Blogger Link Colors

I had the hardest time trying to fix the color of the links here in Blogger. It used to look like this.

Actually the blue was quite a bit brighter and it virtually strobed against the pale tan background. I wanted a less jarring color. I was also unable to change the color of the title.

I tried to fix it through the Customize/Layout/Fonts and Colors - controls and that made no difference.

I Googled around and found some suggestion on how to change them in HTML. I opened up the HTML through Customize/Layout/Edit HTML and saved a copy before making any changes. The variables for the text were already set to the colors which I had specified in Fonts and Colors. Further down, in the section with "a:link", "a:visited", and "a:hover" the colors were assigned by the variables above. I tried substituting specific colors for the variables, but it made no difference. (I'm only barely HTML literate.)

Then I tried radical changes. I selected a whole different template. That changes the appearance dramatically, but the color of the title and of the links didn't change. That indicated to me that template was not causing the trouble. I thought it was the template which controlled this, but apparently not.

Then I learned something interesting. If I only display one posting at a time, the colors I specified do show! That indicated again that there must be some code outside of the template which caused the problem. The only remaining place is in a posting. I don't know why this would be, but it seems to have been the case. Here's what I did:

1. I decreased the number of posts displayed at a time - Customize/Settings/Formatting/Show - to five. At that point, my first page looked right. It didn't have the bright colored links.
   
2. I went through the pages to see when the bright colored links were displayed.
   
3. On the page which had the problem, I looked at each posting one at a time. I found the one posting which had the problem.
   
4. I looked at the HTML for that posting and there was a ton of code which I didn't recognize and which didn't seem to have anything to do with the text or the pictures. (I was experimenting with using an online blog editor and MS Word to write blog posts at about that time. I suspect that caused the problem.) I deleted the excess HTML and the bright blue was gone!
   

I had to go back and play with the settings again, but I could do it with the standard controls at Customize/Layout/Fonts and Colors and it worked just fine.

Eureka!

Before I figured this out I had no luck finding instructions which related to the problem above. Now that I figured it out and wrote the above, all the instructions which comes up relate to this problem. (frustration) I recommend the clear instructions by BlogBulk.

Shaker Workbench

I really want to put together a nice post showing my workbench and explaining why it was built the way it is, but I haven't done it. So, here are a bunch of pictures.











Click on "Workbench" to see more posts about my Shaker Workbench or "plans" to see the SketchUp model

Since these pictures, I've added the anti-wrack pad to the front vise (which works great by the way) covered the chops of the front vise with leather, added the sliding board jack, and probably other things too.

More pictures and better formatting later.

Leather work

I used leather to cover the faces of my woodworking vise. I had extra leather and I found that I can make some very useful things without too much work.

The only tool that was really necessary was the needles. I already had them from some rope work I had done in the past, so they didn't cost anything.

I bought an overstitch wheel which marks the stitches before drilling the holes. I used it only on the second larger sheath. I think the even stitches there look much better and they certainly were easier to mark.

The chunk of scrap walnut has been smoothed on one end on a belt sander. It is used to smooth the edges and rub the leather into place.

The piece of hickory with the hook on the end is for marking where the stitches go in comparison with an edge. It is effectively a marking gauge.

I might invest in a v-gouge at some point in the future. It is used after the marking gauge to cut a thin depression for the threads. This decreases wear on the threads. It is also used to cut thin strips from the inside of the leather to make it fold more easily.

I also used contact cement, a utility knife, a straight edge, a hammer, and my belt sander. That's about it for tools.

The lower sheath was stained with walnut Watco Danish Oil. I don't really think it looks better, but it was interesting to try. It worked very well even though it's not meant for leather. It has also been heavily waxed. The upper sheath has not been treated at all. I'd like to try soaking it in hot paraffin, but haven't got that far yet.

The button is one I cut from a branch and soaked for a couple of months in boiled linseed oil.

I couldn't find any description of how one might make a draw knife sheath or a push knife sheath anywhere on the web. If I made the push knife sheath again, I'd cut it so that it was narrower where the straps come over the back so that the straps would bear directly on the back of the blade. I'd also attach the button further from the back of the blade. The draw knife sheath might be improved a little. It should fit the length of the blade more closely and the flap should be larger and looser (I'm not sure how to make it looser). That would allow the knife to be inserted without having to hold the sheath right near the very-sharp blade.

This is not sophisticated leather work, but I was quite pleased with what I was able to do without a lot of practice or training.

Rendering – Industrial Machine

Some people like to make photo-realistic images out of 3D models. It’s very cool to see something I’ve made in SketchUp turn into something which looks real.

A real photograph taken from the Grizzly catalog:

Two views of a SketchUp model which I made:

A rendering of my SketchUp model by Solo:

I enjoy making models in SketchUp. When I am modeling something which already exists, I am effectively reverse engineering the object. In doing that, often I can understand better how it works and why it works that way.

Rendering is something very different. Changing the characteristics of the surfaces and changing the light sources makes the picture more realistic, informative or interesting. The process is iterative in that a light source can create another light source in the form of a reflection. Then that reflection’s presence has to be accounted for by the program. There is judgment and subtlety involved in preparing a rendering. The computing power and time is huge, and while it’s not too hard to do a decent job, doing an excellent job is very challenging. Then, if you are doing it in a 3D animation, each frame of the animation must be rendered, and that takes even longer.

In this example, I was planning a future workshop layout and I wanted to see how I might fit a knife belt sander into the plan. I only needed a rough model for this purpose, but as I tend to do, I went overboard and made it as close to perfect as I could. I posted it as part of a contest in the SketchUcation forums.

Similarly, there’s a rendering of a wrist watch which I modeled here.

Experiments with Glue and Wood

I wanted to glue up a workbench top 90” long, 30” wide, 4” thick using laminations of 1 1/4" southern yellow pine (SYP). To do it with clamps every six inches would require 14 clamps. I only had four clamps over 30" long, so I wanted to try another way. I got advice on the SawMill Creek forum and decided to try applying glue, screwing each board to the next one, then after the glue dried, removing the screws and gluing on another board.

I wasn’t sure if the screws would apply enough pressure, so I experimented.

Experiment 1:

Two pieces of SYP each 9" x 7”.
Screws centered on long axis 1.5” and 4.5” from one end.
#10 x 2-3/8" ProMax Round Washer Head Dry Lube Screws - from McFeely’s
Titebond glue spread with the objective of just getting a bit of squeeze out after clamping

As you can just see in the lower left corner, the screw force caused the two pieces to bend away from each other on one side. In the right picture, the 9 x 7 grid has been cut into 63 pieces for testing.

I laboriously started measuring the breaking force for each joint.

It's not a sophisticated system, but I was able to fill the bucket hanging at the end of that cord with junk until the joint broke, then weigh the bucket. I soon noticed that there was a lot of variation in measurements but there were just three kinds of breaks: 1) at the glue line, 2) partly at the glue line, and 3) all in the wood. I stopped weighing the bucket and started breaking the blocks and recording the kind of break.

Here are the results of experiment 1:

Clearly, the bent boards I mentioned above caused the upper board to bend away from the lower along the left edge and there was little or no clamping force there.

Conclusions from Experiment 1:

The screw force was sufficient to make a good glue joint.
The clamping force appeared to extend along the grain at least 4 1/2”
The clamping force appeared to extend across the grain 1 1/2".
Wood bends when you squeeze it.






Experiment 2:

Two pieces of SYP each 13" x 4”.
Screws centered on long axis 2.5” and 4.5” from one end.
#10 x 2-3/8" ProMax Round Washer Head Dry Lube Screws
Titebond glue spread with the objective of just getting a bit of squeeze out after clamping

Here’s a SketchUp version of the setup.














I cut one inch squares the long way and short way. Because my circular saw doesn’t cut deep enough, I finished the remaining short cuts with a handsaw. It wasn’t necessary to finish the long cuts.

You can see in the right side picture that the square on the right edge of the glue-up, furthest from the screws broke off cleanly with no adhesion.




Here are the results:

There is a definite anomaly here. It was caused by an oily knot which was on the bottom of the top board. I had hand-planed both boards, but I had not wiped them with naptha as is suggested for SYP.

Conclusions from Experiment 2:

The clamping force appeared to extend along the grain at least 8 1/2”
The clamping force appeared to extend across the grain 2 1/2".
Oily spots, usually related to knots, do not glue well.















Experiment 3:

Based on the data above, I had a theory. A screw puts a point load on the top surface of the wood where the head rests. Force spreads from that point in an oval which is substantially longer in the direction of the grain. For a 1 1/2" board, it covers about 4” side to side and about 12” along the grain. Because it’s an oval, it doesn’t cover the corners of that rectangle.

Based on that theory, I thought I could glue up my 4” wide board with screws every 5” staggered from side to side and an extra at each end. It would look like this, but it would be 90" long.



Two pieces of SYP each 26x4.”
Screws centered on long axis 2.5” and 4.5” from one end.
#10 x 2-3/8" ProMax Round Washer Head Dry Lube Screws
Titebond glue spread with the objective of just getting a bit of squeeze out after clamping

This is what it looked like after I had broken each 1" square and put all the samples back in place. All had broken in the wood, meaning that the glued joint was stronger than the wood itself.


Because the screws were removed after the glue dried and before the next board was glued on, there is no metal in the top. And, it worked. It's been about six months and my top has not shown any cracks nor has it warped at all since the first month.

Conclusion: A $7.50 box of screws can substitute for $500 worth of clamps. But it will take a whole lot more time and effort.

IceBike Shoes

I used to bicycle a lot; hence my web handle "SchreiberBike". Riding year-round in Illinois means learning how to ride in cold weather. I put this information up in a web site back in 2001. It was hosted on Tripod.com and they took it down without notice because of an error I made. Although I had some good feedback, I didn't put it back up.

Here's the information from 2001, with a few edits, pulled from the Internet Archive's WaybackMachine. I recovered most of the pictures from my PC.

Shoes for Cold Weather Cycling

How I Keep my Big Feet Warm

For me, keeping my feet warm has always been the hardest part of winter cycling. I've just made up a pair of shoes that have finally solved the problem for me. I've tried many suggestions from other ice bikers, but I haven't been able to do the most effective one: wearing extra layers of socks inside extra large loose fitting shoes.

When you normally wear size 13 shoes (48 Euro) it's hard to find cycling shoes that are large enough to allow for wearing extra layers of socks without constricting your feet. There are a couple brands available in up to size 50, but at prices over $150.00 US I couldn't justify them. (Doing things inexpensively is critical.) I settled on finding shoes and modifying them to use SPD cleats.

Shoes
A logical spot for big shoes is the nearby University of Illinois football program. Almost all of the football players are big guys so I thought they might have some old shoes they could spare. I found that they were going to donate last year's shoes to Goodwill soon. I haunted Goodwill and was rewarded with a pair of size-15 high-top turf shoes for $8.00. The turf shoes have a rubber sole with dozens of quarter-inch nubs rather than metal cleats.

Cleats
At this point I had two options for the shoes: clipless pedals or PowerGrips. I had done my winter riding with PowerGrips for years, but my rat trap pedals were shot and I'd have to get the extra-large size which were about $30.00 US; so I decided to go with off-brand SPD compatible pedals from Nashbar. The good news was that they were on sale for about $20.00. If price was no object I would have liked the SpeedPlay Frogs, but the cost was too high.

Cleat location
I eyeballed the shoes next to my present shoes and used a marker to show where I thought the cleat should go on the new shoes. The standard mounting for SPDs have about 3/4 inch of front to back adjustment and 1/4 inch of side to side. I was glad I had that much adjustment available, because I found it difficult to estimate where the ball of my foot would fall in the over-size shoe and how much clearance I would need for the crank arm.

I didn't want to attach the SPD cleat directly to the shoe because the localized stress would likely destroy the sole of the shoe, and because I wanted to distribute the pressure of the cleat over more of my foot (diagram below). I visualized a metal plate that would be as large as possible, but no heavier than necessary, and not so large as to disturb the natural shape of the sole. The bottom of the shoe is not flat. I drew a shape on the sole of the shoe about three and three quarter inches by two inches and cut the nubs off of that area with a utility knife.

Hardware
The next step was to make up a metal plate. An engineer could tell me more about the appropriate material to transfer the forces, but what I did was go to the hardware store and hold samples of metal in my hand until I found one that felt right. It was also terribly cheap. I used a steel cover plate for an electrical box. I bought two, but was able to make do with one. Cost 44 cents. Tracing the space on the shoe onto cardboard, I copied the shape onto the metal, cut it with a jigsaw, and shaped it with a file. The fore and aft position of the cleat is adjustable along 1/4 by 3/4-inch slots. I cut the slots in the metal by drilling holes then filing along the lines where the cleat would go.

To attach the metal plate to the shoes I drilled holes in the shoes that matched the holes on the mounting plates. I then used 1/4-inch machine screws screwed into "T" nuts, which are inside the shoe beneath the insole. "T" nuts come with a variety of shank lengths. So long as the length of the shank is less than the thickness of sole of the shoe when the screw compresses it, the shank should be as long as possible. This gives more flexibility in the length of the screws. The 1/4 inch size I used is probably overkill. If I did it again I'd use the next size down and I'd also use a screw with a less prominent head.

The length of the screws is difficult to estimate. I bought mine longer than necessary then ground them down to the correct length. To find the correct length I attached the plate to the shoe with the over long screws then measured the excess by placing Allen wrenches of various sizes next to the portion of the screw sticking out inside the shoe. This told me how many mm needed to be ground off.

The metal plate is flat and the sole of the shoe is slightly curved. Something is needed to compensate for the difference in their shapes. First I tried using household caulk, but that broke down almost immediately. I was ultimately successful with double stick foam tape. The tape compensated for the slight rocker in the sole of the shoe and after some riding, I used some more tape on the outboard side of the plate to level the shoe on the pedal.



Insole
The bottom of the inside of the shoe is not too smoothly shaped after all this. I upgraded the insole by adding two layers of polypro and some reflective Mylar. This is less than perfect, but it has been adequate for rides up to four hours. I'm still trying to come up with an improvement for this. I've also been putting a small scrap of wadded up nylon sock in the toe of the shoe to fill in the extra space there.

Finishing touches
After riding a bit with an Allen wrench and a screwdriver, I was able to get the cleats adjusted just right. I applied copious amounts of reflective tape. The tape I used is DOT C2, the kind used on tractor-trailer bodies in the US. I followed the pattern of the leather on the shoe and it has been quite durable. Almost as important it covers some of the corporate logos.

Results
The shoes are very comfortable with two pairs of thick wool socks. There is room for chemical toe warmers, but I haven't had the need as we've hardly gotten below zero F here this year.

We haven't had much extreme cold here this winter, so I haven't put them to a real test, but they are a great improvement. Rides over a couple hours at 40 F used to leave me with numb toes. With these shoes I've had a couple of rides longer than that in the single digits Fahrenheit without any problem.

I ride primarily on road, but have also done some riding in the woods with these shoes. The shoes have had good traction in the snow. Despite the experience of many IceBikers, I haven't had any trouble with the SPD pedals icing up. In the worst conditions off-road a spinning kick to the pedal cleared up any clogging.

Amazingly, despite much riding on salted roads, the plate and the screw have shown no corrosion. I think the electrical cover plate that I used may be galvanized, but the edges aren't showing any corrosion either. There has been a bit of rust on the SPD cleat, but nothing serious.

The cleat does stick out on the bottom of the shoe, but the shoe is very walkable. It would only be a problem if I were somewhere where I was concerned about scratching the floor.

10:10 almost exactly

Watches in advertisements are usually shown with the minute and hour hands in roughly the 10:10 position. For the watch model below, I wanted to know exactly what time it was when the hands are in opposing positions.

I can't get an exact time because the answer appears to be an irrational number. Based on the results of the spreadsheet summarized below, the time is approximately 10:09:13.8461472.


time decimal 0.423076923 days

--------------------Hour--------Minute-----Second
--------------------hand---------hand-------hand

Units Indicated--10.15384615---9.23076912 13.8461472
Angle-----------304.6153846°--55.38461472°
Angle from vert.-55.38461544°-55.38461472°
Difference--------0.00000072°

Modeling and Rendering

Perhaps the best, certainly the most detailed SketchUp (SU) model I ever made is of my wrist-watch. In the past I’ve bought decent low priced ($30-$40) wrist watches and they last two or three years before completely failing. Three years ago I bought a titanium Seiko SNA139. Though the watch retailed at the time for about $350, I got a used one on ebay for about 1/4th of that. It still looks almost new. If it lasts another six years the purchase is justified financially. Either way I like it and I (mostly) don’t feel like spending that much on a luxury was wrong.

To the point, one day when I was home sick from work, I started a SketchUp (SU) model of the watch. I used a magnifying glass, calipers which measure to about two-thousands of an inch, and to read the smallest letters (just over 1/100th of an inch) on the inner edge of the dial, the eyes of a five-year-old. All together I put (wasted) about 12 hours creating the model. To the best of my knowledge, I captured every detail of the real watch. The SU model if this Seiko SNA139 is available in the 3d warehouse.

Google Androids Dreaming of Electric Sheep

Originally posted 4/14/09
I don’t usually think much about pop culture, especially things like viral videos sponsored by big corporations. BUT, I couldn’t help but admire the popular video of “Extreme Shepherding” by the "Baaa-Studs." Available here on YouTube.

My first question was how much of it was done with real sheep and how much with computers. The best non-answer I found comes from the 2009/03/24 NYT Technology Blog. They quoted from the agency which produced the video; "the team used some digital trickery to heighten the visual effects." That leaves a lot of wiggle room. I don't doubt that some impressive shepherding was involved, but I'm sure it's mostly CGI.

The agency which produced the video was The Viral Factory at the request of Samsung as a way of bringing attention to their LED based televisions (must have worked). So, it didn't really have anything to do with some Welshmen with too much time on their hands and a strange love of sheep and LEDs. It was a big time corporate advertising campaign which got all of the attention. In searching around, I see that very few people realize this. That's significant, but it's not the reason I wanted to write a blog posting.

What interested me was that when I first saw the video, I thought of the science fiction title "Do Androids Dream of Electric Sheep?". Then I saw that Samsung was involved and I remembered that Samsung was releasing a new phone to go with Google's Android software.

So, it IS:

Sheep > LEDs > Samsung TVs.

Is it ALSO:

Electric Sheep > Androids > Google Android > Samsung Cell Phones?

I often think I'm clever. It's good to be able to put together wide ranging ideas, but sometimes what I'm thinking has no relationship with reality, and that's not so clever.

SketchUp to Reality - bedside table

I use Google SketchUp (SU) to do a lot of doodling and planning. Sometimes I actually produce something which I have modeled. This is a nightstand I made for my daughter.


Besides basic assumptions for a bed side table, the design criteria were:

1. It had to fit stylistically with her present furniture and be neutral enough to fit with future furniture.

2. It needed to incorporate some bins.

3. It had to minimally block the vent which it will be on top of.

4. It had to be quick & easy to build.


After Googling around looking at bedside tables, I drew first on paper just to think about the design, then started in SU. I modeled the height of her bed and the window behind it to help get the size and proportions right. We purchased bins and I modeled those.

I used pocket screw joinery to assemble it. I bought a "Kreg Jig" system earlier this year and I have found it to be extremely useful. It is a quick and strong way of assembling things. It's not "fine woodworking," but not everything needs to be.

In addition to the model of the nightstand in place, I copied the components onto imaginary boards to optimize lumber usage. One of the "scenes" in the model shows the diagram below. I had planned on gluing up lumber to get 12" wide panels, but when I went to purchase the wood, lumber which would dress to 11 1/4" was available. I quickly updated the model and was good to go. The pine lumber was nominal 3/4" thick.

I used hand planes to prepare the wood, a guide and a circular saw for the crosscuts, and handsaws for the rips and curves. My router broke just before I started this project, so I was not able to use a router bit to round the edges. I did that with a block plane and sandpaper which resulted in a nice handmade look.

I sanded and scraped the pine and on first assembly, found that the wood was substantially less than 3/4" Using 1 1/4" Kreg screws for that size resulted in some of the screw points coming through the sides of the boards. I disassembled it, moved the boards slightly and replaced the screws with smaller ones.

I do not find finishing intuitive, but I am pleased with how this turned out. After I assembled the piece the first time, I disassembled it and finished it flat. Finishing individual pieces is much easier than working with completed furniture. The downside is that all the glue surfaces are covered with finish. If I were doing it again, I would mask the glue surfaces. Between the lack of glue and the smaller screws, I have some concern about structural integrity, but it feels solid enough.

I wanted to learn about shellac because I had never used it before. Shellac’s great fault is that it is soluble in alcohol, so you wouldn't want to use it anyplace where an alcoholic drink might be placed on it.

I used Zinsser Bulls Eye shellac. The shellac as it comes out of the can is very thick. Finishers refer to that thickness as a three-pound cut. A three-pound cut means - if the shellac had been made from flakes, as it used to be, it would have used three pounds of flakes mixed in one gallon of denatured alcohol. For the first coat, I added two units of hardware store denatured alcohol, making it a one-pound cut. Subsequent coats were a two-pound cut. I sanded with 400 grit to smooth the surface between coats. I also applied a gel stain between the 2nd and third layers.

After the four layers of shellac, I applied Trewax, Mahogany wax with 000 steel wool, then buffed it. Unfortunately, where I had put the wax on in a swirling pattern, there was a slight stain on the shellac. Repeating the waxing as below, eliminated almost all of the stains.

Next time, I will not leave as much wax on the surface while it is drying. It should be just enough to create a haze with no ridges. I would also apply the wax with the grain. The final finish has substantial depth.

The customer is happy and I'm pleased to be able to show it to friends. The model is available in the 3d warehouse here.

A Shaker Workbench - Hybrid Design

I'm building a workbench. The model above was produced in SketchUp. SketchUp is a relatively easy to learn, free, 3D modeling program. You can download the model here.

The workbench is a hybrid of what are called Shaker and Roubo designs. Shaker refers to the type of workbenches used by members of the Shaker religious community and the aesthetic style which they developed in their furniture. The other source of the design is André Jacob Roubo, a 16th century Frenchman who wrote a book on woodworking. A very heavy and solid style of bench takes its name from him.

Many of the ideas came from Christopher Schwarz's book and blog. Scott Landis's book was also very useful. Surfing the net and seeing other people's benches in the real world helped too. Although there is very little which is actually new in woodworking, I think a few of the ideas are original. originally posted 2008/05/30