My Kohler Light PUZZLE

Diagram of Piston top & rod base

Diagram B: Piston top & rod base detail.

Content Tools

26 Mott Place Rockaway, New Jersey 07866-3022

This story begins on a cool August evening in 1993. The North Jersey Antique Engine and Machine Club was holding one of its monthly meetings at the Sussex County Farm and Horse Show grounds. At the end of the meeting Jim Quince, the club's president, asked me if I wanted a Kohler 4-cylinder light plant he had. said, 'It's all apart, but know where second unit is that you can use for parts.' told had tried to locate parts his at home, could not find what needed, and thought maybe do something with the two units combined. him was interested, we set date get together.

About two weeks later I went to Jim's home in order to pick up the two light plants. I came home with more than I'd asked for! First, Jim showed me several mid-sized boxes of parts he had in his garage. These included nuts and bolts of all sizes, a lot of broken rings and ring pieces (?!!) and more. 'Where did all of the rings come from?' I asked. 'Well, we kind a broke a few getting the pistons out,' Jim said. 'A few!,' I thought. Other boxed items included a semi-stripped Eiseman magneto (still hot, but the spark distribution assembly had been removed for unknown reasons), a valve rocker shaft with a broken support (that broke too; it was stuck), a couple of broken and whole sparkplugs, as well as a whole lot of pushrods, rockers, and springs galore! What did I get into now, I thought! I soon found out. Next, we went out to collect the parts outfit. It had been used in a sand and gravel pit up until about five years ago, to provide light for nighttime operations at the pit. When A.C. power was installed, the Kohler was taken off line and was abandoned to the elements. The quarry owner said that I could have the unit so long as I took it right away. 'Load it up,' I said and the owner walked away.

This unit was a 1,000 watt, 110 VDC generator which looked like it was all there. It had a Kohler 4-cylinder manual-start engine, a Zenith carb, and an American-Bosch mag that was all there. A tin can had been placed over the upturned exhaust pipe, to keep out water and other uninvited guests and their related materials. A quick look at the oil dipstick told another story though the crankcase was full to the top with water and emulsified oil (you know that milky white corrosive gunk that gets into everything). Then I noticed that the tin can had rusted through over the exhaust pipe. I also noted that the nose bearing that supports the hand crank had been recently broken. I later found out that the owner had stood and jumped on the crank in an effort to free the stuck engine, in order to sell it, but the support broke instead.

Well about now, the owner comes up to us in one of the quarry's big loaders in order to pick up the heavy outfit. It was loaded post-haste with the expert help of the operator. You should see this thing! You could fit two of my Grand Caravan automobiles into the bucket of this huge machine, and yet it set the Kohler generator onto my trailer as gently as a feather! We thanked the owner for the machine and the help loading it and headed back to Jim's house in order to pick up the unit he had.

I backed into Jim's yard, got out of my car and went to look at my 'puzzle.' This was a true basket case! The entire unit had been disassembled soup-to-nuts, with the exception of the field cores in the stator housing, the control box for the auto start mechanism (although the cover and the wiring had been removed unlabeled of course), and the generator's rotor and armature was still mounted on the end of the engine crankshaft. All of this was on the deck of Jim's trailer. Oh yes! I forgot to mention the trailer!!! The three-foot by five-foot trailer was covered by an Army green tarp that was well weather worn. Jim said it had sat out back about two years, and now he needed the trailer to haul engines to the local shows.

Well, we lifted the tarp off the trailer and there was the entire Kohler outfit, less the two boxes in the garage, as I had mentioned, as well as three or four field mice, all over the trailer deck! Two large cardboard boxes of parts, one large bushel basket of goodies like unnumbered ignition wires, starter control wiring, a harness full of eye type connections, a solenoid of unknown type and more. Like I said, a basket case, basket and all, all laid out like a giant jigsaw puzzle on this little trailer.

Jim and I transferred all of the large items onto my trailer. The last big piece was the combination crankshaft-rotor assembly. When we lifted it, out ran two more field mice! This turned out to be the heaviest piece to move (over 100 pounds), and the longest (almost five feet in length). Most of the pieces were in fair to good shape, with the exception of the cardboard boxes which the mice had chewed with abandon, and the crankshaft, which had some rust on the journals (no grease to protect them from the elements) and stink from mice in the rotor assembly! The armature did not escape mouse damage either, the segments being heavily corroded, again, compliments of the mice. After all of the other boxes, parts and the peach basket were loaded, Jim said, 'Oh yeah, just one more thing. This has to go too.' 'This' was an engineless David Bradley garden tractor. This also was winched aboard the trusty and now overloaded trailer for the ride back to Rockaway. Before I left I got some history on the puzzle unit.

Jim told me he had located the outfit while on a trip in Maine. It had been brought into a hunting camp in 1920, over a frozen lake by sledge, as the campsite at the time was inaccessible by land. Fuel and oil was brought in as needed by boat, or over the ice in winter, until sometime in 1986, when several camp owners in the area persuaded the local power company to bring power overland to their sites, at a cost of about $5,000 each! Jim had acquired the unit about three years ago, and removed it from the site by snow mobile over the frozen lake that it had come across so many years ago. It had been in a shed all its life but had seized from disuse, during its five years of idleness.

Jim and a friend, Henry Boutelette, had taken the Kohler apart in an effort to restore it. First, they removed the valve cover and carb from the head. Then the rocker shaft supports (one broken), pushrods and rockers were removed. The head was pulled off, the valves were removed, ground and reinstalled on the head, and the rest of the entire engine was disassembled. All four pistons were stuck tight in their respective bores, so the head was also reinstalled, using the old head gasket. A special fitting was made up out of an old 18mm sparkplug and a zirk fitting, and grease was forced into the cylinders pushing the pistons out. Jim said they used eight tubes of grease to remove those stuck pistons! The rings were stuck in the cast iron pistons by rust and carbon. Jim had broken over half of them trying to free them up. The oil spy line rocker shaft support was broken when the three shaft support bolts were loosened before the oil lines and their fittings were removed. The stress of the pressure from the loosening bolts split the fitting top to bottom, rendering it useless. In a way, I'm lucky that the supply side support did not also bust. After looking for the needed parts, Jim and Henry were stymied in their efforts to locate parts. The stripped down generator was then loaded onto the trailer and relegated to the backyard, where I finally picked it up two years later.

When I got home I took a couple of pictures of the load on the trailer and glanced over the puzzle unit's needs. I figured that it needed at least the following: a complete gasket set, complete ring set, hone job on the cylinders (they didn't look pitted), and a new rocker support. Several connecting rod throws on the crank needed some TLC and besides these items, some nuts and washers, this seemed like a workable project. Basically it was all there.

The first thing I did was to unload the David Bradley onto the driveway. No small feat, considering it had no motor, and a flat tire. The 'puzzle' generator was then unloaded into several piles in my garage, where I took a few more pictures, then turned my attention on the quarry unit.

I took a couple of pictures of this unit from different views and then started work in earnest. First thing on the agenda was to drain the water out of the crankcase. There were almost three gallons of water in it! I left a small pan under the oil drain and tilted the block in a way that any residue in the oil pan would eventually find its way out. Next, I took off the automatic choke relay rod, the air filter, cast iron valve and rocker cover. Then I started to remove the rocker shaft support bolts and nearly made the same mistake that Jim and Henry made. I noticed that as the support bolts loosened, the bolts were putting a lot of tension on the oil lines and their fittings. I suddenly realized what had happened to the support on the other engine and hastened to tighten them back up. After a few minutes' study I recognized the problem, and removed the spy line pipe, and loosened the oil supply pipe. It took a bit of coaxing to get the oil supply pipe off, it was really stuck! I then removed the two compression ells from the rocker shaft supports, and again loosened the support bolts. This time they came loose without incident and the shaft was finally free of the head. I removed the shaft from the head and proceeded to loosen all of the head nuts and remove them and their washers. When this was done, I hit the head with a heavy rubber mallet until it was free of the cylinder deck on the block, then lifted it off.

The copper head gasket separated cleanly, and I looked into the cylinders. Three were full of water (Boo!-Hiss!) and one was clean and empty (Yeah!). All of the pistons had stopped at nearly mid-stroke, that is halfway down the bores of the cylinders.

I should note here that I paid careful attention to the disassembly of the top end, in order to reassemble the piles of parts in the garage. As the pushrods were removed, they were numbered, so they could be returned to their respective locations. Since the lifters, pushrods, and rockers have been mated at the factory, it is not wise to disturb the wear patterns unless a new replacement is necessary. In the case of my 'puzzle engine,' I will have to put up with the excess wear and valve adjustments until the parts wear themselves into their new patterns. As near as I could tell, both units appear to be nearly identical in their internal parts.

Here are some items of note: 2' bore, 4' stroke, 18mm spark plug set in the block at the top of piston travel, but about 1 down from the top of the cylinder deck. The engine has about a 4:1 compression ratio. Mean compression pressure is around 50 PSI. The engine only develops 3 HP at 1,000 RPM so long as the compression on all cylinders is equal I won't worry about the low reading. The head is milled flat, with the heads of the valves protruding into the cylinders. The pistons are made of cast iron with flat tops. I didn't take them out, but the ones on the other unit only have three compression rings, no oil rings. The only real differences were that the quarry engine did not have the auto-start option and there was a slight difference in the valve lash adjusters.

At this point in time, I took a picture of the water on top of the pistons, and then got back to work. I removed the water from the cylinders, and mopped the rest out with a dry rag. Although the cylinder walls had some black, slimy residue on them, they looked relatively clean, there being no pitting or red rust visible. I wiped out the cylinders again and sprayed WD-40 and some Kerosene oil into them on top of the pistons. I engaged the hand crank into its receiver and attempted to turn the engine over. Two of the pistons, the end ones (1 & 4), moved, and two didn't. The center two (2 &. 3) were stuck fast in their bores. Then I took a four foot long piece of 1 hexagonal brass rod (see picture 1), and fit one end into the #2 cylinder and bopped the head of the stuck piston from the height of the cylinder deck. This was repeated in the #3 cylinder and then back and forth between the two stuck pistons until suddenly the crank shaft turned quite easily. Now turning the engine over with the hand crank was a cinch, with no sticking or binding evident. I wiped the cylinders out again and noted that they were quite clean and had apparently suffered no damage from the water that had been present.

I checked the pan I had left below the drain, and it had about half an inch of muck in it, but no more was coming out of the crankcase. Then I removed the screen filter from the drain and found it was pretty clean. I flushed about a gallon of clean kerosene through the upper open areas, and collected it as it ran out of the drain. There was virtually no water in it, and so I replaced the screen and drain plug, proceeded to fill the crankcase with SAE 10W-30 oil. The pan took two gallons of Shop Rite oil! I had my son Andrew turn the engine over with the hand crank until oil began to flow out of the oil passage in the cylinder deck. At this point I took over, as his arm was tiring, and I cranked until we saw no more water come out of the supply port, we only lost about a pint, most of which we saved and separated for reuse in one of my other engines. I took another picture of my son turning the engine over, and then proceeded to lap in all of the valves. This could be done without pulling the valves out of the head, so it didn't take long. All of the valve faces cleaned up, just using a fine lapping compound, the valves being in good shape.

I reassembled the top end, using all of the old parts. Even the head gasket was in good shape! I purposely left the rocker cover off so as to adjust the valve clearances. I set them at .010 intake, and .012 for the exhaust, following lawnmower practice.

I then had my son help me check the mag for spark (this only works one time, you know), by holding onto one of the ignition wires as I turned over the engine. When I heard a loud snap and saw him jump, I knew that it had spark. Andrew said 'That was neat Dad, can we try it on Mom?' I said, 'Not right now. She's still mad from the last time.' (I had asked her to pull one of the low tension wires off of an engine to stop it, but she had grabbed the high tension instead and got bit pretty good.) Well, we put in the 18mm plugs after cleaning them off, primed the carb, and turned the engine over. After the second turn it fired up for a few seconds and then quit. We reassembled the rest of the parts onto the engine, filled the tank with gas and tried to restart the engine. I cranked until I was blue in the face, but the engine wouldn't make a peep. Then Andrew primed the carb. The engine started, ran a few seconds and then quit again. I took apart the Zenith carb and found it was full of dried debris and the float was jammed in the closed position. It was soon cleaned out and reinstalled, ready to go. The engine started on the first turn and quickly got up to its governed speed. I hooked a 110 volt trouble light to the wires on the control box, and when I threw the switch, to my surprise, the bulb lit brightly!

After taking another picture of the quarry generator running, I put in a call to Jim at his home in Sussex, and asked him, 'Hey Jim, how would you like to buy a nice little Kohler 4-cylinder light plant to take to the shows? It starts real easy, and it's cheap! Listen to this!' and with that I pulled on the crank and the engine fired within half a turn of the crank! I held the travel phone up to the running outfit to let him hear it run. 'Well I'll be a son of a b,' he said, 'you actually got it running?' I heard some other expletives and then he hung up the phone. From the time it left his house, until it was running, was just under five hours including the time it took me to get back home!!!

The only item that I could not repair here at home was the broken nose bearing. I felt that a professional at welding cast iron should repair it. I could have brazed it myself, but I thought a weld would be stronger in this application. I decided to exchange it with the one on the 'puzzle engine,' only to find after half an hour looking at parts, that the nose bearing for that unit had also been broken. I later learned from Jim that both he and Henry had stood on a three foot pipe wrench on the starter, trying to free the stuck pistons when the nose bearing broke. I took both pieces to a local welding shop in Denville, New Jersey (now closed), and asked how much the repairs would cost. About $15.00 a piece, the man responded. I said, 'Fine. When can I pick them up?' The man said they would be ready by tomorrow night, come back and pick them up then. Well, the next evening I went over to the welding shop to retrieve my parts, and got a surprise. One of the bearings was finished (the one for the puzzle engine), and the other wasn't touched. The welder said, 'It'll cost you $15.00 just like I said, for the one I fixed, but I lost money. They used such a crappy metal in casting these pieces, it took me five hours to do this one, and I wouldn't touch the other one for $100.00!' The one he did looked like brand new. You couldn't tell where it was welded after it was installed (see picture 2). I located another welder to repair the other bearing, but when I got it back it was a mess. I took it to a machinist friend of mine, Doug Kimble, and asked him if he could repair the damage to the mounting threads, where an arc weld had fused several threads, and cut a new mounting face on it, because the old one was severely undercut during the repair. Several days later, Doug dropped off the machined part and I installed it and had the crank back onto the engine.

I ran the engine for about five hours, and drained the oil. It looked pretty clean, so I saved it and used it up in my hit and miss engines at the shows. I refilled the crankcase with SAE 30 non-detergent oil, and after taking it to a couple of shows, and taking a few more pictures, it was sold at the 1994 Hudson Valley Auction. On a side note, a fellow club member was interested in the David Bradley tractor and that, too, was sold (making my wife a very happy person). In the meantime, I started to take interest in my upcoming project, the puzzle engine and generator set.

I took another picture of the quarry engine, copied the information from both units, and sent a query into GEM's Reflections column. My sincere thanks to Mr. Wendel and all of the people who responded to my questions about the Kohlers (29-6-11), in the June '94 issue. I did make one mistake when I sent that information in, the 1923 unit had the American-Bosch magneto, and the 1920 has an Eiseman magneto. If you think that the first part of this story was complicated, you isn't seen nothing yet! Imagine being given about 1,000 parts in a pile and without any guide, except your general knowledge, you have to make something out of it. That's all I had to work with. I had a 1,000 piece puzzle to solve, and the first thing I decided to tackle was the magneto.

The magneto is an Eiseman model GL-4 that is directly coupled to a brass splined assembly at the rear of the engine speed governor. It is a simple, two-pole magneto, with no impulse, that has a built-in spark distribution system to send the ignition spark to the correct plug

After locating all of the mag's missing parts in three separate boxes, I reassembled the magneto and mounted it in a vise on my workbench. When I spun it, I found that it had no spark. I checked to see that I was turning it in the proper direction (I was), and still no spark. Then I pulled the front of the mag apart again and checked the point gap (.016), cleaned them off and checked over the mag one more time. I noticed a small brass screw this time around that I hadn't seen before, because it was under the horseshoe magnet, and it was loose. In order to get at the screw, the magnet had to be removed from the magneto frame. I removed the two brass screws holding the magnet to the frame and lifted it off. I noted that the magnet was not very strong, and put a steel keeper bar on it. Then I removed the small brass screw by hand, in order to see what its function on the mag was. It turned out to be a cap screw with a spring centering post on it. The spring and the small, flat carbon brush that was supposed to be in the hole were missing. In looking in the hole where the screw was, I saw a brass contact strip on the armature where the brush was supposed to ride. The magneto base and I took a ride to my local goodies supply the Rockaway Sentry Hardware. I found an old drawer at the rear of the store that was full of old brush and spring sets. I only looked for about two minutes and found a set that fit perfectly. Fifty cents later I was on my way home.

When I got back home I put the magneto back together, with the exception of the magnet. At this time I checked the magnet again it was very weak. I could lift it off a ' thick steel plate I have, with just two fingers! I own a magnet charger that I use as part of my display of three Fairbanks-Morse Home Light Plants at the shows (see GEM April '86, page 2, 'Let There Be Light'). Even with the 36 volt battery pack providing 60 amps, with the outfit making power, the magnet charger still draws an additional 40 amps off the generator. Man, you should hear that Fairbanks bark! Well, I removed the keeper, and placed the magnet on the charger lightly, in order to check the magnet's polarity. The poles on the magnet were marked at the factory, as well as the pole positioning on the magneto base. The poles on the magnet charger are also marked, so I make sure that the switching wires are always hooked up the same way, as not to reverse the polarity on the charger. Fortunately, the magnet polarity was correct. I have found several cases where the polarity on a horseshoe magnet has been reversed, but the metal in these magnets has a memory, and the magnets usually weaken in a short time.

I started the light plant up and set the generator at full current. After letting the light plant recharge the battery pack from the starting cycle, I sent a current through the magnet charger for about 30 seconds. During this time I rapped the magnet sharply with a brass drift pin, to help the iron set its lines of magnetic force in alignment. The current was cut and I allowed the batteries to re-energize for about a minute, when the process was repeated. This was done three times. I cut back the charging current on the Fairbanks so it would not cook the batteries, and then placed the steel keeper across the magnet again (see picture 3). I took the magnet off the magnet charger and placed it on a piece of ' steel plate I have on my work bench, and removed the keeper. I should note here that it took a good pull to remove the magnet from the charger, even though the keeper was shorting the magnetic field across the poles. After removing the keeper, I tried to lift the magnet off of the ' plate, by lifting the magnet straight up. It took both hands to lift it off. This was one strong magnet!! I reset the magnet on the charger and went through the charging procedure one more time just to be sure, and again put the keeper on the magnet. This time I placed the magnet onto the magneto base in its proper position, and slid off the keeper bar. The two brass x20 screws then were reinstalled to anchor the magnets in place.

It's important to use a keeper after the magnet is removed from its mount. Some magnets will lose up to 70% of their magnetic strength, if a keeper is not used, and in the case of the Fairbanks AB 33 &. 34 mags, the magnetism will actually swap sides, rendering the magnets useless.

I set the magneto back in the vise and turned the armature drive. The resistance due to the cutting lines of magnetic force was much greater, and the mag produced a thin, blue-white spark about ' long. Spinning the mag with a drill opened the spark to about ?' but that didn't affect the quality of the spark any. As I found out later, an ominous sign of things to come. At this time I devoted my attention to the 4-cylinder engine itself.

The first thing I did was to set the engine base right side up (it had been dumped unceremoniously, upside down on a pile of other tangled parts). I cleaned all of the mouse garbage out and flushed the base out with clean kerosene, which I set aside. I noted that something should have been mounted inside the base (there were tapped mounting lugs all over), and that the interior had been primed thoroughly with a red primer paint. I also noted that there were no crankshaft supports. (Boy, I'm in trouble already!) I decided at this point to lay out the entire puzzle by group, as best I could, on the garage floor, in order to see how assembly of this 1,000-plus piece puzzle was to proceed. I placed cardboard on the floor of my garage, and then proceeded to lay out parts roughly in their assembly order.

In the layout process I had emptied the five boxes of parts, and the peach basket, as well as rearranging all of the other larger pieces that didn't fit. All of this work (some of this stuff is heavy!), took up the better part of the day, and when I was finished, also about a third of my car garage! I took a picture of the laid out parts, and went to bed.

The next two days were spent in sorting the good and bad parts, and in general, cleaning up the existing pieces. Most of the internal parts were covered with a coating of a thick, waxy, dirty grease. I bought a cheap plastic cement mixing pan from my friends at the hardware (these guys have everything!), and filled it with most of the small and medium sized parts, as well as two quarts of concentrated GUNK engine degreaser and the leftover kerosene from cleaning the engine base. After a lot of scrubbing, the parts were blasted as clean as possible with a high pressure washer supplied by Rich and Walt at, you guessed it, The Sentry Hardware in Rockaway. The parts were then blown dry with air, hand dried, and finally, coated with either fresh kerosene or WD-40, to prevent rust. When all of the washing, drying and coating were done, the pieces were returned to their respective assembly layout areas on the floor. The larger parts received the same treatment, with the exception of the crank and rotor assembly. When everything was squeaky-clean and coated, I did some serious checking of the parts. I took another picture of the cleaned-up parts and got to work.

The block was in good shape. The cylinders were not scored, they just needed a light honing to clean them up. This was done, then the block was washed off to make sure there was no grit left inside. The cylinders were then sprayed with WD-40 to protect them from rust, before assembly began.

Next I looked at the pistons and rings (see Diagram A). The pistons are 2' in diameter, 5' tall, with the diameter wrist pin set about a third of the way down from the top. The wrist pins are full floating, kept in place but not anchored, by a x 20 set bolt and locknut. Each piston had only three compression rings, no oil ring. Although there was no rust on the pistons, the rings (the few that were left), were stuck tight in their grooves. With a lot of tapping with a plastic hammer and a lot of WD-40, I got the rest of the rings out of their grooves, only breaking one. With the rest either missing or broken, I needed an entire new set. I cleaned all of the carbon out of the grooves with one of the broken rings, and set them aside after spraying them with WD-40.

I located the wrist pins, connecting rods, the bearings and checked them out. The pins and big end bearings were in good shape. The pistons and connecting rods are marked by cylinder and an arrow points toward the front of the engine, making my job of assembly much easier when installment time comes (see Diagram B).

At this time I called Otto Gas Engine Supplies and asked about buying a ring set for the engine. He gave me a price that I thought was not unreasonable, when he asked me what the rings were for. (I had only given him a piston diameter and ring width when I had asked for the rings.) I told him that they were for an early Kohler 4-cylinder light plant, and he said that I was in luck. He had a new-old stock complete ring set for the Kohler right in his shop! Then I asked him if he knew of anyone with a complete gasket set. He said, 'I just might!' When I asked him who it was, that I might contact him, he said, 'ME,' and that was all. I said, 'Me?' and he said, 'yes, ME!' He went on to explain that at one time he had had several kits at one time, as he wanted to rebuild several of these light plants. Over the years though, he had sold one unit and had parted another out. I ordered the gasket and ring sets (and I got a better price too), and asked if he possibly had a rocker support from his parted out engine as well. While I was waiting for the parts to arrive, I began to assemble what I could onto the engine block.

I installed the magneto drive assembly as well as the governor and its related parts, no mean feat as the parts were scattered all over counter weights to one side, shaft on another, gear up front, and washers and thrust bearings all over. It must have taken me over two hours just to assemble the governor itself! Next came the lifters and the camshaft. I cleaned out the lifter bores with a clean rag, and used some Bardahl assembly oil my dad had left me to lubricate the lifters, cam lobes and bearings. This stuff is like STP Oil Treatment, thick as honey, dark brown in color, and it won't come off! The lifters weren't appreciably worn, and even the cam lobes were looking like new. When this work was all done, another day had bitten the dust. At least I had another three feet of floor space back in the garage.

The following day was spent entirely on the head. Although the valves are different between the exhaust and intake, the springs looked the same, all eight of them! Some were weaker than others though, and I separated them according to strength. I must have sat for an hour squeezing springs. I'll tell you, I'll have the strongest fingers in town for a while! The weak springs went on the intake valves, and the strong on the exhaust according to lawnmower practice. I had no other guide. All of the valves looked to be in good shape. They were lapped in with coarse and then fine lapping compound, by hand. The head and the valves were rinsed off with kerosene, blown dry, wiped off with a dry rag, in order to make sure that all of the lapping compound had been removed (see diagram C). The valves, springs, retainers and automotive style keepers were then installed, after being coated with assembly oil. I took a photo of the completed head, and called it a day.

The crankshaft was next on the agenda. I first wire brushed the rust off of the journals with a fine brass hand held brush. I noted some fine pitting on several journals, and I proceeded to clean them up with some 600 grit emery cloth. When the pits appeared to be gone I took a clean rag, and using some DuPont #7 auto polish, I polished the journals until they were shiny bright. Then I took each connecting rod and tried it on its respective journal to see if there were any high spots. Unfortunately for me, there were a few. It took the better part of two days to smooth them out. The crank itself rides on only two mains and actually hangs suspended from the block! The mains are located at the very front of the block, and at the rear, about ' from the end of the block. The crankshaft has no oil seals at the front none is necessary because the crankshaft does not extend out of the block. At the rear a curved slinger is used to control oil loss. The tail bearing on the end of the generator shaft is a ball bearing. All four rods ride on separate journals on the central section of the crank. #1 is at 0° , #2 and #3 are at 180°, and #4 is at 360° (see diagram D).

At this time I blew all of the grime off, and out of the rotor armature assembly. The armature segments were then cleaned with some super fine sand paper, and the segment mica was clearanced, too. The rotor shaft bearing was washed out with kerosene and was repacked with bearing grease. All of the crankshaft journals were then sprayed with WD-40 and I had to wait for parts to arrive.

About a week later a package arrived from Otto Engine Supplies. Inside was a Hastings complete ring set for the Kohler 4-cylinder engines, and a N.O.S. Kohler engine gasket set. There was also a note, saying that apparently the rocker shaft supports had been sold previously, and he didn't know who might have a replacement. Well, two out of three isn't so bad!

I found out that the ring set was made for the later engines with four rings per piston (three compression, and one oil ring). I just installed the compression rings as per the instructions supplied with the set and they fit perfectly. The pistons were set on their respective rods, the wrist pins and their retainers and lock nuts were installed. The steel wrist pins fit snugly in the piston pin bosses, and the bronze wrist pin bushings in the rods felt like the fit there was good too. The wrist pins were liberally coated with the Bardahl oil, the pistons and rings given a liberal coating of WD-40. The individual rings on each piston were set at 120 degrees apart, and the assemblies were placed on newspaper and set aside to await installation.

I put the engine block on a clean piece of cardboard, on the cylinder head deck, the inside facing the ceiling. The pistons were then fed into the cylinders from the crankcase side, as the big ends of the connecting rods will not fit through the 2' cylinder bores. There is a large gentle taper at the base of each cylinder that acts as a ring compressor. As the piston is pushed into the cylinder, the rings are slowly compressed into the ring grooves, with only a little coaxing. Slow and easy does it!

My son Andrew and I then placed the crankshaft-rotor assembly onto the block, and as he balanced the rotor, I installed the bearing caps on the block mounts. It took several tries to get the correct fit, as shims had to be adjusted in order to get the bearing clearances correct. Only the removal of a .002 shim was necessary to make the fit right. A small crate was placed under the rotor to take the weight off of Andrew's arms; it served to balance the assembly while we installed the connecting rod caps. Fitting the rods took some time to accomplish, and we quit for the night after all of the rod and crank bearing assemblies were coated with the assembly oil.

 The next day I turned my attention to the oil pan crankcase base again. In studying the inside of the sump again, I noticed that there were actually two sets of mounting lugs. I suddenly realized what they were for one set (the lower one) was for the plunger driven oil pump; the second set was for the mounting of the connecting rod oil pick up pan, which is suspended above the oil reservoir in the base. The first item to be installed was the oil pump itself. I located the pump body and the rest of the related parts in a small box, among all the other stuff on the garage floor. I looked over the parts and reassembled the pump. The pump was primed with 10W-30 oil, and was mounted onto its place in the base. The pick up pan was then wiped clean and was also bolted into position. The oil pump pushrod, which is driven off a lobe on the camshaft, was placed in its shaft way in the pump, anchored there with a liberal coating of heavy grease.

I used 3-M Black Weather Strip Adhesive to anchor the base to block gaskets in place. I made especially sure to keep the adhesive away from the oil transfer ports through the mating flanges. It wouldn't do to seize the engine with a full crankcase of oil, would it? (I will detail the oil flow through the engine later.) The trick to putting the heavy base onto the block is to have at least two people. To this end I again asked my son to help out. The top of the gasket surface was given a coat of the 3-M adhesive and Andrew and I then inverted the 60 pound base over the upside-down block (see Diagram E).

I had previously made up a support rod out of a piece of  ? threaded rod, bolted to opposite sides of the bottom of the base.

After taking the entire weight of the base myself, with the help of the rod, I stood straddling the block, while my son Andrew helped align the oil pump shaft (Okay Dad, down a half!) and then the base mount holes. The base uses four ? bolts, and twenty 5/16 bolts to mate up the flanges between the base and the block. When the pump rod was in its proper spot, I lowered the base to within ' of the block, and Andrew then started the four ? bolts about two turns each, after a few positioning corrections. When the fourth bolt was caught, the base was then lowered down and the bolts snugged up, finger tight. I then removed my support rod, and with Andrew's help, we set the engine right side up on its base. We each then took a side, and started the 20 small bolts. At the same time we checked the face alignment of the cam gear cover, the gasket, and the generator mounting surface, and after a few minor adjustments with a heavy rubber mallet, we tightened the bolts to 15 inch pounds of torque, with a ? drive torque wrench. I took another picture of the now partially assembled puzzle and started to work on the front of the engine. In the meantime, I had Andrew again place the crate under the rotor in order to take the 75 pound weight off of the rear main bearing.

Andrew and I turned the camshaft to the point where the lobes for #1 cylinder were in between the intake and exhaust strokes (see Diagram F). The crankshaft was then turned to bring #1 piston to top dead center. The cam gear was installed on its keyway, and with a slight movement, the gear itself was mated to the crank gear and the governor drive gear. As the gears were meshed, the cam gear retaining nut was snugged up. The crank was then rotated 360 degrees, and the timing marks on the gears lined up perfectly. I found that the governor gear also had timing marks, and by pure luck, or by the grace of God, these marks also were in alignment at TDC! The cam gear retaining nut was then tightened, and the retainer folded to keep the nut from backing off. Andrew noticed a hole in the camshaft end that I hadn't paid any attention to earlier. It was about 2' deep and about inch in diameter. He asked me what it was there for, and I replied that I didn't know, but we'd soon find out! As it turned out we found out very quickly. When we looked inside the timing gear cover, we found a spot that was polished to a mirror finish, exactly opposite the hole in the cam gear. We looked at our dwindling pile of puzzle pieces and my son spied a piece of red material that looked like a dowel, that would fit the hole. The rod looks like it is made of the same type of material that is used to make the old red distributor caps. It is about 1' long and ' in diameter, with one end polished, and the other worn around the outside edge.

We placed it in the hole in the cam, and it fit perfectly. The only trouble was that it went totally inside the cam! I put some bearing grease on a pencil's eraser end and used it to retrieve the pin. In a few minutes searching on the floor, I found a fairly strong spring that was nearly the same diameter as the pin. I placed the spring in the hole, and then the dowel pin. The pin protruded about ' out of the cam and when pressure was applied to the end, the cam was pushed back against the face of the thrust bearing behind the cam gear. Problem solved! The bake lite dowel (if that's what it is), acts as a thrust bearing keeping the cam in one place. We applied a thin film of 3-M Blue HI Temp Silicone RTV sealant to both sides of the timing gear case cover gasket. The cover was set in place, the bolts snugged, and after double checking everything for alignment, were then torqued to the 15 inch pound limit I had set, in a diametrically opposed pattern (see Diagram G). This was done to prevent the gasket from walking out of place during the tightening sequence. We then quit for another night.

The following day was to be a busy one! Andrew and I started working on the top end. The copper clad head gasket was sprayed on both sides with Form A Gasket Head Gasket Adhesive. After the time specified on the can, the gasket was placed over the head studs, onto the cylinder deck. You have to make sure that the head gasket is placed properly, both 'up' and 'front.' The head gasket can be flipped the studs still line up, but the oil passage will be blocked, and some of the water passage holes will extend over the lifter ports, allowing water to enter the crankcase. The head was then set in place, the head nuts and washers snugged and then tightened in three stages, to 35 foot pounds of torque. At this point, several of the head nuts refused to torque further, due to stretching of the studs. I decided to stop here, so as not to either snap a stud, or to warp the head by torquing to different pressures. Also, I didn't know if new studs were available. The pushrods were placed through the lifter ports onto the lifters, and I got ready to install the rocker shaft. As I got ready to set the shaft, Andrew said, 'Dad, we still have to fix the support.' I had completely forgotten the busted one! Just as I was thinking about how I was going to fix it, my machinist friend Doug Kimble called, saying he had a problem with his plumbing. Well, since I am a plumber by trade, and he's a machinist, I said, 'Doug, let's make a deal you fix my problem, I'll fix yours.' I went over to Doug's and it took me all of two hours to fix his leaks. It also took him about two hours, and a block of aluminum supplied by Jim Quince, to make me a new oil outflow-support. A good trade I think, thanks, Doug! Just as I was leaving Doug's house I received a call from the Rockaway Police, saying that my middle son Robert had been injured in a bicycle accident. So a trip to the hospital put the kibosh on any further work this evening. Fortunately, Robert's injuries were not too serious, although next time he might listen when his mother says, 'No. Not one more jump!'

A few more days passed and then Andrew and I were back to work on the 'puzzle,' with a sidelined Robert as an interested observer! The replacement rocker support-oil outflow fit perfectly as I reinstalled the rocker shaft. I had Andrew tighten the mounting nuts, as I aligned the pushrods with their respective rockers. When they were all snugged, I torqued them to 20 inch pounds, and then proceeded to set the valve lash, using the same operations used on the quarry engine mentioned earlier in the article.

I had Andrew install the hand crank and the repaired nose bearing into the timing gear cover. At the same time I had Robert covering the ornate brass ID plate with a coating of grease. I took a picture, then gave the entire assembly two coats of Krylon gloss black spray paint. I should note that all open holes were plugged, and the cast iron valve cover had been dropped in place temporarily. After lunch we went back to the garage and took another picture. I then removed the rocker cover and installed the magneto mount bracket. I had Andrew put eight quarts of Shop Rite 10-30 oil into the crankcase, and had an idle Robert turn the hand crank until oil appeared in the oil outflow pipe on the rocker shaft. This is how the oil flows in this engine: as the engine turns, the cam moves a pushrod that drives the piston in the oil pump. Oil is pulled from the surrounding two gallons of oil past a ball check in the pump chamber. On the down stroke, the oil is forced past a second ball check into a copper supply line that is routed to a passage in the base. Oil then flows through this passage into a gallery that feeds the crankshaft mains and the camshaft bearings. A second copper pipe then carries oil to the top of the cylinder head. At this point, the oil enters the rocker oil supply pipe attached to the front rocker, through a swage connection (see Diagram H). Oil then travels the length of the rocker shaft, supplying oil to the rockers and valves. This oil gravity feeds the pushrods and lifters as it returns to the sump. The excess oil supply to the rocker shaft is forced through a restriction in the oil outflow pipe. As the oil leaves the oil outflow it is visible through an inspection port in the valve cover (see Photo 8). This oil then falls into the connecting rod oil pick up pan mounted in the base, where the rods pick it up to lube the big end bearings. At this point I might add that a hole had also been drilled in the top of the big end bearing. This hole supplied oil to the top of the connecting rod (wrist pin), and the piston skirt. Although there are eight quarts of oil in the engine, when the engine is running, about a quart must be in the air inside!

I had Robert turn the engine over until the #1 cylinder was at TDC compression, then I aligned the timing marks on the brass coupler with those on the magneto installing the same at its 3-1 firing point, as marked on the built-in distributor. The mag was then bolted in place. The Hi tension ignition harness was installed, and four new Auto light Blue-Crown spark plugs were put in, and connected to the wires.

The crate was removed from under the rotor and some gas was sprayed into the intake manifold. Andrew then attempted to start the engine. He gave the crank a mighty yank and promptly fell on his rear end, as the crank dogs had not engaged the crankshaft. He decided to try again, and this time made sure the crank was engaged. With the second turn the engine fired up and ran for a few seconds.

We were getting close to solving our puzzle! The brass Kohler carb was cleaned up and put in, along with the governor linkages. The exhaust manifold was painted with Hi Heat silver paint, and that too, was installed. Now comes the hard part. The field core housings had been previously painted and set aside to dry, and were now ready to install. As I held the 75 pound-plus assembly, Robert helped center the tail shaft bearing, and Andrew hit the housing with the big rubber mallet to set everything in place. The hardest part of this particular job was to align the four mount bolts with their respective holes in the block, while the entire operation takes place! After about 20 minutes and three or four tries, we finally synchronized our efforts, and got all four bolts started. When this was done, the brushes and springs were installed in their holders and the four bolts were then tightened to 30 foot pounds of torque.

Now another hard part. As fewer puzzle parts remained on the floor, it was harder to figure where the remaining pieces went. The nuts and bolts were easy, the electrics were not.

I found that the solenoid in the box was for the choke. Another mechanical switch assembly was mounted off the end of the governor linkage, to operate the starter circuit. I also looked over the wiring harness that mounted on the carb side of the engine. A few obvious connections could be discerned, the choke and the magneto kill. However, the other wires could be swapped, so some circuit testing had to be done. On the other end of the harness, I had no clue as to where they went. I decided to mount the auto start control box and to research the wiring in earnest.

Ah Yes! The Kohler auto start system what a pain in the butt!!! On my 'puzzle' unit there are 10 possible connections. Four terminals are on the base (24 VDC + and-), and 110 VDC output (unmarked + and -). On the 'front' left side there is a terminal block with four more connections, marked one through four. To the right, mounted on the box itself, is a combination fuse block, holding two fuses, and a two pole knife switch, as well as two more terminals (110 VDC output unmarked + and -). The control box mounts on top of the generator's stator housing, in front of the radiator. It is fastened to the mounting lugs with four ? bolts, which were a real pain to catch and tighten.

I did not have two 12 volt batteries with which to test the circuits, so I used a 12 volt charger I have, that hums loudly when a load is placed on it. I hooked it up to the plus and minus terminals on the 24 volt side of the control. I also decided to color code the four terminals at the side of the box. They are: 1-green, 2-yellow, 3-red, and 4-blue (see Diagram I). By using a voltmeter, I found: 1-dead, 2-dead, 3-short to ground (+), and 4-dead, with no other connections made. When the knife switch was closed, a loud hum came from the battery charger and a solenoid when a 200 watt bulb was placed in the output circuit. I pushed in the relay contact bridge, but nothing happened. I then had Andrew hold the bridge in, while I tested the circuits again. This time I found: 1-Hot, 2-Hot, 3-Open and 4-only 1 or 2 volts. I figured that #3 was indeed the kill switch. Number 1 was the choke power, #2 was the ground connection for the start erchoke relay, and #4 was the ground for the charging circuit solenoid. I colored the wires to match the terminals to ease reinstallation, should the harness have to be removed in the future.

The self starter operates as follows: First a load of 25 watts or greater is placed in the power circuit. When the knife switch is closed this completes a circuit between the 24 volt battery and the primary relay field. (Also, any switch in the circuit that closes it will work). This in turn closes the relay's contact bridge and does three things: (1) supplies power to the choke circuit (activating the choke coil and closing the choke); (2) opens the magneto kill circuit (allowing the mag to provide spark); and (3) sends power to a second relay that actually sends the 24 volt charge into the generator, causing it to act as a motor, turning the engine over. As the engine builds speed and begins to fire up, the governor begins to close the throttle. As it does so, the grounding contacts are broken and the switch then closes a circuit activating a third relay, at the same time disengaging the starter relay from the circuit.

As the unit starts making power, the third relay engages its contacts and it sends 110 VDC to the power terminals, and through a wire-wound resister, supplies between 26 to 28 VDC back to the batteries in order to recharge them. Any disruption of the power circuit, blown fuse all lights and load off etc., will cause the #1 relay to trip out, thus killing the mag. As the engine RPMs fall off, the charge power relay also trips out, putting you back to the beginning.

There are two other features that should be brought to your attention. First is a manually reset circuit breaker in the starter circuit. This breaker trips out if the starter circuit is engaged for more than two minutes continuously. It prevents an unattended outfit from draining the batteries dead if the unit happens to run out of fuel. It also prevents the armature from overheating due to the starting load. The second is that if the batteries should be dead, or removed from the generator, the ignition kill circuits can be disabled, and the unit can be manually started with the hand crank. There is also a provision to manually choke the carb, should it be necessary. A note of caution is necessary here. DO NOT RUN THE AUTO START units without a load on the power circuit, with the magneto kill disconnected. If there is no load on the power circuit, a disastrous voltage buildup may occur, leading to the destruction of the relays, and possibly anarc fire at the armature itself. Once the engine has started, the power solenoid will activate, and the 120 VDC will be available. If the batteries are still on line, they will be recharged. Reconnecting the mag kill-kill circuit will restore the unit to full operation.

I finished installing the wiring harness and its connections, and also installed the radiator and its related plumbing. The fan assembly and the oil cap was also painted and installed, and the flat belt that drove the fan was hooked up. The belt was worn, but I felt I could get a season out of it. I was wrong! After about five hours of running time, the laminated belt self destructed, and I had a 'bugger' of a time getting what was left out and the new one in. Another day had passed and still we were not finished!

I had Andrew help me install the Stewart vacuum fuel pump, and its related piping, and rigged a fuel tank for a test run. The Stewart vacuum pump utilizes engine manifold vacuum created by the engine to draw fuel into a sealed canister. As the fuel level rises in the canister, a float also rises, eventually tripping a valve to dump the vacuum to atmosphere. A check valve holds the fuel in the can, where it is gravity fed to the carb, as the level drops when the fuel is used, the valve is again tripped, starting the process over. The only problem with this system is that if the engine is under constant heavy load, and the fuel level is low, the unit may not be able to create enough vacuum to draw the fuel from the tank, and the engine will starve for fuel (see Diagram J).

We put the entire generator outfit on a small railroad skid that I had bought at a garage sale for $10.00. New wood and a paint job had this item looking like new. I bought two new Sears 45 batteries and mounted them beside the generator section of the outfit. The two 12 volt batteries do a good job of turning the light plant over.

I filled the fuel tank (about two gallons), and switched on the knife switch. This engaged a 100 watt bulb I had in the power circuit, and the unit began to motor over. After about a minute and a half of steady cranking, the Stewart pump finally supplied enough fuel for the engine and it sputtered to life and began to pick up speed. I heard the relays click and then the bulb lit! After all this time (about 200 hours by my guess), the puzzle was finally completed! (See picture 5.) Well, almost. I decided to have the power plant well broken in by the time I took it to the local shows, and to sort out the bugs if any. If only I had known! First, after five hours the fan belt broke as I had mentioned before. After about 25 hours of running time I changed the oil in the crankcase. I didn't waste the used oil as it was quite clean, I just used it in my other hit and miss engines. As the engine was run further, I noticed that it was turning over slower than it used to. At first I thought it was just the compression, but one day I noticed that the mag was quite warm to the touch. Finally, one day, the auto start would not turn the engine over. After about two minutes I determined that the mag had seized! I removed it from its mount and the engine spun freely. The mag was stuck tight.

I took the mag apart and found a paraffin-like substance all over the armature and the case interior. I cleaned all the gunk out with gas and a dry rag, and reassembled the mag. I had the engine running again in less than an hour! Twenty five minutes later, the engine slowed and quit again, laboring all the way to the stop. AW NUTS! Now what, I thought! As it turned out, it was the mag again. I pulled it apart a second time and it was full of gunk again! I cleaned it up again, and left the parts out to dry overnight.

The next day I rebuilt the mag and reinstalled it on the engine. At this time I could only get a weak ? inch spark to ground. I reset the points, thinking that perhaps I made a mistake in setting them, but learned quite quickly that this was not the case. After about five minutes running, the mag got real hot, almost too hot to touch, and suddenly smoke and sparks issued from under the distributor cap and the points cover. The engine stopped in about two seconds with a grunt, and the mag was locked up again. This time when the mag was opened up, I could see that the armature was toast. It was fried to a crisp. I put in a call to my good friend John Corigliano, who does magneto repair. John said, 'Bring it over, and we'll have a look.' John has been rebuilding mags for over 25 year she has all kinds of neat stuff in his shop, only he didn't have an armature for my Eiseman. He said, 'Wait here,' and proceeded to make some calls. He called a place in Massachusetts, no luck. Two places in Ohio, the same result (one guy said, 'fat chance'). Next went a call to Nebraska. The gentleman there said to call so-and-so out in Michigan. Voila! He has one, but he doesn't know where it is, and it lists for about $280.00. Woof!!! John said, 'Well, if you find it call me back,' and hung up. John knows most of his suppliers on a first name basis, so he knew the man would call if the part was found. John then told me he would keep his eyes open for the armature, but I should find out why the armature decided to eat itself. I told him what had happened to the mag before it fired, and he said, 'Oh, I know these things, check the drain cock on the Stewart reservoirs, the bad leak gas onto the mag, it gets inside and eats the shellac out of the windings in the armature shorting them out. Screws them up real good.'

'I know,' I said, looking at my box of mag parts in my hands.

I left John's shop, wondering where in the blazes am I going to get another mag like this one? Two days later John gave me a call. The man from Michigan had called back, and he had found the armature for my mag. John also talked him down a lot in price under the condition that I would not tell anyone what I actually paid, and that I would return the old armature as part of the exchange. I agreed, and three days later the core was over at John's shop. I went over to his place and we disassembled my armature and installed the new core. I took the assembly home and again rebuilt the mag. I also recharged the magnet again, in case the heat generated when the mag quit had affected the magnet.

I didn't bother to bench test the mag (my mistake), and I remounted the mag back onto the engine. All the wires in place, switch on, light switched on, fuel on, and . . . nothing! Absolutely nothing! The engine turned over, but it didn't give a hint of ignition. I double checked everything, and found that the mag had no spark, although it was slightly warm again. I called John and he said to bring it right over and we'll take a look at it. Thank God he lives close by. John mounted the mag onto his testing machine, and after about 10 turns, he said, 'Shorted.' He took the mag apart and checked the wiring I had done. 'Looks okay,' he said. Then he took out an old rectifier probe to the contacts, and the old unit read 'Short.' John clipped the connections to the core and began testing each lead. Lo Tension to ground open, okay, Hi Tension to ground open, okay, Hi Tension to Lo Tension open, okay, Lo Tension to Lo Tension Lo Resistance, okay, Hi Tension to Hi Tension Hi resistance, okay. 'Darn, it checks out, what now?' John asked. I said 'For the heck of it, why not check the condenser?' John said that these condensers never go bad (he hadn't had to change one in over 20 years, they were that reliable!). Well, we checked this one out. Case to ground short, okay, lead to ground short-fault! 'Well I'll be!' said John. He took the condenser out of the armature and double checked it with the same result: the condenser was no good! John then checked his parts inventory, and located three condensers like mine. Just to be safe, we decided to check them, too, to make sure that they were good before installation.

Well, the first one checked out with internal leakage, it would not hold a charge. The second one checked bad too. The third one checked bad too, with an internal short, like mine! I went home with my box of mag parts again, as John tried to locate another condenser for me.

After another week, John called to say he had located another condenser in a junk mag he had acquired for parts. I practically flew to his house to pick it up! 'I just hung up!' he said. We again rebuilt the mag using the used condenser (which had checked out okay), and set it on his testing rig. By turning the mag by hand, it gave off a fat blue spark for each cylinder. Spun at 500 RPM, the spark jumped a Vi' gap, but no more. John explained that the mag had an internal Vi' gap built in in order to protect the armature core from building up too much secondary voltage. A combination of the gas and the bad condenser had caused the core failure. While the mag was still off of the engine, I relocated the piping on the Stewart unit so the fuel would not drip on the mag, should it leak. Now if the carb doesn't leak too, I'll be okay. The mag was then reinstalled and the outfit was test run perfectly.

After about 100 hours of running time, I still have two problems with my Kohler outfit. First is an inherent problem with the carb. In humid conditions, the carb builds up ice on the throttle body inside. This appears to be a design defect. Later units use a different carb and do not have this problem. The other problem relates to the Stewart pump. I believe that the switching valve may be worn, and is not seating properly all the time. Once in a while the valve doesn't allow the pump to draw fuel and the engine quits for lack of fuel. When the auto start feature activates, the pump will then cycle and the engine will restart and run normally.

I am going to try to replace or rebuild the pump soon, as I have been given several possible parts sources. My most sincere thanks to all of the people who responded to a query I had posed in the Reflections column in the June issue of the GEM.

As sort of a postscript to my article, many of you may be wondering where are all of these pictures he took? Well, I am too. I took two 36-exposure rolls of film to my local Shop Rite store for processing, and they never came back. I got two rolls of film for free, but I cannot reclaim the lost sights and memories. I have taken a few replacements, but they are a poor second to the originals. The few supplied were taken at my home in Rockaway, and at the Hudson Valley Old Time Power Association Show in August '94 in Hudson, New York.