Construction Summary

This document summarizes the methods of construction of the log home. Several factors are responsible for these methods being quite different from those of most builders. These factors included 1) The site is an irregular rock, 2) the house is a log house built using Skip Ellsworth's methods, 3) my desire to do as much of the work as possible by myself, 4) my preference for hand tools over power tools wherever possible, 5) no time pressure to finish the project before any certain date, and 6) the availability of the particular tools and materials that I happened to have. These factors should help you understand why I made the choices I did as I built. Here's how I did it.


The site is on a ledge of a big granite outcropping. The ledge was half filled (0 inches deep at the low end and about 5 feet deep at the high end) with rocks and dirt. It was mostly rocks with dirt only in the spaces between the rocks. The rocks varied in size from pebbles to some that probably weighed 500 lb. I wanted to separate the dirt from the rocks, so I could use the rocks to build a foundation and the dirt to make the upper roadway. I excavated the whole thing by hand with a pick, shovel, and a big steel bar. I wheelbarrowed the dirt down the hill to make my road and I stacked the rocks inside the site for future use in the foundation.

Once I had the site as flat as I could get it, I determined the size, shape, and location of the eventual building. Then I dug the foundation trench down to bedrock. In some places the bedrock was already at the surface. In the deepest place, the trench was about 18” deep.

Next I built forms for the footings. I had to have an engineer design the footings because the building department told me that they knew how to build a foundation on dirt but they weren’t sure one could be built on solid granite. The engineer had me build them 18” wide and a minimum of 10” deep.

I decided to lay up a concrete block foundation rather than pour a continuous concrete foundation. That way, I could do the work all by myself on an intermittent schedule. I planned the footing for the blocks so they would come out exactly to the half-block. I stepped the forms to match this plan. This added a lot of complexity and time that most other people wouldn’t do. I had to place two #5 rebar continuous all around the bottom of the footing forms, and I placed a vertical #5 rebar every 4 feet. These rebar pieces stuck up above the forms and they had a six inch 90 degree bend that was embedded six inches or so into the footing. I also planned ahead for a couple of penetrations through the footings. These were pipes for the water supply line, and for drainage at the lowest point in the footing.

When the forms and the rebar were ready, I had a ready-mix concrete truck and a big pumper come to pour the concrete. The pumper had a 42 meter boom and was able to reach my foundation by parking down below the cliff. Before the concrete was set, all the vertical rebar were straightened up, the threshold for the door was set in the footing, and the CB88s were placed in the column pads.

Before starting the block laying, I wired a piece of #5 rebar to the vertical rebars sticking out of the footing that weren't already long enough to reach the first bond beam level.. These rebar were carefully planned so that they would run through the cores of the blocks as they were laid up. The 2x10 door frame was also placed on the footing so the blocks could be laid up against it on both sides. The door frame had 10" galvanized flashing attached to the top and sides to protect it from moisture from the blocks, and a length of 1/2" EMT was nailed vertically to the center of both 2x10s. This conduit went into the slot in the center of the concrete blocks that mated with the door frame and that keeps the frame in place.

To lay the blocks, I built up what are called ‘leads’ in each corner (actually it was only two corners where the wall was going to be 8 feet high. The other two corners would only be one block high.) I used a story pole at each of the two high corners with marks on them to make sure each course was at exactly the right height, and each time I laid a block on the corner, I used a plumb bob to make sure the outside corner was exactly plumb.

Once the first corner blocks were in place, I snapped a chalk line on the footing between the blocks to mark where the first course of blocks should go. For the rest of the courses, a tight mason’s string was stretched between the leads and the blocks between the leads were kept straight by the string. The blocks that had the vertical rebar running through them, had to be lifted up and down over the rebar. This lifting was extra work, but it got less and less as the courses went up.

I placed K-web in the joints between every other course of blocks. I also placed wall ties under each block that stuck out of the wall for tying into the rock veneer I will later lay up on the outside of the block wall.

At two courses below the four foot level, I wired the next 4 foot sections of vertical rebar to the ones below. That had to be done while there was still the overlap amount sticking up.

At the four foot level, I laid a course of special blocks to make a bond beam. These have notches cut out of the tops of each end. That way, there is a channel that can hold two horizontal #5 rebar which makes the bond beam. Before placing the rebar, I filled all the cores with concrete up to the bottom of the channel. I made sure I tamped the concrete into the cores good with a piece of rebar. Then with all the cores filled, I placed the two #5 rebar in the channel and then filled the channel with concrete and leveled the top to make a base for the next course of blocks.

One trick I did that paid off big time for me was when I placed the bond-beam rebar, I also placed a U-shaped piece of rebar in the same seam at each corner of the foundation so that the U stuck out of the wall about 3 or 4 inches. These provided super strong anchor points for attaching my log moving rigging to later on. I used them a lot and I considered them indispensable.

From that bond-beam level, it was pretty much the same process as it was from the footing itself. I laid up leads at each corner and then laid up the blocks between the leads up to a second bond-beam that ran completely around the foundation. There were other details like placing pipes, conduits, vents, etc. through the foundation, and the door frame, the vent openings, and two girder pockets. The vertical rebar was switched from #5 to #4 to stick out the top of the foundation and go through the first course of logs as Skip taught us. A layer of felt paper, foam sill sealer, and the pressure treated sill were threaded down over these rebar and then the logs went on top of that.


The logs were placed in the walls using the butt and pass method. At each corner and at each course, one log would butt up against the other and the other would stick out of the building 2 feet. That end of both logs would either be butts or tops. Butts and tops alternate with each course as did the direction of which log butted and which one passed.

The logs were fastened by steel dowels made of square-cut #4 rebar driven vertically into each log every 20 inches along its length in the wall. These dowels not only keep the logs in their proper positions, but they also support the entire weight of the building by transmitting the load forces down to the foundation. This is the reason that log houses built in this manner do not settle as a result of log shrinkage. When the logs shrink, they shrink about their own center lines, and these center lines maintain their exact positions in the original geometry of the building. This fact greatly simplifies construction compared to other log construction methods in that there is no need for slip joints, jacks, and other devices necessary for vertical members like columns, windows, cabinets, stairs, plumbing runs, etc.

The dowels were driven into a half-inch vertical hole bored in the log being placed and from there into the log directly below with no pre-drilled hole in the second log. The dowels were cut so that they reached to the center of the second log. At the corners, a dowel was also driven horizontally through the passing log and six inches into the end grain of the butting log.

Prior to placing a log, it was planed down using a Log Wizard, which I affectionately called my "Gwizard". This tool is a 3 1/4 inch planer mounted on the end of a chainsaw bar. The tool removes bark, discolored and even rotten wood, and leaves the log with a drawknife-looking finish. When this was done, the entire log was sprayed with a couple of applications of a water-soluble borate solution to protect the wood against insects, fungi, and bacteria.

I placed 2x10 door frames and window frames in the walls as I went. For the window frames, I would chisel a flat surface on the log which would receive the bottom of the frame and then nail the frame down onto that surface after caulking in between them. Logs that met the window or door frames were cut square on the end and fastened to the frame with 3 60d galvanized nails. These joints were caulked later after the walls were built.

To lift the logs, I built a crane which was eventually fastened to the back log wall and was incrementally moved up as the building grew. To reach the walls, I built a scaffold system on the inside of the building which also was moved up as the walls grew. In addition to the scaffold system, I built and used a portable platform which hung from a bracket of 3/4" EMT that could be hung over a log, either a log in the wall, or one of the projecting logs at a corner. This platform was used mostly to give me a place to stand in order to drive the horizontal rebar pins into the corners to pin the butt and pass joints.


The roof is supported by a log ridgepole and six log purlins, two on one side and four on the other side of the ridge. Each purlin is supported by three log columns, one near the center of the building, one just inside or just outside a gable wall. Two of the purlins are outside the building and support the portion of the roof that covers the porch deck. Each of these is also supported by three columns. In cases where columns intersect beams, the column is interrupted with the beam being continuous. The beam is flattened above and below to provide a bearing surface for the square-cut ends of the column segments.

Each column rests on a concrete pad or pier which was poured on clean, solid bedrock. The pads for the RPSLs and the PSLs along the Grid 1 and 3 walls are integrated into the footings and were poured at the same time. The piers for the columns inside the building were poured separately. Each pad was sized for the load it would bear and a suitable grid of rebar was incorporated in the bottom of each pad or pier. A CB88 was embedded in each pad as it was poured and serves to fasten the column to the pad.


The roof is constructed using 11 7/8" TJI rafters 16 inches on center. This is sheathed on top using 3/4" OSB, with tar paper and finally 16 inch wide steel roofing panels on top of that. R38 insulation will be placed between the rafters from the inside, and then 1x8 T & G pine boards will be nailed to the undersides of the rafters to form the ceiling. Prior to placing the rafters on the purlins, one of these pine boards was nailed to the flattened top of each purlin and the rafters were then placed on the tops of those boards. That way, the ceiling boards can mate with the tongues and grooves of those boards already in place.

A 32 inch square hole was framed into the rafters near the ridge to accommodate an eventual masonry chimney. A short rafter was screwed into this hole and the sheathing and roofing covered the hole. A metal chimney was installed in this space. The metal chimney will serve a wood stove until a masonry chimney is built, which may or may not ever happen.


The subfloors are constructed using TJI joists 16 inches on center with 3/4 inch T & G plywood glued and nailed or screwed to the joists. The joists are supported by beams in the center of the building and rim joists on the Grid 1 wall, both for the main floor and for the loft floor. On the Grid 3 wall, the main floor is supported by a series of double 2x10 beams supported by beam hangers attached to the three columns next to that wall, and by special steel straps on each end.

A hole was framed into the main floor above a concrete footing designed to support an eventual masonry stove or fireplace. Face mounted joist hangers were installed on two sides of this hole and TJI joists were installed on these hangers. The hole was covered with plywood to make a continuous floor to serve until such time as a masonry structure is built.

A circular hole was cut in the floor directly under where a wood stove is installed for the purpose of supplying combustion air to the stove. The hole is plugged with mortar but can be unplugged by removing screws from the underside of the floor. It isn't clear whether or not this hole will be necessary or helpful, but if it turns out to be, then it is ready to go.

The plywood subfloor is continuous under the thresholds of both the front and back doors. That exposes the edge of the plywood to the elements on the outside of the building. It is no problem on the front because the big porch protects the plywood from moisture. But at the back door, snow drifts onto threshold of the door frame, and wind-blown rain can also reach the threshold. After seeing early signs of water damage to the subfloor under the back door, aluminum flashing was installed to keep the water from reaching the plywood subfloor. A narrow lip of the metal engages a slot near the top of the inner section of the aluminum threshold. From there, the flashing runs straight down to the top of the subfloor, makes a right=angle bend so it runs over the top of the subfloor to its outer edge, and from there it makes another right-angle bend and runs down over the top of the sill log. The other three parts making up the threshold were replaced over the top of the flashing once it was in place.


The porch structure is supported by a 6x8 pressure treated ledger bolted against the Grid E foundation wall and three log beams. The log beams are supported by log columns resting on concrete piers each poured on solid bedrock. The columns are fastened to the piers with CB88s. There is a full-length beam at Grid F, a beam running between Grid G1 and G2, and a beam running between Grid F.5,2 and F.5,3. These beams and ledger support a series of log joists on 2-foot centers. The deck, consisting of wide hand-hewn planks of 2 to 3 inch thickness is screwed to the joist system in counter-bored holes. The holes are then plugged with wooden dowels. Columns supporting the roof over the porch rest on bearing surfaces cut into the Grid F and G beams. Each such joint is fastened by a vertical length of #4 rebar which runs through the beam and penetrates the end grain of the columns above and below to a depth of 6 inches.

A log staircase provides access to the porch. Two log stringers rest on concrete abutments at the bottom and a thick plank header fastened to a log joist and beams at the top. Half, or third, log treads are fastened to notches in these stringers with lag screws. The stringers, and a log newel post are fastened to the concrete abutments using CB66s.


The porch structure is supported by two log columns fastened using CB66s to concrete pads poured on bedrock The columns extend high enough to support the guard rail for the porch. Two shorter log beams span the space between the columns and the Grid A sill log. These beams are fastened to the columns and sill log with 4x4x4 inch mortise and tenon joints. The joints are secured by pins made of #3 rebar running through the joint.

The deck is made of four half-logs ripped down the middle and installed flat side up notched into the short beams and pinned with lengths of #3 rebar. rough

The staircase is supported by two log stringers which rest against a short beam at the top and on concrete abutments at the bottom using CB66s. These abutments are, of course, poured on bedrock. The stair treads are log slabs fastened to notches in the stringers with lag screws.


The rough window opening frames were made of off the shelf 2x10s and which were installed as the log walls grew up around them. Milgard vinyl windows were installed in each of these openings in the standard way. A shim on the bottom supported the window in the correct position, then the flange on the window was caulked and screwed to the edges of the 2x10 frames. The trim of 1x2s was then nailed to the frame over the top of the flange.

The outside 1x2 trim was stained to match the exterior log walls. The inside was finished by scraping and varnishing the inside surfaces of the 2x10 frames and then mitering and installing cove molding between the vinyl window and the frame. This presents a more-or-less gradual and pleasing transition from the factory finish of the window, to the nicely finished cove molding, to the more rustic 2x10 frame, to the very rustic and checked log walls.

The installation from this standpoint was simple and standard. However, all window installation was done by me without any help other than some help carrying some of the windows up the roadway to the building. Installing the big windows on the front of the building, particularly the high ones, presented a challenge. My method was to set up sufficient scaffolding in order to gain access to the window openings, and then to use rigging attached to anchor hooks on the foundation wall and the Grid B, C, and D purlins in order to lift, position, and hold the windows so that I could align, shim, trim, caulk, screw, and trim the windows.


The exterior log walls were finished in a 6 step process of preparing, spraying, staining, insulating, nailing, and chinking. Access to the walls for this work was provided in most cases by my scaffold system which in this case hung on the outside of the wall rather than on the inside.

Preparing the wood involved removing the discoloration that was mostly due to UV damage during the long interval between erecting the wall and the time I finished them. This discoloration was removed by a combination of a power planer, scrapers, a big gouge, chisels, rasps, sandpaper, and even a pocketknife. Even though the damaged wood was very thin and near the surface, there was a very visible difference once it was removed.

Once the wood was bright again, a borate solution was sprayed on to protect against mildew, fungus, bacteria, and insect damage. The surfaces were then stained with TWP. Next, strips of unbacked fiberglass insulation were stuffed in the spaces between the log courses. This not only provides insulation to conserve heat, but it provides a wick which will allow any moisture that condenses inside the walls to make its way to the outside where it can evaporate.

With the insulation in place, a galvanized nail is driven into the lower log in each seam up against the insulation. These nails stick up an inch or two and are bent up so that they end up vertical and up against the insulation. These nails provide the mechanical support necessary to hold the chinking in place.

The final step is to chink the seams using brick mortar. The mortar is troweled into the seams up against the insulation with a nice tight interface to the log surfaces on the top and the bottom. The bottom is not much of a problem, but it takes a delicate touch with the tip of a small trowel to press the top margin of the mortar up against the top log without having is slump away and form a crack between the mortar and the log. It is essential to have the mortar consistency just right and it helps to have it consistent between batches. For this reason it works best to carefully measure out the water and the mortar mix to get it just right.

When the mortar cures, it will form vertical hair-line cracks every 6 to 8 inches or so. That means that there are two or three nails in each chunk of mortar which will hold the mortar in place indefinitely.


The interior partition walls are standard wood 2x4 stud walls covered with drywall.

The interior log walls were finished in a similar manner to the exterior log walls. On the interior walls, the step of applying borate was skipped. This is because the logs had previously been treated with borate and being inside the building, they were no longer so vulnerable to insects or other deteriorating agents.

Another difference is that the interior walls were varnished with three coats of Varathane high-gloss polyurethane rather than the TWP that was used on the outside.

The insulation, nailing, and chinking on the inside was done exactly the same way as it was on the outside.


The electrical installation in this project differs from a standard installation in two major respects. Wiring in log walls is different from wiring in frame walls, and the unusual site of the building being high on a granite rock some 150 feet away from the transformer in the street presented some unusual problems.

For the most part, the wiring runs ran in the two floors, the roof rafters, and the partition stud walls. These runs were done in the standard way. There were, however, three cases of where wiring was done in the logs or the log walls.

The first case was the installation of baseboard receptacles in the log walls. This was done using metal boxes recessed in a space between two logs so that the cover plate just touched the log surface above and below. These boxes were connected with 3/4" EMT conduit which ran nestled in the space between the two logs. The galvanized nails for the chinking were typically longer for these seams so that they could reach up and over the EMT. The mortar in these seams is a little wider than typical seams. During chinking, the mortar was troweled up against the vertical edges of the boxes so that the cover plates fit flat against the chinking surface.

The second case was the problem of running wires from inside the loft floor to inside the roof rafters. I had not made vertical raceways in the log walls for this purpose when I built the walls (a mistake in hindsight that I wouldn't make the next time) so there was no unobtrusive way of running those wires except in the corners of the walls. Again, I had not thought of this problem when I built those corners, so I had to deal with them the way they were built. In the Grid E1 corner, the joints just happened to have gaps between the logs that allowed wires to be tucked into the joints with no problem. The Grid A1 corner was more difficult.

In both corners, there was the problem of running the wires up through the cap log, which is the top log in the non-gable wall. The cap logs are purlins, so they run all the way outside to the eaves and don't provide any way of getting a wire around or through them. With considerable work, I was able to drill holes through these cap logs and make wire runs that could be buried under the insulation and chinking that would go over them.

The third case was the installation of fixtures on the logs. These are things like ceiling fixtures mounted on the undersides of purlins, flood lights mounted on the outside of wall logs, and switches and light fixtures mounted on columns.

To mount a fixture to a log, after doing it the hard way with hammer and chisel a couple of times, I used a 6-inch hole saw to make a flat surface for the fixture base. When the hole saw started, the kerf would start on the two high points of the log and as the saw went deeper, the two kerfs would widen and curve toward each other. When the saw was deep enough, the kerfs would meet and I would stop at that point.

Then using a hammer and chisel, I would remove the wood from inside the circular kerf making a more-or-less flat circular surface for the fixture. This wasn't flattened very precisely because the next step was to switch to a 4-inch hole saw and using the same center guide hole, cut a hole for the pancake electrical box. The one-inch margin around this hole was then flattened more carefully to provide a flat surface for the fixture base.

Next, the pancake box was inserted in the hole to make sure it fit right. Once it fit right, the knockout for the wire was chosen and removed, and the green screw was screwed all the way into the box. Then with the box back in the hole, the sites of the knockout and the green screw were marked on the wood.

A 1 1/4" spade bit was used to bore a hole for the cable clamp and a smaller bit to bore a hole to accommodate the green screw.

Next, a long augur was used to drill from the cable clamp recess all the way through the log for the wire. The wood was then cleaned up with a chisel, the edges of the big hole were chamfered, and finally, the newly cut wood was stained or varnished to match the log surface.

The electrical box and fixture were then wired up in the normal way and installed in the log.

To install a switch in a log column, I first chiseled out the rectangular hole to hold the box. Then, using a long augur, I drilled a hole from the inside lip of the rectangular hole down at the steepest angle I could through the log so that the augur emerged from the log under the floor level. This required removing the bit very frequently to remove the chips, otherwise the chips building up behind the augur head would have plugged the hole making it impossible to remove the bit.

Some additional chiseling was required inside the hole to accommodate the clamp and so that the box could be inserted into the hole with the wire clamped in place. Then it was wired and closed in the normal way.

Providing electrical service to my site on that granite cliff required some expert help from a good friend. Since the building is more than 150 feet away from the transformer, it presents special problems for how to install the service.

The first question is where the meter should be located, and, in my case, the next question is what the route of the wires should be. The solution I chose was to install a meter-disconnect down by the road very near to the transformer. Then run a feeder line from there to the service panel in the building. Since that feeder line is protected by a fused disconnect down by the transformer, the feeder line does not have to be buried as deep as an unfused primary line. Also, the distance between my service panel and where the feeder entered the building was not limited because, again, the line was fused. In my case, that run inside the building was 30 feet or so and wouldn't be allowed for a primary line.

Another problem, which was the biggest one of all, was that for about 60 feet of the feeder line, the run went over solid granite bedrock at or near the surface so that a trench was not a good option. The PUD informed me that I could lay the conduit over the rock and then vault it with 2 inches of concrete around the conduit. My son Bill suggested that I might as well build a concrete staircase over the top of the conduit since I had to pour concrete up and down that cliff anyway. I took his advice and excavated the run down to bedrock, removing many very large roots in the process. Then I built forms for the staircase and got some wonderful help from sons Bill and Dave, brother John, Dave's friend Jim, neighbor Kerry, and John's and my childhood friend Harold who built an impressive staircase for me.

Dave also gave me some good advice. He suggested installing a wiring closet for centrally locating all low voltage and computerized applications in a central place. I chose to install it at the top of the pantry, at Dave's suggestion. For this reason, I installed a set of 120v receptacles up there for whatever devices might need it. I also installed a 2" conduit raceway from the wiring closet down through the floor to the crawlspace for routing low voltage and communication wires.

Except for these exceptional cases, the electrical installation is fairly standard.


(Still in the works)


A wood burning stove is installed in the center of the main floor and it will serve as the secondary heat source. The primary heat source will be some type of electric heating system, yet to be determined.

Go To Home Page

©2009 Paul R. Martin, All rights reserved.