The finished shelf unit in place.

I originally wanted to do it in MDF with PVC pipe or galvanized pipe, but my wife thought it was too raw. Here’s the prototype I whipped up of that concept using about $15 worth of Home Depot parts. (I used white “Sharkbite” PET tubing and particleboard shelving, with plumbing strapping on the diagonal in back for stiffening.) I kind of like it as is, and in fact, used it for a year in the house with no complaints.

Prototype shelf made from Home Depot parts.

To accommodate the need for something more refined, I designed a version that uses stainless steel tubing and baltic birch plywood. I had the baltic birch cut by Ponoko, using their CNC routing process. I’m including all the details here so if you want to replicate this shelf you can easily do so. The parts are not expensive. The routed plywood parts cost $100 or so depending on how many you make and the freight. Someone with patience and skill could easily cut those parts in a home shop. I like the precision of the routed parts and I very much like having them just show up via UPS ready to assemble.

Details:

The unit has four plywood parts. The plywood is 0.740″ (18mm) baltic birch, one of the standard Ponoko materials.

(2) top and bottom shelves (same part used twice)

(1) middle shelf

(1) back piece

These parts are shown here.

Parts come bubble wrapped from Ponoko and ship via UPS.

“

Parts unpacked (from left) back piece, middle shelf, top/bottom shelf.

Then, you need some hardware.

(4) 2-1/2″ Casters with 5/16″-18  threaded stud. (McMaster-Carr)

(4) 5/16″-18 x 2″  zinc-plated or stainless steel flat-head cap screw (McMaster-Carr)

(8) 5/16″-18 threaded “tube connector” for tube ID 0.620-0.652″ (McMaster-Carr)

(4) 5/16″ ID flat washer (McMaster-Carr)

(2) 1-1/2″ self-tapping screw

Hardware from McMaster-Carr.

You also need some tubing. I used 3/4″ OD stainless tubing from McMaster-Carr. You could use any 3/4″ tubing. If you use the dimensions in my drawings, then you cut this tubing to be 17-7/8″ long. This allows you to get exactly 4 pieces out of one length of 6 foot tubing, including an allowance for the kerf of the hacksaw.

3/4" OD stainless tubing

Assembly is pretty easy. Start by threading a 5/16  cap screw into a tube insert. Then tap the insert into the end of the tube. Tap it in just barely past the end or else the stud on the caster won’t engage the thread fully.

Tap in thread insert after threading a screw in. Only tap it in just past the end of the tube.

With the inserts in each end of the four lengths of tubing, you can assemble. The top side is retained with the flat-head screws. You need to chamfer the hole to allow the head of the screw to be flush.

Hold top shelf in place with cap screws. Chamfer holes first.

On the bottom side, the casters themselves hold the assembly in place. You just thread the caster in, putting a washer between the caster and the plywood.

Thread caster into tube from bottom side, including a washer as shown.

The back edge of the middle shelf is held in place by the back piece. However, the front edge is not supported. So, you insert a 1-1/2″ screw in from the sides and into the tubing. You need to pre-drill. Measure to be sure the shelf spacing is the same front and back.

Insert 1-1/2" self-tapping screw from side to position front edge of the middle shelf.

The assembled units look like this, ready for finishing.

Units ready for finishing.

The first of these I built, I finished the plywood before assembly. That’s probably a good way to do it. However, I hate flipping the parts over during finishing. So, the second batch I assembled first and then finished. You need to mask the tubing to prevent it from being messed up by whatever finish you use. I use Minwax Wipe-On Polyurethane. I put a coat on first, before any sanding. Then, I sand the edges and faces lightly. Be careful as the edges will splinter and there is a “nap” to the surface as well. I use a 220 grit sanding block.

Then, apply a bunch of coats. It wipes on easily. I probably apply about 4-5 light coats. (In Utah where I was finishing these, the coats dry really fast, under an hour.) I sand lightly with a 220 grit sanding block between coats.

Shelf unit masked and ready for Minwax wipe-on polyurethane.

Finished shelves…

Finished unit.

Edge detail.

Joint between back piece and bottom shelf.

Back piece has two notches that locate the back edge of the middle shelf.

Files for Ponoko

(You need two top/bottom shelves and a middle shelf plus the back piece. These files include a mix of different components to make a total of three shelf units, I think. I also filled the empty space with rectangles as you pay for the whole sheet of material and a flat rate for freight.) These are .eps files, so you’ll need Illustrator or Inkscape (or equivalent graphics program) to open and edit these.

File-A

File-B

Please note that the first time I ran these parts, I specified the holes for the tubing to be exactly 0.750″ diameter. Inexplicably, they came back from Ponoko at 0.730. Ponoko had some story about cutter speed that doesn’t really make any sense to me…every other dimension was dead on. Still, I simply increased the specified diameter by 0.020″ on the second batch, and holes did indeed come back at 0.750 the second time. It is possible this is some idiosyncratic feature of the vendor Ponoko uses, and that this won’t work exactly right in the future.

These notes explain why, fundamentally, this conversion is beneficial. This is for the financially and mathematically sophisticated investor. If you don’t even know what a Roth IRA is, then go read the Motley Fool guide or something.

First, consider the very basic situation is which your tax rate today is the same as your tax rate in the future. For concreteness, assume your total tax rate is 40% (assuming for instance, a 35% federal rate and a 5% state rate and a state with policy similar to the federal). Assume your total long-term capital gains tax is 25%, now and in the future.

Assume also that you have $100,000, which were contributed with pre-tax funds, in a conventional IRA.

Assume also that you have $40,000 in after-tax funds available for investment. These funds are a reserve you have set aside to pay taxes.

If you stick with the conventional IRA, you leave these investments where they are until liquidation. Assume that you invest in a way that results in any investment increasing by a factor of 10, all very tax efficiently (e.g., all long-term capital gain). A 10x return would, for instance, correspond approximately to an 8% annual return for 30 years.

At liquidation, you would have, before paying taxes, $1,000,000 in the IRA and $400,000 in the after-tax side investment. Then, you pay the taxes and you are left with $600,000 from the IRA, and you’re left with $310,000 from the after-tax side investment. (You lose $90,000 because you pay a 25% long-term capital gain tax on the $360,000 in gain you got on that side investment). Your total after-tax proceeds are $910,000.

Now, consider what would have happened if you had done the Roth conversion.

First, you would send the $40,000 in after-tax funds to the tax collector. (This is the painful part.) Then, you would invest the $100,000 tax free. At liquidation you would have $1,000,000 with no tax due. Compare this to the $910,000 you would have with the conventional IRA.

Thus, in this example the Roth is about 10% better even when you assume the tax rate today and in the future are the same. However, this is only true if you have after-tax funds available to pay your taxes now. If you pay taxes from the pre-tax funds, the investments come out exactly the same.

Huh? How can it matter whether you pay the taxes from the proceeds of the IRA or from after-tax dollars? The reason it matters is that in the first scenario — staying with the conventional IRA — even though you have set aside 40% of the value of your IRA to pay future taxes, you will pay a capital gains tax on the growth in those funds that allows them to remain at 40% of the value of the IRA. That will reduce your value by $90,000 in the future. However, in the case of the Roth conversion, you pay only the $40,000 in taxes, and essentially you get the benefit of the future appreciated value of these funds (which have already been paid to the IRS) tax free.

Of course these factors make the Roth even more attractive:

- Any increase in future tax rates.

- The benefits of no-minimum distributions, as exist with conventional IRAs, and the ability to bequeath the funds to heirs tax free.

These factors would diminish the value of the Roth conversion:

- Lower future tax rates, as could occur if the U.S. went to a VAT tax or federal sales tax, or if you believe you will fall to a lower tax bracket with lower marginal rates.

My 13-month old Bosch Vision 300 Series Washer displayed error code E:27 (E27). Of course, I was just out of warranty.

This error code is not in the user manual, but some research on line indicates that it is a door latch fault. The problem is that the washer will not operate and the door is latched closed.

You can manually release the latch by removing the lower panel (Torx fasteners, I think) and finding a wire lanyard that releases the latch. However, I had read that there was a diagnostic code that could be used to reset the latch solenoid via the control panel.

There is no service manual available on line, as far as I can tell.

Here is what worked…

1. Set the knob to off.

2. Hold down both the Spin Selector button and the Allergy Rinse button simultaneously.

3. While holding these two buttons down, rotate the knob one click to the left, to the “cold” position.

4. At that point, the display will show a diagnostic code and the unit will enter a self-test mode. In my case, the solenoid latch started clicking repeatedly. I pressed the Spin Selector button once or twice, which resulted in the display of new codes and then the release of the latch.

5. I set the knob back to off, and the washer now appears to be operating.

Last Winter we had a pipe freeze while we were on vacation and this level of the house flooded…thus the floor covering (among other items) had to be replaced. I decided to use FLOR again. In six years, the system has been improved significantly. Originally, the carpet tiles were adhered to the floor beneath with double-sticky “dots.” This didn’t work very well for two reasons. First, the dots didn’t hold the carpet down very well, and secondly when the dots did adhere, they pulled up the floor finish when the tiles were removed. The new system adheres the tiles to each other. You simply place the new dots, with single-sided adhesive, at the corners of the tiles with the sticky side up, and join the four corners of adjacent tiles together. This is a much better approach. The adhesive is formulated to work with the tile. The tiles can float a bit with respect to the floor, which make installation much less fussy. My 14-year-old son and I  laid down 63 tiles (7×9) in about an hour. Following are the photos. This particular tile is “Rake me over” in sunny, butter, red, and pewter. The teenagers are pretty happy with the results. The cost of materials for this 11 ft. x 14 ft. area was about $1000…in IKEA range for covering that much area.

Snap two perpendicular chalklines (teaching 3-4-5 triangles along the way) and lay down a "cross."

Work out from the corners using dots to adhere tiles to each other.

Goes really fast...

Put the teenager to work vacuuming the loose strands of carpet.

Nice look and feel for the boys' TV room.

IKEA sells LED lighting strips with power supplies (“Ledberg”). These are roughly 24 inches long and they have a modular connector system so several can be ganged together. I simply mounted these to the bottom of the vanity about 3 inches back from the front edge with the cord fed through a hole in the bottom of the vanity.

This was easy and inexpensive. The light color is a little cooler than I’d like, but overall I’m pretty happy with the results.

Ledberg IKEA LED lighting strips. Note hole drilled in bottom of vanity with 11/16 butterfly bit.

LED strip mounted to bottom of vanity with a couple of little screws (provided).

Power supply and most of cord sits inside vanity.

"Nightlight" effect from below vanity.

Amenity Hut from back side showing outdoor sink.

It’s so nice to have a place to wash hands and brush teeth right next to the Pavilion. We placed a Katadyn gravity-fed water filter next to the sink to provide drinking water there, as the lake water can present giardia risks.

Amenity Hut from lake side, showing shower enclosure.

We get a lot of sun, so I figured we could do a crude solar water heater on site. I found a black steel pressure tank intended for storage of compressed air for tractor trailer braking systems. The plumber installed this in line with the shower (we split the line into a cold and “hot” side and ran them to a conventional Moen shower valve). This sort of works. There really isn’t enough surface area here relative to the 2o or so gallons of water in the tank to get the water very warm. Still, the lake water runs 65-70F and I think the “hot” side of the shower was about 80F. That makes a pretty big difference. The shower at that temperature is very nice on a warm summer afternoon. The less hardy members of the family supplement with one of those black plastic solar shower bags, which get up to about 105F. I think with some fussing I could get this to work better, but it’s not a bad first try.

Solar water heater plumbed in-line to shower...a black steel air pressure tank for a tractor trailer braking system.

I’ve spent a fair bit of time over the last two years researching composting toilet systems. I’ve discussed them with friends and have done some experiments at my cabin in Vermont. I’ve also had almost 10 years of experience with a commercial system (Envirolet) at my Vermont cabin.

Commercial composting systems are really set up for steady demand by a few people. Our use pattern is very different. We hit the system very hard for about four weeks a year and then it sits idle for the other 48 weeks.

Furthermore, most commercially available composting systems benefit substantially from electric heaters built into them. These use a fair bit of energy. My system in Vermont is not powered and it does not work very well, by which I mean the waste takes at least a year to decompose. I’ve concluded that we wouldn’t be happy with the performance of a conventional composting system at the Montana sleeping pavilion.

Instead, I’ve installed a “Lovable Loo” which I’m using in an unusual way…anaerobic composting instead of aerobic composting.

The Lovable Loo is basically a housing for a 5-gallon plastic bucket with a toilet seat on it. After each use, the user scoops a generous amount of saw dust or peat moss over the waste in the bucket. (We use peat moss because it’s easy to buy at the home center.) This essentially eliminates odors and provides a more attractive appearance than the alternative. When the bucket is full, the lid to the box is lifted, and the bucket lifted out and replaced with a new one. There are thousands of these systems in use and they are apparently used very successfully even in conditioned indoor spaces.

Lovable Loo installed in the "amenity hut"

Nearly all Lovable Loo users adopt an aerobic composting system. Basically they have a wire/wood bin on their lot somewhere and they empty the bins of waste into the bin to compost. If the mix of organic material, water, and air is maintained appropriately, these compost piles are efficient and fast at decomposing the waste. Aerobic composting does not smell, as the product of decomposition is largely C02, an odorless gas. The temperatures of composting are high enough that dangerous microbes are killed. However, this would not work on our site. First, I’m not wild about disturbing the landscape with a sizable composting bin. Second, I am also worried about maintaining the right conditions for composting when unattended.

Instead we are trying anaerobic composting, which is composting in a sealed chamber without oxygen. (Basically there are two kinds of bacteria…aerobic and anaerobic. We’re using the anaerobic bacteria.) Here’s how it works. Each bucket of waste is capped tightly with the standard bucket lids (which are extremely secure…darn near impossible to get off without a tool). We mark the bucket with the date using a Sharpie. We then simply let the contents decompose anaerobically for at least one season. (It might take two, but that remains to be determined…report in Summer 2012) Once decomposed, the contents (which should be essentially soil and water) can be returned to the land. I punched a small hole in the lid (1/16″ diameter) to prevent any internal pressure from the build up of biogas (mostly methane). Incidentally, the holding tank of a septic system is basically an anaerobic composter…waste settles to the bottom and decomposes anaerobically. This is also how most sewage treatment plants work.

The Lovable Loo system is relatively inexpensive, does not require any permanent fixtures in the hut, and thus can be abandoned at low cost if we don’t like it. However, if it works, it scales nicely to fewer/more people and is a pretty reliable low-fuss solution. Details on the Lovable Loo here. I also found the Humanure Handbook, written by the hippies who created the Lovable Loo, strangely interesting.

In our first season, which comprised an average of three users for four weeks (12 person-weeks) we filled five buckets. We adopted a protocol in which male users (and willing females) eliminated liquid wastes the way the other large mammals on our island do…this minimizes the volume of waste the system has to handle. After the first season, I can attest that we had no odors and the use experience was quite good, even for kids and city slickers. This has been a very good solution for our use case. I’m hopeful that the latency required for composting is less than one year, in which case we’ll have only five or so buckets digesting at any one time. If it takes two years, we’ll have ten, which would be manageable. (Incidentally, the optimal moisture level for anaerobic composting is very high…essentially you want to fill the bucket with as much water as it will hold before sealing.) More a year from now…

Significant settling of paver surface from water draining down through it.

Gutter assembly flipped upside down. 3 inch c-channel in center, angles on edges, 8 inch plates on top.

Removed pavers to allow for 16-inch wide gutter assembly to be recessed, protruding off the edge by 1-2 inches.

Recessed gutter installed flush with surface of pavers.

Installed gutter system. I hung a chain from the awning gutter to direct the flow of water and avoid as much splashing as possible. Works pretty well.

I re-laid the pavers in the areas where the terrace had settled. This time I used polymeric sand to fill the joints. (As best I can tell, polymeric sand is a mixture of fine sand and a water-curing glue…it feels like Elmer’s glue and sand when wet, but sets up to seal the joints. It seems to work fairly well. It may be belts-and-suspenders…we’ll know next spring.)