Fixing a 2-1/2 HP Ottawa Gas Engine

Peter Rooke fixes the round fuel tank, Webster magneto and more on his 2-1/2 HP Ottawa gas engine.

HP Ottawa

Peter Rooke’s unrestored 2-1/2 HP Ottawa

Photo By Peter Rooke

Content Tools

Fuel tank

The original round fuel tank on the 2-1/2 HP Ottawa was 9.50 inches in diameter with 0.250-inch seams, 3.25 inches high, constructed from 26-gauge (0.18-inch) galvanized steel. This struck me as being a little thin for durability, so I used 18-gauge (0.48-inch) ungalvanized steel left over from an earlier project.

I first decided how to cut out the three pieces needed for this tank while wasting the smallest amount of sheet metal. I drew two circles 10.125 inches in diameter, including 0.625-inch extra for folding the seams, and a piece 28.8 inches long by 3.75 inches wide for the side. I also drew lines marking the folding points for the seams before the three pieces were cut out with heavy metal shears.

To start, I clamped the long-side piece between two strips of flat bar at the marked line for one of the seams. I held a hardwood block against this seam that I then folded it over by hitting with a hammer. Once the two seams were folded, I turned my attention to the two round pieces for the top and bottom, with a former being needed to create the lip on them.

I found a cast iron wheel with a good, clean, square edge, but it was not quite big enough. I bent a length of scrap metal sheet then tack welded it to the wheel, thus providing a former the exact size needed, 9.50 inches. I used this to create a 90-degree bend to form the lips of the top and bottom pieces.

I rolled the side piece to the rough diameter of the base by gently easing it around the wheel former with hand pressure. After I removed short sections where the top and bottom lips of the side piece overlapped, I adjusted this side piece to fit inside the base. Once satisfied with the fit, I cleaned the mating surfaces where the ends met and fluxed then tinned with high temperature solder. I again fitted the side piece inside the base piece so that I could solder its ends together.

Four tabs fitted to a ring of 0.091-inch thick wire resting on the bottom lip of the tank would hold the tank in place on the cart timber. I cut a length of this wire and made a half lap joint in each end before I brazed them together to form a ring that would just slide over the side of the tank.

I cleaned and fluxed the seam of the base before tinning with solder. I again fluxed the tinning and put the side/bottom together, then folded over the seam with a hammer before applying heat to seal the tinned faces together. Before starting to finish the top, I filled the base with hot, soapy water that would remove any surplus flux. This also allowed me to check for leaks.

Before fitting the top to the tank, I checked the positioning of the fuel filler cap and the hole for the fuel takeoff. I already knew the dimensions of the wooden skid, so I placed the fuel tank in position to ensure that when it was permanently fixed on the wood the fuel takeoff would not face any obstruction in lining up with the mixer, and the filler would be clear of the rocker arm so that I could easily pour the fuel into the tank.

Once satisfied with the positioning of the filler caps, I drilled and filed to size two pilot holes and then soldered the caps in place using high temperature solder.

The final steps were to clean, tin, bend over and heat the top seam to solder it before again filling the completed tank with hot soapy water for cleaning and testing. All that remained was to fit the four tabs to the wire ring to secure the tank, done by folding them around the wire ring and adding a dab of solder. Finally, I gave the tank several coats of paint.


This engine had been converted to high-tension ignition at some stage in its life. My aim with this restoration was to restore it as a Webster-powered low-tension engine. It is likely that this engine did not have a magneto originally, as there was an arm on the pushrod for an igniter trip to pivot on. It is unlikely this would have been provided if a Webster magneto was fitted.

I had already found a trip arm and igniter bracket, and they looked sound enough that they only needed to be tidied up and new springs installed. It took a while to find the Webster K that was eventually purchased as “not working.” It was clear when it arrived that it needed a fair amount of attention, a bit more than just new springs and rollers. I fitted some modern crosshead screws in place of the originals, so I would have to make replacements, and then I removed the roller on one end cap.

I stripped down the magneto, using a marker pen to clearly identify where parts had been fitted. It soon became clear that it had gotten quite wet inside as some of the laminations on the pole pieces and armature had been spread apart through rust forming between them.

As soon as I touched the insulation block at the top of the magneto it disintegrated, so I made a replacement from some dense plastic, cutting a slot in the middle for the old terminal block that I then epoxied in place.

I cleaned the rust off the armature laminations using a small brass brush. At each end of the laminations rust had either eaten away or spread out the end. When I turned the armature, these were rubbing on the pole pieces, so I cut them off and lightly oiled the armature.

The pole pieces inside the magneto had also suffered rust damage and the cotton wrapping around the coils had seen better days, falling apart at the lightest touch and exposing corroded windings. In view of the need to clean up the pole pieces to remove rust and the damaged laminations that were interfering with the armature, I decided to remove the old coils and fit modern replacements that would be far hotter. Before starting to remove the coils, I noted the position of them and the wiring. If new coils are fitted the wrong way round then the winding direction of the coils could be reversed with the result that they will cancel each other out and not generate a spark.

Due to rust the laminations had spread out and firmly locked the old coils in place, so the only way to remove them was to use snips to cut away at the coil wire.

Once I had removed the coils, I brushed the laminations clean and removed the deformed ones with a cold chisel. I lightly stroked the edges at each end of the laminations with a file to round them, so they would not cut into the protective layers of the new coils, then I lightly oiled the inside of the magneto. The new coils were purchased from Mitch Malcolm at Lightning Magneto. I then fitted the end plates and armature and conducted a test to ensure the armature moved without snagging. There did not appear to be any excess play in the armature bearings when conducting this test.

It was straightforward to fit the new coils over the pole pieces, but as it was cold in the workshop the coils were gently warmed in the kitchen first to make them more flexible. It was also essential to scrape the protective insulation off the ends of the connecting wires before hooking them up so that a good electrical contact would be made.

Once I fitted the coils, I assembled the magneto, fitting new springs and making a new roller to replace a damaged one. I had already made new screws on the lathe to hold the side plates in place. After a couple of quick bursts on the magnet charger, I removed the springs and gave the magneto a spin test with the drill. A good 11.5 volts registered on the ammeter at 500 RPM. Perfect.

My next step was to strip the igniter bracket, clean it up and fit new mica insulation, as the old insulation was in a poor state. The shaft of the moving electrode appeared badly worn just below its head so it wobbled a little when turned. I was able to pass a piece of 0.375-inch drill rod through the body of the igniter, indicating that the hole was good and that the shaft was the only problem.

Rather than make a new shaft, I built up the worn area with weld and then machined it down to size. Before adding the weld, I fitted two steel washers on the shaft, either side of the area to be repaired to prevent the spread of weld or splatter. I added weld and then, when cool, turned the shaft to size on the lathe.

I bolted the magneto to the bracket and adjusted it, in particular moving the adjustment screw on the electrode arm so that it was just touching the tail of the push-finger. I then operated the magneto using the starting lever and it generated a fat spark when tripped.

Crank guard

The factory crank guard would have been made from cast iron. By comparing photographs of originals against this engine, I was able to work out the approximate dimensions. My first step was to make a rough frame from some scrap sheet that could be bolted to the engine then bent and manipulated to get both the clearance for the big end bolts and shape of the crank guard right.

I would make the guard in sections, bolting the base section to the engine from 0.250-inch thick plate and the body of the guard from 12-gauge (0.104-inch) sheet steel. This would be a little thinner than any casting but any thicker would make it much more difficult to shape. I used a filet to join the main part of the guard to the mounting flange, making it a simpler exercise than trying to shape it all out of a single piece.

The first step was to form the main section by hammering out a gentle hollow in the middle and then beginning to bend it to the profile of the pattern, pushing gently by hand over the anvil, a short length of rail track. The second stage was to fold over the edges and get this bend without creasing the metal. Several V’s were cut out to be made good later with weld.

It was easy to cut, drill and gently curve the mounting plate to fit the engine casting. The more difficult phase was to dome out some more 12-gauge sheet as a filet between the mounting flange and the main part of the guard.

After a fair amount of trial and error, using a grinder and files a good fit was obtained. I then MIG welded the three parts together before cleaning with a grinder.


The original muffler was a dome, 5.25 inches in diameter, threaded for a 1-inch NPT connecting pipe, each half of the dome being 1.375 inches tall. The thickness of the metal at the edge was 0.25-inch, so I would form the muffler from metal sheet of this thickness. The first step was to mark the center point on each piece, both of them being cut oversize so as to provide a good flange to make it easier to work on them. I hit a small hollow in both pieces so that they would easily rest in the section of hollow tube 4 inches in diameter, which would hold when the dome was being hammered out.

I heated the metal on the brazing hearth to make the forming easier, holding it with a long-handled wrench to steady it on the forming tube when hitting it. This was a lengthy process to form the dome, and the process is fairly noisy so I wore hearing protection. I used a cardboard template so that progress in forming the dome could be assessed.

When both domes were the correct shape, I held the smooth, rounded end of a cold chisel in the vice to be used as a planishing stake. I rested the dome on this stake then moved the stake around while the dome was hit with a smooth-headed hammer to remove any small indentations, creating a level surface.

I then sawed off the surplus metal around the muffler prior to it being ground, leaving a 0.625-inch wide rim.

I turned the threaded boss for the connection pipe to a diameter of 2.125 inches before cutting a shoulder to sit the muffler on, then drilled and threaded 1.00 NPT. I held one half of the muffler in the three-jaw chuck so that a hole could be bored to fit up to the shoulder cut on the mounting boss. While I was still holding the piece in the three-jaw chuck, I marked the centerline of each jaw on the rim of the muffler as a reference point for the three holes that I drilled for the rivets to join both halves. I then made the three 0.250-inch rivets and their 0.125-inch spacers, and the muffler was ready for assembly.

I first brazed the boss in place, then I used the rivets to bind the two halves together. I added a short length of threaded pipe to complete the muffler.


The pushrod had suffered badly from both wear and rust and was a sloppy fit in its guides. In order to get a stable platform so that the igniter trip works precisely the same time and time again, I made a new one from 0.375-by-0.570-inch thick steel.

The original pushrod had a pivot post on the side for a trip arm that was used with a battery-operated igniter, so this was not needed. I could reuse the original roller and latch-out plate. The screws holding the latch plate in place were rusted in, so I used penetrating fluid then an impact driver to free them. I trued the edge on the latch plate square with an oil stone.

I also made a small hook from some 0.125-inch steel and fitted it to the pushrod to act as one of the anchor points for the return spring.

The pushrod should be held in place by a metal plate, with spacers top and bottom. I cleaned up the spacers, using a file to get square bearing surfaces. The two metal plates had suffered badly from rust, so I made new ones from some 0.125-inch steel plate. To ensure a precise fit, I needed to add some shims to the spacers so the pushrod moved freely, yet did not exhibit any unwanted movement.

One of the spacers should have a small hook to anchor the rebound spring for the pushrod. I cleaned out the remains of the old one from the spacer and made a new hook, again from 0.125-inch steel.

Drip feed oil pipe

When stripping the engine, I snapped the pipe for the drip oiler off flush with the top of the cylinder and part of the pipe was left in the threads of the hole. I drilled this, starting with a 0.312-inch drill. I gradually increased the drill diameter up to 0.406-inch, taking care to stop short of touching the threads. I then used a steel pick to remove the start of the broken thread, before using a 0.250-inch NPT tap to remove the remainder and clean up the thread.

Once removed, I cut threads at both ends of some 0.250-inch nominal NPT pipe and added a connector to couple up with the oiler.

Read more about the beginning stages of the restoration in Ottawa Engine Restoration and 2-1/2 HP Ottawa.

Contact Peter Rooke at Hardigate House, Hardigate Rd., Cropwell Butler, Nottingham, NG12 3AH, England •