheap and will suit the project. Here is my best choice from what I've seen:

http://www.x10.com/x10-home-automation/interfaces/cm19a.html

This is the obvious choice. It is only $30 and is by the same company that made the FireCracker CM17A. This is the new USB version of the device, the CM19A transceiver. Just for reference for those unfamiliar, a transceiver is a transmitter and a receiver, so it can send and receive X10 commands. The CM17A is the same, but of course it is a Serial device, not USB, which causes many problems with modern computers.

Solar Panel Update
It seems Don is hard to get a hold of these days. I've left a message earlier in the week but he has not yet called me back. I've tried calling him again with no luck, straight to voice mail. It's possible he could be on vacation, but I'll just have to keep trying to get in touch with him.

For future reference Don Clark's phone number is - 603-669-9000 - Ask for Don Clark when the receptionist picks up and she'll take you right to him.

I should note again that this step isn't crucial by any means, but it is still somewhat nice to have this sooner than later.

Current Progress
This week I spent a majority of the time trying to get a virtual box version of XP to work with the Firecracker and a usb device. Eventually after speaking with Jonas I realized there are serial ports on the backs of the machines, and although I lose the ability to tinker with it at home or at work, I can at least spend some time in the server room to get it started there.

Marcel has also been a lot of help with doing tests of his own on Linux, and I'll be speaking with him this week in more detail. Apparently he also has a FireCracker module and is successfully using it on a Linux machine with little effort. This is a good sign as it means this may be easier than expected. Even still I need to get it working on the correct version of Linux, RedHat specifically, and have it transmit on and off signals. Once that happens I will be done with the first step. The next step would be programming a solution that sends signals out of the FireCracker to turn on and off specific machines, and generally make a useful interface that people can remote into and take a look at to view the status of any particular server.

Solar Panel Info
Spoke with Don Clark and he was unable to access his archived emails to send me the report for last year. I asked him to inquire again about it and he said he'd get it from the right person and send it over to me.

I was able to get some preliminary numbers out of him, so at least we have a ballpark. He stated that we collected 151791kWH of energy in about three years time, which means we get about 50,000kWH per year. I won't go into too much analysis until I get the report from him, and I'll share it here in these logs. I hope to expect it by next week.

Week Ending July 29th, 2015
Extremely sick this week.

Week Ending August 5th, 2015
Still very sick this week, nothing to report.

Week Ending August 12th, 2015
Finally I am no longer as sick as I was, so I'm starting to work on this project once again. I plan on going into the next meeting to accomplish several things:

1. Speak with Marcel who said he had done some interesting experiments via Linux and a Firecracker, and that he knows how to implement it. Essentially he says that it is simple to communicate with the device, which has been my biggest issue, but also he was able to make a simple Python program to issue commands which is even better.

2. I have obtained a license for Red Hat Linux. I plan on installing this onto a spare laptop, of which I believe Jonas has one for me that I can borrow.

3. Catch up on anything else I missed. In the next log I will give a summary and hopefully show some progress.

Solar Panel News: I am still having trouble with Don Clark to an incredibly frustrating degree, as he seems incapable of sending a simple file to me. I'll have more on that, and hopefully the actual document, next week.

Week Ending August 26th, 2015
As this is the final week of the Summer I will outline the status of everything that has been accomplished and what some of the goals will be for next Spring when I take a serious look into this project. I definitely made some significant progress over the Summer, perhaps not as much as professor Jonas would have liked, but I am satisifed with how far I have come.

Laptop Situation
I have finally been able to obtain a spare laptop and install Red Hat Linux 6.6 onto it. This allows me to do my own testing at work or at home without much hassel, and as I am using the exact same operating system as Caesar any solution I create should be easy to move over.

CM17A FireCracker Driver Installation
Marcel was able to find a useful guide for installing drivers for the FireCracker module.

This is the guide:

The steps to prepare for it are simple, just making sure you have the basics such as Python and PyUSB installed. The only issue I had with this step was figuring out how to get into root, as my base user login did not belong to the sudoers group. To fix this I just switched users and logged in as root to avoid any other issues with permissions.

The next step is to download the package (At the bottom of the website), and unpack it. Then navigate via terminal to the top-most level of that package and create several folders as described.