Hercules Restoration: A New Webster Magneto

The Hercules engine gets a Webster type M magneto and some finishing touches.

Hercules Model F

Circa-1923 Hercules Model F

Photo by Peter Rooke

Content Tools

Igniter Bracket

This engine was restored to use a Webster Tri-Polar magneto. I found a well-worn igniter bracket (type number 303M1A) on eBay, recognizing it would require a little work. As usual, once received and examined, it needed a lot more work than anticipated.

It was clear that the advance/retard lever spring, pivot and roller would need replacing. The pivot pin had completely lost its shape with wear, the roller had worn through and the spring on the adjustment arm was broken.

The spring was easy to fix; the old rivets were drilled out, the holes cleaned out and a new spring riveted in place. Fixing the pivot pin was a relatively straight-forward turning task to create the eccentric. Some 0.50-inch diameter steel was set in the 4-jaw chuck, but with a piece of 0.125-inch steel between one of the jaws and the rod as a spacer. This was adjusted to run true, then the opposite jaw to the one with the spacer was slackened so the spacer could be removed. This jaw was then tightened so the steel rod moved across to rest against the opposite jaw. This resulted in the steel rod moving its center by 0.125-inch creating the eccentric cut. The rod was trimmed to length after turning it and holes were drilled at each end for the retaining split pins.

I had an old but serviceable roller in the oddment box, so I used that as a replacement. The hole in the igniter bracket for the pivot pin was distorted, but I left that until later to see if it affected the operation of the trip rod.

When I first tried turning the moving electrode it appeared reasonably tight in the bracket, but on stripping the igniter it was clear this had been achieved by bending it slightly. The first step was to pass a 0.3125-inch reamer through the igniter body to check shaft size. This wobbled, but a 0.343-inch reamer gave a clean hole. A moving electrode shaft would have to be made to fit, which could be done by cutting off the electrode head and brazing it to a new oversize shaft.

The overall length was measured and a spot punch was used to mark the head in precise alignment with the keyway in the shaft. This would help set the alignment for the new shaft. The head of the moving electrode was held in the 4-jaw chuck. Measuring with a dial gauge ensured the section of old shaft near the head was running true. The shaft of the electrode was cut off 0.50-inch from the head and that stub was turned down to 0.1875-inch diameter.

The next step was to make the new shaft. First, a section of 0.375-inch drill rod was turned to a diameter of 0.343-inch. Next, 0.300-inch of one end was turned to a diameter of 0.250-inch, ready to be threaded 1/4 UNC. The lathe compound slide was then offset to cut the taper near the end of the shaft. This was set to mirror the original and a taper was cut. To check the fit the taper was given a thin coating of engineer’s blue then the arm fitted. To get a perfect fit the taper was re-cut, using a dial indicator to measure the adjustment of the compound slide.

After cutting the thread for the end nut the shaft was held in the milling machine vise and a 0.063-inch slot was cut for the woodruff key. The shaft was then trimmed to length based upon the original overall measurement, but deducting the length of the head, less the spigot. A hole was drilled in the cut end for the spigot, taking care that it was true and the precise size so that the head and shaft would be in perfect alignment.

The head was brazed to the shaft after being cleaned and the parts fluxed. Lines were cut in the spigot with a scriber to trap flux when it was assembled. The slot in the shaft was aligned with the earlier punch mark and after brazing the surplus was cleaned off, first by turning using the lathe then with fine emery cloth. After adding a new contact, machined from a nail with the end flattened to hold it in place, the moving electrode was finished.

The igniter bracket was then assembled for testing, with a new mica tube and washers fitted to the fixed electrode. As I had suspected, the oval hole in the bracket meant the advance/retard roller had too much movement. To fix this, the bracket was held in the vise and the milling machine centered over the “round” part of the hole. Using a boring bar the hole was opened to 0.625-inch diameter. The boring bar was used as a drill bit would want to move towards the oval center of the hole.

A 0.50-inch ID bushing 0.002-inch larger OD than the hole was made and pressed into place. After fitting, the bushing was cleaned out with a reamer as it had tightened very slightly when it was fitted. After assembly of the bracket, the only outstanding task was fitting the magneto and adjusting the screw on the moving electrode arm so it barely touches the magneto’s push finger.

Igniter trip assembly

When the engine arrived it had the clamp fitted to the pushrod for a Wico magneto, but nothing else. Rather than search for a push finger for the Webster Tri-Polar, the whole assembly was made by copying one off another engine in my collection.

The first task was to make the trip pushrod assembly, starting with the trip journal, which was made in two pieces. First, a 1-inch diameter bush with a 0.50-inch ID was trimmed to a length of 0.70-inch. A length of steel was turned to a diameter of 0.6-inch, drilled and threaded 5/16 UNC for the trip pushrod thread. The end of this part was bored to be an offset fit to the bushing before being cut to length.

The two parts were then brazed together and a small tab of metal filed, then drilled 0.125-inch for the trip rod spring before it was brazed in place. This journal was then filed to shape and finally the small oil hole drilled in the top to lubricate the pivot. The trip rod was quickly made by turning and threading a section of square 0.3125-inch steel to 5/16 UNC for 2.75 inches. The addition of lock nuts completed this part of the assembly.

Not possessing the tools to broach a square hole for the trip rod wedge to fit on the trip finger, it would have to be made another way. One method is to mill a 0.312-inch square slot in steel and then braze a flat plate on top, thus creating the square hole.

A block of steel 0.625-inch thick by 1.125 inches wide and 1.50 inches long was held in the milling vise to cut this slot before a piece of 0.125-inch plate was brazed to it. After cleaning, the 0.625-inch boss for the locking screw was cut by holding the block in the 4-jaw chuck and turning it on the lathe. Finally, the mill was used to remove the cut-out at the front of the block and taper the front bottom edge, with the final shaping of the wedge being accomplished using a grinder and file.

The clamp block was a little more involved. First, a 0.375-inch deep by 0.625-inch wide slot was cut in a block of steel 0.625-inch thick by 1.75 inches long and 1.50-inch tall. Some 0.250-inch thick plate was cut to fit this block and was then brazed to it to create the rectangular hole for the pushrod. A further block of steel was then welded to the first block (see page 20), with extra weld to create a curve between the two blocks.

Sawing, milling and grinding got the outline of the clamp right before drilling holes for the pivot pin and the two locking bolts to the pushrod. After the addition of the pivot pin, held in place by a spot of weld inside the hole, the lug for the trip finger spring was added before shaping the clamp by grinding and filing. Once the trip assembly was completed and assembled it was fitted to the pushrod and set up ready for testing.

Webster magneto

The Webster type M magneto, another eBay purchase, was stripped down and cleaned, given a charge, then tested for voltage. This indicated 6-1/2 volts; below the suggested 8-1/2 volts for this type of magneto. The magneto fired the engine, but it was not consistent, so a new set of coils was ordered from Mitch Malcolm at Lightning Magneto. After stripping the magneto the wiring of the old coils was recorded so the new coils would be fitted correctly. Fortunately, it proved easy to gently ease the coils off the laminations with a screwdriver.

The laminations were carefully cleaned and checked for any sharp edges or corners that might cut into the new coils. If the new coils were tight it would have been necessary to remove one layer of laminations so they could slide on. The new coils were gently pressed in place and the connections made. A dab of glue was used on each wire to hold them in place clear of the armature. The magneto was then partially assembled, magnets re-charged and given a spin test, this time registering 8.5 volts. This gave a very healthy spark.


Before assembling, the engine was brushed clean of dirt and given a coat of ATF to bring out the color of the remaining original paint. The connecting rod cap bearing had moved, the small stub locating it in the cap had broken off at the bearing. The bearing still had plenty of life, so it was repaired. To do this, some white metal was heated in a ladle. At the same time, the underside of the bearing was heated using a soldering iron, applying heat through the stub hole. Once the bearing started to show signs of melting, metal was poured into the hole to bond with the bearing. After cooling the cap was cleaned off with a file.

The dimensions for the bore and piston had been checked earlier. The diameter at the midpoint of the piston measured 3.246 inches and the bore of the cylinder 3.254 inches. There was still plenty of spring and thickness to the piston rings so they were retained. The piston and cylinder were coated with oil and the piston was pushed into the cylinder with the aid of a ring compressor.

The crankshaft was then greased around the bearings and fitted, and the timing gear was installed and aligned with the timing gear wheel. There were one or two nicks in the keyways and on the crankshaft; these were cleaned off with a file then polished using emery cloth before fitting the flywheels.

New gaskets for the cylinder head and igniter were cut and fitted, along with the governor and pushrod. The governor detent blade was adjusted first. With the exhaust valve fully open the detent blade should be pushed behind the catch block on the pushrod. The detent blade should be adjusted so there is only the thickness of a postcard between it and the catch plate. With the speed lever set nearest the cam gear the detent blade should be 3/4-inch out from the pushrod. If necessary, it can be bent using pliers to get the correct measurement.

Any timing marks on the flywheel had long gone, so as a first step the flywheel was turned until the piston was at TDC, then a line was drawn on the flywheel face aligning with the top of the pushrod as a reference point. The timing of the exhaust valve should be set first, to open 45 degrees before bottom dead center and close 3-5 degrees before top dead center. These were set by eye, the adjustment screw in the rocker arm being moved to achieve this. For ignition with a Tri-Polar magneto, the general rule is 8 degrees of advance for every 100 rpm of engine speed, so 40 degrees advance was used for a 350 rpm maximum speed. One-ninth of the flywheel circumference was measured and a further mark made.

The igniter was set up when it was assembled, so it was just a case of setting the pushrod. The Webster cocking lever was used to tension the springs, with the flywheel in the ignite position and the retard lever set to run, and with the wedge moved out of the way. The pushrod should just touch the push finger and be tightened in this position, then the wedge moved into play so that, without moving the flywheel, the bottom edge of the pushrod is level with the top of the push finger, and the locking screw tightened. The cocking lever can be removed and the flywheels then turned a couple of time to check the magneto is firing at the right point.

With both the cylinder head and fuel tank in place the fuel pipe was bent and fitted. With the two elbows in place on the mixer and a check-valve fitted to the “T” a short piece of wire was bent to fit between them as a template for the brass tube. Some 0.250-inch diameter brass tube, longer than needed, was then annealed to make it soft, heating it red hot for a short while before letting it cool gradually. A small bending tool was used to match the template and after a trial fit the ends were trimmed to size.

Once the engine was assembled it was bolted onto the new cart. After oiling all appropriate points and filling and turning down the grease cups, fuel was added. The mixer needle was opened one full turn, but this proved too much and was reduced to one-half turn. After a number of attempts, the engine fired. It proved to be running too slow then cutting out as the governor kept latching, so rather than bend the governor detent blade, the spring in the governor was replaced with a stronger one.

This engine was set to run at a lower speed than the factory norm of 550 rpm. Without a load the speed drops to 280 rpm before firing to a maximum of around 350 rpm. To get smooth running, the choke plate on the mixer was set barely open to get a strong airflow for the fuel vapor. At these settings the engine ticks over nicely all day.

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