1919 Eaton Restoration – Part 2 of 3

Making governor weights, a latch arm and muffler for the Eaton from scratch.

eaton 1 hp

1919 1 hp T. Eaton Co. engine

Photo by Peter Rooke

Content Tools

This is Part 2 of Peter Rooke’s series on restoring a 1919 T. Eaton Co. engine. Start at the beginning with Part 1.

Governor weight

The governor weights were missing from the Eaton, as well as the latch arm that locked the pushrod and thus held the exhaust open. Fortunately, several photographs were found, including some very useful ones on eBay where the vendor had helpfully laid a ruler alongside the weight. There was only one weight for sale and the shipping costs and duty alone would have added another $35 to the price, so the photographs were a great asset in making the weights.

A single long piece of 1-inch-diameter steel rod was used to fabricate the two weights, making it easier to hold the work in the lathe chuck and vice. First, the 0.25-inch holes were drilled at the end of each weight for the springs, spaced 2.75 inches apart. Allowance was made between each of the two weights for parting off and the holes for each weight were drilled at 90 degrees to each other.

The steel rod was then machined on the mill, cutting rectangular 0.25- by 0.50-inch profiles at each end of a weight before transferring it to the lathe.

Using a ball turning tool set to a radius of 2.5 inches, the bodies of the weights were shaped. It was not possible to shape right up to the rectangular sections at each end in view of their different orientation. The cutter was stopped short and the radius later finished with a file.

To complete the weight body, the recess for the flywheel hub was filed, along with a flat where the arm would be welded. To mark the center line and the depth to file to, a hacksaw was first used to cut a shallow slot.

In order to make the remainder of the weight, the curved arm and pivot pin support, a wooden pattern was made. First, a hole was drilled in some 0.625-inch-thick wood so that it could be held in place by the pivot pin. The wood was then sawn and filed until it would both fit in position and engage with the governor sleeve. A piece of 1-inch-diameter steel was inserted through the flywheel hub to act as a temporary crankshaft so that the moving governor sleeve could be fitted to check the shape and working of the pattern.

Using this pattern as a guide, the outline of two parts was marked out on a single piece of 0.625-inch-thick steel. The next task was to drill the 0.25-inch holes for the pivot pins in the steel so they could be used as the basis for measurement and shaping. Using one piece of steel again made it easier to hold the work for the majority of the machining.

A temporary pivot pin was then put through one of the holes and this piece of steel was held in the vice, aligned so that the milling cutter could cut away metal to form the lug to fit in the governor sleeve. The pivot pin was used as a reference point for the measurements to align the cutter.

To roughly mill the rounded portions of the arms, a threaded pivot was held in the milling vice and the arm bolted to it so the piece could be turned by hand when the cutter removed metal. This is where the length of the work-piece provided leverage/control.

The block of steel was then removed from the mill and held in the vice so the arms could be filed to size. The two arms were then separated and center lines marked.

Two steel 0.75-inch-diameter washers were then cut to fit either side of the pivot pin hole. These were secured with a nut and bolt and used as filing guides to shape the rounded portion of the governor weight.

The arm linking the weight to the pivot area was then rough filed to shape, its length being assessed by fitting it to the hub and holding the weight in position. Once the length had been set, the end of the arm was ground to a point to provide the maximum area for weld to adhere to. The arm was then fitted to the hub with a pivot pin and the weight clamped in place, the marked center line abutting the end of the arm. Once satisfied that the two pieces were correctly aligned, they were welded together. A thin welding rod was used to get the puddle of weld to the bottom of the ground-out section. Once this first pass had been completed it was cleaned up with a pick and wire brush before more weld was added. Care was taken to ensure that there was complete penetration of the weld to fully secure the weight to the arm.

Once the top and the sides had been welded the weight was removed and the underside was welded. After cooling, the weld was ground to shape, further weld being added to fill any imperfections. The second weight was then completed using the same method.

Once the weights were completed the moving sleeve was fitted and the movement of the two weights checked. The pivots were made by trimming the temporary pins to length and drilling holes in both ends for the split pins to hold them in place. To provide a means to alter engine speed, two threaded adjustment screws were made to allow changes in the tension of the springs, thus providing engine speed adjustment.

Governor latch arm

Photographs emailed by David Swanson provided the necessary information to make a latch arm. These were scaled against the other parts of the cam assembly and the first part made was the area around the pivot pin, based on two 0.625-inch-diameter washers 0.375-inch thick.

These would be welded to 0.3125-inch-thick rod to form the pivot point. A strip of steel bar, 0.50 inch wide, 0.25 inch thick, to hold the latch block at its extreme end would be joined to the middle of this pivot. To simplify accurate bending of the rod to a precise shape, notches were cut at each bend point of the rod, these later being filled with weld.

Once this part of the latch assembly was completed, a small boss was made from some 0.5625-inch round steel that was then brazed to the outer face of the latch before drilling part through from the inside face to form a recess for a 0.375-inch rebound spring. The latch was then fitted in position and the cam gear turned to move the catch on the cam assembly to its farthest point away from the crankshaft. The length of the latch arm was then trimmed to leave a gap of 0.0625 inch between it and the catch.

A recess 0.65 inch long was then milled across the inner face of the latch for the catch plate, with a 0.25-inch hole drilled at the midpoint. This hole was countersunk on the outer side. A 0.15-inch-thick catch plate was then cut from some high carbon steel, sized to fit the slot in the latch, but fractionally longer. A 0.25-inch UNC thread was cut in the middle of the plate. After fitting the catch plate to the recess, and trimming to allow a 0.0625-inch clearance between it and the catch, it was removed then heated red hot before quenching in oil to harden it.

The next step to complete the latch arm was to weld more round rod to the crankshaft end of the latch to hold the follower that acted against the governor sleeve. Setting the position of the governor sleeve on the crankshaft was easy as there were deep wear marks on it from use.

This final stage took a little time to set up, allowing both sufficient clearance to the cam on the gear and permit full movement of the sleeve. This meant a fair bit of trial and error involving numerous tack welds and grinding off before the result worked satisfactorily.


Before the engine was purchased, a reproduction muffler had been listed on eBay, but by the time the engine was acquired, it was no longer for sale. Nothing else was seen that looked remotely right, so once more a replacement was made. (Once work started to make the muffler another reproduction was offered for sale!)

The muffler had to be 5.25 inches in diameter, fitted to a 0.75-inch NPT threaded pipe, a slight dome on the outer face, and elongated dome/cone on the inner side. Fortunately the one earlier listed on eBay included several pictures, including one full-on from the side. Copies had been taken of these images. This meant the dimensions could be worked out by scaling these photographs.

The first step was to rummage through the scrap bin to see what was available, ending up with three pieces of steel to make the main body and outer cover. The pieces were chosen to reduce the amount of machining needed to remove surplus metal.

The main piece of steel was only 5 inches in diameter. As a first step, a ring of steel 0.25 inches thick was sawn from some 6-inch tube, a section cut out before it was compressed to fit around the main piece and welded in place. The other piece of steel was 2.5 inches in diameter and already had a 1-inch hole through the middle, and would be welded to the main piece later.

Holding the steel in the chuck was always going to be a problem once it was shaped, so the order of machining was carefully thought through. First the outer, large face was squared up and a 1.0-inch hole drilled through the middle.

The rim of this part jutted out from the jaws of the chuck and the width of the rim was machined to the required diameter of 5.25 inches and to give a lip of 0.50 inch wide. The rest of the body of this main piece that was clear of the chuck jaws was then reduced as far as possible. This piece was then reversed in the chuck and the remainder of the body trimmed to the same parallel diameter, 4.25 inches.

The embryonic muffler was moved in the chuck so that the machining of the inside could start. The reference photograph of the muffler had been scaled on the computer so that when printed the major diameter was 5.25 inches, enabling fairly accurate measurements to be taken from this image.

Vertical lines were drawn on the photograph at 0.25-inch spacing so that the diameter could be measured at each line. The inside was then bored out to these measurements, less 0.50 inch of the diameter thus giving a 0.25-inch wall thickness. This was ignored for the first 0.50 inches from the outer face, which was bored to a smaller diameter to provide a deep square edge and good inner face in order that the chuck could grip the inside of the muffler.

The second main part of steel was then faced true and leave a spigot to fit inside the end of the main piece before the two were brazed together. When it was re-fitted on the lathe the muffler was clamped at the outer edges of the rim so that the body could be finish machined to shape.

In a similar manner to the boring of the inside, the outer wall was machined in 0.25-inch steps to get a profile matching the photograph and then, when completed, it was machined again in 0.125-inch steps, using half of the combined 0.25-inch measurements.

A copy of the muffler photograph showing the side view was glued to some thin card and then the profile of the muffler was cut out to provide a template. Using a large round nose tool the body of the muffler was shaped, essentially by removing the steps created in the earlier machining, progress being checked against the template.

Once satisfied with the shape of the outside, the rear hole inside the muffler was opened up to 1.05 inches to fit the 0.75-inch NPT threaded pipe, which was cut to a length of 3.75 inches. Two inches of this was fixed inside the muffler as a form of baffle and the pipe was then welded in place. The muffler was then held in the lathe chuck by this pipe in order that the steps cut on the inside could be smoothed out with a round nose tool.

For the muffler cap there was an off-cut of steel the right diameter, and this was machined to give a rim 0.25 inch thick and 0.50 inch wide. This was then reversed in the chuck to start machining the inside. At 5.25 inches diameter the arc cutting tool would not fit inside, so again the coordinates were calculated and steps cut to the inside and then smoothed out with the round nose tool. After reversing the work piece in the chuck, the outside was similarly shaped. A little work with a grinding wheel rounded any sharp edges.

All that remained was to cut three spacers 0.0625 inches wide to fit between the two halves of the muffler, with three holes drilled equidistant around the rim. Three 0.1875-inch rivets were then used to join the two halves and complete the muffler. All that was now needed was coating with heat-resistant paint.

Pushrod/cam follower

The pushrod follower wheel that made contact with the cam gear was badly worn and out of shape, so a replacement was turned using some high carbon steel to give better durability.

The cam follower slides on two posts fixed in the cam follower holder, but the bores for the posts in the cam follower were worn. To fix this, the holes in the cam follower were reamed out, increasing the size of the bore from 0.375 inch to 0.4063 inch.

The retaining pins fixing the posts to the holder were removed so the posts could be pushed out. The bores in the holder were increased to the same diameter of 0.4063 inch, then two new posts were made to the same diameter, fitted and cross-drilled for new retaining pins.

The support pin for the cam follower/holder had an oil groove that aligned with a hole in the engine block. At some stage this had been fitted 90 degrees out of line to the posts for the cam follower. The pin was removed, replaced with the correct orientation, then new punch marks made to hold the pin in position.


It has already been mentioned that the bore of the engine was in first-class condition, and close examination later confirmed that it had been sleeved. The diameter of the bore measured 3.125 inches. The piston was then checked and found to have diameters of 3.118 inches at the piston ring lands and 3.120 inches at the skirt. The skirt measurement was a little tight, but acceptable, but the land diameters were slightly oversize. The piston was set to run true on the lathe and the top land was turned down to be .013 inch smaller than the bore diameter, the second 0.009 inch and the third 0.007 inch.

The edges of the ring grooves were already square and the piston rings, two fitted in each groove, had minimal clearance between them. The piston rings were checked to the bore by fitting the piston and then butting a ring against it to measure the ring gap. This gap was found to be a little tight in several cases, well below the 0.012 inch (0.004 inch per inch of bore). The “tight” rings were adjusted by filing the ends of the rings with a needle file.