This is the first in a three part series on Peter Rooke’s restoration of a John Deere Model E. You can read part 2 and part 3 here.
The John Deere Model E was first built in the 1920s and at that time it was a leading design, having an enclosed crankcase when open crank engines were the norm.
Deere & Co. was established in the 1830s when John Deere, a blacksmith, started working as a repairman in Illinois, then began making plows and later other agricultural equipment. In 1918 Deere bought the Waterloo Gas Engine Company to enter the growing tractor market as its own designs had proved unsuccessful. The Waterloo tractors were sold until Deere’s new design, the Model D, emerged in 1923. Waterloo model H and K gas engines were produced from 1919 and were phased out after production of the John Deere Model E started in 1921. The Model E was produced in three ratings, 1-1/2 HP, 3 HP and 6 HP, and apart from the introduction of a crankcase breather later in production, were relatively unchanged over the 23 years they were built.
My engine arrived with the rocker arm broken and, not surprisingly, covered in dirt, but at least it turned over freely. The first thing checked was the engine number, 266,583, indicating that it was built in 1926. I had hoped that this would be the first engine that could be restored without repainting, and that I could leave it in its original state. However, when the dirt was cleaned off there were different colors of green paint. When the newer paint was removed, specks of orange paint lay underneath. Foiled again, yet another repaint job!
Stripping the engine was fairly straightforward. The cylinder head was only held on by a couple of nuts, the others were missing and the fuel line had already been disconnected. Before taking the head off, the igniter was removed. After undoing the three machine screws holding the magneto cover plate, the three bolts were removed from the crankcase cover to release it from the main casting. The big end cap was marked to ensure it would be replaced the same way before removing its bolts so that the piston and connecting rod could be pushed out from the front of the engine. There was no oil to dispose of, only a thick black goo that I washed out with kerosene.
The piston rings were stuck tight in their grooves, so the piston was soaked in a bowl of kerosene to release the build up of old oil and carbon. This enabled the rings to be loosened and removed. Once they had been taken off, it was clear that they would need replacing as they were very thin and had no spring in them. After soaking in kerosene it was easy to clean out the piston ring grooves.
A tricky part in stripping the engine was removing the magneto. First, all the gear wheels were checked to ensure that the factory alignment marks – a small “O” stamped on the rim – were visible. If none of the marks are visible, even after cleaning with emery cloth, then the meshing points of the crankshaft gear, cam gear and magneto must be marked to make it easy to re-fit them in the same position.
The magneto is held in place by two bolts from inside the governor case up into its base. The first step was to remove the six machine screws holding the governor cover plate in place. It was very fiddly to unscrew the magneto bolts; it required a wrench with a small head.
As no work was needed on the flywheels or crankshaft, they were lifted off in one piece. Before being removed, the bearing caps were first marked to ensure they would be replaced in the same place. They proved a bit tight, which might have been from the fact that the end of the shims could have been sealed with shellac or a similar product to prevent oil escaping. After persuading the caps loose with a mallet and brass rod they were removed along with the top bearing shells and the shims. This enabled a mobile hoist to be used to lift out the crank and flywheels.
It was necessary to strip the governor assembly as some teeth were missing off the gear on the governor shaft. The speed change nut was removed before sliding the cam follower off its shaft. The metal disk (welsh plug) at the top of the governor case was removed by unscrewing the plug at the bottom of the pivot rod for the hook up lever, then a brass drift and hammer were used to push it up. It is important to remember that the nut holding the governor shaft and weights in place is threaded with a LEFT hand thread, and this is perhaps the reason for the stripped gear teeth. To unscrew this nut, place a small length of wood through the spokes of the camshaft gear to stop it from turning.
The fuel line was then removed from the tank, before the bolts holding the sub-base and fuel tank were removed so they could all be separated.
The cylinder head appeared to be in reasonable condition, apart from the broken rocker arm and damaged oil cap in the end of its pivot bolt. The nut on the top was unscrewed before removing the pivot bolt.
The mixer choke plate, adjusting screw and internal jet were removed for cleaning.
The exhaust valve was stuck in its guide. The rust on the stem had to be cleaned off before it was soaked in penetrant and later eased out with the help of a copper head hammer.
The valve heads and seats looked as though they only needed to be ground in to return them to good condition.
The muffler was stuck fast in the head and, as there appeared no work needed on it, it was left untouched rather than risk damage to the head by trying to move it forcibly.
The John Deere name on the side of the water hopper was screen printed on and was not a decal. Rather than remove this as part of the cleaning process and replace it with a decal, the decision was made to try and preserve this lettering. To remove the old green paint a small modelling knife was used to cut around the letters before cleaning off the old paint using a thin rotary cutter in a Dremel.
Yellow paint was missing from parts of several letters, so this would need to be built up again.
The remains of all the old gaskets were cleaned off and it was clear that a new set of gaskets would be required. To fully remove several of them that had been fitted with sealant it was necessary to use a scraper to get back to a clean metal surface.
Once the main casting and other large items were clean, they were all given a coat of red oxide primer to protect the metal over the coming winter months.
The fuel tank had been partially filled with water when it was removed, which at some stage leaked through the missing center bolt hole in the crank casing and also corroded the sheet metal top to the fuel tank. This resulted in a number of rust pin holes in one end of the tank.
Before spending money on a replacement tank, an attempt was made to repair the old one by patching or using a fuel-proof epoxy sealer to line the inside of the tank.
To identify the extent of the rust problem, a gas torch flame was first played over the metal of the tank and further pin holes started to pop and appear, but there still appeared to be plenty of sound metal.
The area around the drain plug also looked as though it might leak, so it was heated and the fitting removed, to be refitted later.
The pin holes were around the curved area at the end of the tank, making it a little more difficult to repair as the metal patches would need to be shaped. This was achieved by bending patches on the ball of a large engineer’s hammer, each being individually formed.
As the inside of the tank was so badly pitted, the patches would have to be fixed to the outside of the tank, which was given a good clean with emery cloth. The patch and tank were both fluxed and then tinned with soft solder before a soldering iron was used to fix the first patch in position. Once the first patch was in place, the second was fitted, adjusting the size and edge to the first patch as well as its curvature so that it fitted the tank as closely as possible.
Once all the patches were in place, the drain plug fitting was cleaned and also soft soldered in place. The tank was then tested by filling with hot water, which would wash out any flux in a bad joint, and left to stand to see if any leaks developed.
The end result is not exceptionally pretty, but if covered with several coats of paint the patchwork will not be that noticeable.
The top of the fuel tank with the oil pan, which fits between the main casting and the sub-base, had also been perforated by rust. In this case it was easier to make a new sheet for the base and fix the old oil pan to it rather than try to patch and repair it, as both sides were pitted by rust so the solder would not take.
The original fuel tank top was used as a pattern when cutting a replacement out of some 0.035- inch sheet. The position of the oil pan was measured, so once it was removed it could be fitted to the new one in the identical position.
The new sheet was cut to the same outline using snips and a file. The clearance holes were drilled for the bolts before the larger holes for the fuel filler and outlet were drilled and filed to size. The oil pan, which had been spot welded in place, was removed from the old top and then the base of it was cleaned up.
To fix the old pan to the sheet metal its position was marked out before drilling 4 x 0.250-inch holes through the sheet to align with the base of the oil pan to plug weld it. Clamping the oil pan tight against the sheet, a MIG welder was used to fill each hole with weld, starting in the middle so the initial contact is made with the base of the oil pan, then filling out. For this type of “spot welding” the larger the hole the better, or else the weld will just fill the hole, with the arc being made with the top sheet rather than going through the hole to the metal sheet underneath.
The gear on the governor shaft had some of the teeth broken off and a repair was necessary. As a first step the split pins that held the governor weights in place were removed and the plunger taken out from the middle of the shaft.
The first task was to count the teeth (14) and measure the diameter of the gear (1.60 inches), which identify that in “gear cutting terms” the gear was sized 10 DP (diametric pitch), a common measurement for gears on old gas engines.
Fortunately there was a cutter the correct size and profile in the workshop, so a new gear could be made. The broken one was turned down to the shaft before fixing the new one in its place.
A piece of cast iron was turned down to 1.6 inches o/d and a 0.437-inch hole bored in it to fit a mandrel I had, before it was parted off to a length of 0.500 inch.
The gear was then mounted on the mandrel and the dividing head was set up on the milling machine table. The set up took some time, involving frequent checks with the dial gauge and adjustment to ensure the gear blank was running true.
The dividing head was set for a count of 14 and the gear cutter centered on the blank with all scales on the milling machine zeroed.
The depth of cut for this gear was 0.216 inches, and one by one all the teeth were cut.
Once all the teeth had been cut, a temporary collet was made to hold the gear in a three-jaw chuck, so the center hole could be bored out to a diameter of 0.875 inches.
The next task was to turn down the old gear, leaving a shaft 0.874 inches in diameter that the new gear would slide on to. This stage was deliberately left until after the hole had been bored in the new gear, as it is easier to measure the size of a shaft than a hole.
To set the governor shaft for turning, the winged end was held in a four-jaw chuck and the surface of the bearing area used to check concentricity with a dial indicator. A revolving center, mounted in the tailstock, was used to steady the end of the shaft while the remaining teeth of the old gear were turned down. The depth of cut was reduced when near the required size and the new gear was used as a gauge to check the fit for the last few thousandths of an inch.
Once the gear slid on, the governor shaft was ready for brazing. All contact surfaces were cleaned before both the inside of the gear and the shaft were covered in flux and the two brazed together.
While replacement rocker arms are readily available, an attempt was first made to repair, and if this failed then a replacement would be purchased.
The first step in the process was to clean any loose material from the break, then temporarily join the two parts using super glue to hold them together, while two holes were drilled through the arm for the support rods. A 0.094 inch size drill and steel rod were used.
Once the two holes had been drilled, the joint was broken and the super glue cleaned off both sides using small grinding points and finishing with a brass brush. Before joining the two parts, the old metal was covered with flux and heated to brazing temperature, at which point the flux turned black with the dirt it had drawn from the metal. The cast iron was allowed to cool slowly to avoid it becoming brittle, which is the last thing you want with a rocker arm that is subjected to shock. The dirty flux was cleaned off and the process repeated again until the flux turned more grey than black.
After a final pass with the brass brush, the strengthening rods were inserted after they and the joint were given a good coating of flux, before the two parts were finally brazed together. Once the part had cooled, the over-length pins were trimmed flush and any excess braze filed smooth to give a perfect repair.
The cylinder head looked in good order apart from the valve seats that had a covering of light rust. Both the seats and valves looked in reasonable condition, so all they needed was a bit of grinding in. The valve stems were a good fit in the guides and did not need any attention after the surface rust had been cleaned off.
The seat of each valve was given a thin covering of valve grinding paste before the valve was turned in the seat in a back and forth motion to bed it in. A grinding stick with a rubber suction cap was used, although a screwdriver or brace and bit can also be used, taking benefit of the slot cut in the valve head.
During the grinding process the valve was frequently lifted and the grinding paste redistributed, taking care not to drop any paste onto the valve stem or its guide. Once there was a clear ring of clean metal the job was done and the whole area was then carefully cleaned with kerosene to remove any residue of the paste.