Posted by admin on December 8, 2012
Seeing my blog stats, other than my article on “Math Anxiety”, it appears that the great majority of readers come to read my technical articles or download my e-book or white papers. Because of this, I am thinking of posting more “how tos.” This “how to” is for a somewhat common home repair, fixing damaged door frames. Here are examples of two such repairs. The first was a badly scratched door jam at my fiancé’s house. Her late, beloved Boxer, Copper Penny, scratched at the door when she wanted in. Boxers being a strong breed, there was quite a bit of damage. An attempt had been made to patch it with filler, but when I attempted to fix an air leak with new weather stripping, it did not quite work. Being somewhat ADD, one thing led to another until the door was fixed right. Here is a look at the door jam before the repair.
Before Repair View 1
Having taken basic woodshop in seventh grade summer school, I knew replacing the damaged wood was the way to go but did not have a good idea for how to cut it out. The cut would be too deep to use a circular saw. I thought of using a chainsaw or a Sawzall®, but those tools are rather crude for making fine cuts. I called my Brother-in-Law, who, if you happen to live in South New Jersey, does home renovation and remodeling every bit as good as the pros on the home improvement shows. He suggested a tool that I had not used before, an oscillating tool. Immediately when he described it, I knew it would be the way to go. And, as luck would have it, Lowe’s had a great deal on one in their Black Friday Sale.
Oscillating Tool with Wood Saw Attachment
To repair the jam, first buy or cut a piece of wood to the size of the wood to be replaced. I found a piece of wood exactly the size I needed labeled “hobby wood” in my local Lowe’s. Mark the wood to be removed.
Mark Wood to Be Removed
Remove the damaged wood using a wood-cutting blade attachment on the oscillating tool. There are excellent videos on using the tool at Dremel®’s website.
Damaged Wood Removed
Frame with Wood Removed
Sand with coarse sandpaper enough to eliminate any wood flakes or splinters. You don’t need to fine sand as we are going to use an epoxy filler. Test for fit.
Test for Fit
Use a wood chisel or your sander to remove any high spots.
After More Sanding
This is the filler I used. It is essentially the same stuff as the Bondo® that held together the Mid-Western rust-buckets we grew up driving.
Be ready to work quickly! The filler will set in two to four minutes, depending on how much hardener you mix in. Mix the filler according to the can directions. Spread a thick layer on the area to be repaired. Press the new wood into the filler, squeezing out any excess. Remove the excess using a putty knife and paper towels. Use leftover filler to fill cracks and any surface damage.
Wait for the filler to dry thoroughly. Sand until smooth.
Paint with a good quality primer.
Reattach hardware. Your door frame looks as good as new!
Good As New!
Here is a second example, a similar repair for a door frame that has cracked and split where the door closer is fastened.
Damaged Wood Removed
With an oscillating tool and some basic woodworking skills, do not be afraid to tackle repairs like this yourself! If you feel intimidated, practice on some scrap wood before trying the real repair. You will save money, feel good about yourself, and, perhaps, like I did, find yourself to be the owner of a new power tool!
Posted by admin on December 14, 2011
Last year, my garage doors started opening and closing randomly. At first, I thought one of the neighbors had changed to use the same code, so I changed my code. But, the random openings continued, sometimes I would come home and find the left door open, sometimes it would be the right.
After some research, I came to suspect that the nearby Air Force Academy had switched to using Land Mobile Radio System (LMRS) radios, which operate on the same frequency, 390 MHz, and are known to cause interference. My opener, a Chamberlain 700WHC, has 9 dip switches with + 0 and – positions, allowing 19,683 possible combinations. It seems, though, that the new LMRS transmit 25,000 bits per second, and, sooner or later, the opener senses the right combination of bits and opens. My fix was to retrofit the opener with a new receiver with billions of possible combinations. Unfortunately, the old receiver continued to work; installing the new receiver did not stop the random openings and closings. With online searching revealing that there was no easy way to disable the old receiver, I resorted to good old fashion, tried and true methods, attacking the circuit board with a soldering iron!
Altered Circuit Board
Conveniently, the circuit board is manufactured with a test point between the radio receiver and the open/close logic. Just above the down force adjustment variable resister, labeled “DNF” (click on the photo to see the large version), you can see where I unsoldered one end of a wire to disconnect the receiver from the remainder of the circuit board. Disconnecting this jumper wire disables the radio receiver. If you don’t want to unsolder it, you can cut it.
New Opener Kit
The new receiver kit uses rolling codes and many more symbols to avoid interference. Installation is simple following the supplied instructions.
Parts Supplied in the Kit
The new kit comes with a receiver, remote. “wall wart”-style power supply, and wire. You fasten the receiver to the ceiling near the existing opener, connect the wires, plug into power, and train it to the remote.
This change solved my random garage door opening problem. This is a much simpler, less expensive, and less time-consuming solution than replace the whole garage door opener just to get around a radio interference problem!
Posted by admin on June 27, 2011
I will be teaching CS265, Algorithms, at Colorado Tech starting next week. There are still openings available. Here is the course information:
- Students are introduced to the basic concepts of algorithm design analysis, including searching and sorting, hashing and information retrieval.
- Average and asymptotic behaviors are discussed.
- Complexity issues are explored.
- Describe and in basic problems apply the methods of analysis to the algorithms that solve those problems. This includes the derivation of sequences, series and recurrence equations that define the growth function for a pseudo code fragment.
- Find the bounding asymptotic functions for various growth functions. Much dependency will be placed on applying relevant theorems and formulas without requiring their proof.
- Classify algorithms according to their bounding big oh or theta functions.
- Recognized what algorithm design type might be applied to solve a given problem type.
- Trace the execution of an algorithm based on a given design type as applied to a specific problem solution. The student will use representations of data structures to display clearly how the algorithm works.
Sign up here!
Posted by admin on July 25, 2010
I have a copy of my short presentation on the subject Engineering Disasters and Learning from Failure here.
Posted by admin on March 1, 2010
The rapidly-changing business environment and the ubiquity of the Internet and the World-Wide Web have led to the emergence of platform-independent, web-based technologies as the standard building blocks for enterprise integration. These technologies are called “Service Oriented Architecture” (SOA). Fundamental to SOA are the concepts of Web services and the Enterprise Service Bus (ESB). But SOA is also the enterprise Information Technology (IT) infrastructure – Web portals, networks, common software services, web-enabled legacy applications, and databases that support delivery of Web applications. My e-book explores the evolution and concepts of SOA in both contexts, the enabling technologies and as an enterprise IT infrastructure. You can download it here:
Introduction to SOA
Posted by admin on February 21, 2010
It is the mark of an instructed mind to rest satisfied with the degree of precision which the nature of the subject permits, and not to seek an exactness where only an approximation of the truth is possible.
Have you ever listened to the news and heard one of the talking heads make a particularly nonsensical assertion? Did you think to yourself, “That is just not believable?” When you have that experience, do you try to prove to yourself that the assertion is far fetched?
With this type of problem, you generally need only a rough estimate to demonstrate that the assertion is feasible or improbable. Many day-to-day problems are of this type. Engineers and scientists use estimates to quickly evaluate these problems. Learning and using simple estimating tricks will greatly improve your ability to critically evaluate events and opportunities in your day-to-day life.
An extremely simple yet useful estimate is to determine before beginning a calculation if the answer is greater than or less than 1. This comes in handy when converting fractions to decimals. Take 8/5 for example. When we convert this to a decimal, should the answer be greater than one or less than one? Since 8 in the numerator is greater than 5 in the denominator, we know the answer should be greater than 1. If we do the division correctly, we should get this:
However, if we mistakenly do this:
We've Made an Error
We know right away that we’ve made an error because our answer is less than 1.
Another simple but useful estimating trick is to determine if an answer should be positive or negative before beginning. It is said that the single most common error made by engineers using advanced mathematics is making a plus or minus sign error. Take for example, – ½ * ¾ * -2
We see that there are two minus signs so our answer should be positive. If we punch the numbers in our calculator and come out with a negative answer, we know immediately again that we have made an error.
The third simple estimating trick is using an order of magnitude estimate. An order of magnitude simply means a power of ten. When we estimate to an order of magnitude, we ask simply, is it closer to 1 or to 10, to 10 or 100, to 100 or 1,000, and so on. We can also estimate small numbers similarly. Is it closer to 1 or 1/10, to 1/10 or 1/100?
Using order of magnitude estimating, we calculate a “ballpark” answer before beginning a problem. If we get a final answer that is very much different, we know that we have made an error. Take for example, 7543 / 0.4359. Since 7543 is closer to 10 thousand than to 1 thousand, we round it to 10 thousand. Since 0.4359 is closer to 0.1 than to 1, we round it to 0.1. Dividing 10 thousand by 1 tenth, we get 100 thousand. The true answer, 17,204.4276 is within an order of magnitude of 100 thousand. If we made a mistake, say multiplying instead of dividing, we’d get 3,287.9937, which is closer to 1 thousand and is more than an order of magnitude from the correct answer, letting us know that we made a mistake.
Order of magnitude estimates are useful when an exact answer is not required. They can help understand big questions readily. For example, if you were to take off in a space ship just powerful enough to leave earth orbit, would it be feasible to travel to our nearest star, Proxima Centauri?
Taking escape velocity as 11.2 km per second, the distance to Proxima Centauri as 39.7 quadrillion km, and 31.5 million seconds in a year, we would come up with an order of magnitude estimate of 10 quadrillion km divided by 10 km/sec all divided by 10 million seconds per year. Using the back of an envelope, we write 10 quadrillion km / 10 km/second. We cancel one zero from the top and bottom leaving 1 quadrillion seconds. We then divide 1 quadrillion seconds by 10 million seconds per year. We cancel 7 zeros from the top and bottom, leaving 1 with 5 zeros, 100 thousand years, as our final answer. Something will certainly break on our spacecraft before 100 thousand years has passed. So, the answer is that our spacecraft may eventually get there, but it will in all likelihood be broken.
Back of the Envelope Order of Magnitude Estimate
If we do the real math, we find that our answer is not appreciably different:
3.97 × 10^13 km / 11.2 km/sec = 3.54 × 10^12 secs
3.54 × 10^12 secs / 31.5 × 10^6 seconds/ year = 112,528 years
Whether it takes 100 thousand years or 112 thousand years is not going to make a difference. None of us will be around by the time the spacecraft gets there. In this example, the order of magnitude estimate comes very close to the calculated answer. Other times, it will be further off. For our purpose, we don’t care. We know that the answer is between 50 thousand and 500 thousand years. Long enough for us to decide that this endeavor is not feasible.
What’s with the 10^13 stuff, you may ask? That’s called “Scientific Notation.” It’s an easy way to work with very large or very small numbers, numbers with a lot of zeros. It will be the subject of the next article.
Math Is Your Hammer!
In this post, we’ve talked about three easy to use estimating techniques. It is a good practice to use one or more of these when we start to solve a math problem. It is easy to make errors when performing complex calculations. Having a rough guess at the correct answer before beginning the calculation gives us a way to check to see if we have performed the calculation correctly.
In addition to improving our math, these techniques can be used to understand problems that involve large numbers. A calculation using order of magnitude estimates, usually easy enough to be done in your head or on a scrap of paper, can answer many questions of the type, “Is this feasible?” or “Is this plausible?”
Try using these techniques to answer questions in your everyday life. I would love to hear about your experience. Do you find them useful? Please leave a comment to let me know what you think!