Another Radial - this time 18 Cylinders

[mayhugh1]

I've been out of town and at my son's wedding for the past five days. When I returned, I decided to resume construction with the connecting rods instead of the distributors because I may have some unfinished machining to do on the main bearings. I may have to cut some additional openings in some of the main bearings to help with assembly of the crankshaft sections when the rods are installed, and I'd like to tie up those loose ends. I'm having trouble visualizing an order of assembly without the actual pieces in my hands. A lot will depend on whether the rod assemblies can be assembled onto the crankshaft sections outside the engine and then inserted in the crankcase sections or whether everything has to be assembled within the crankcase as was necessary with my 9 cylinder.
I started with the slave rods and a goal of 20 completed parts. I designed my own rods for my 9 cylinder engine, and I'm using the same design here. These rods have the maximum possible amount of material in their stressed areas and a worst-case .020" minimum clearance to other components of the engine when running. I gathered up all the half-inch known 7075 (approx. 75% higher strength than 6061) aluminum plate I had on hand; and, with no screw-ups, I should have just enough for the job. I very much wanted bronze bearings at each end of the slave rods, but doing so would have compromised their strength especially at the master rod end because they would have replaced significant amounts of high strength aluminum in critical areas.
I used the CNC program that I had previously developed but with a few minor tweaks to nest the parts among my scrap plates. Four arrays of either four or eight rods were machined at a time. The holes at either end were first drilled and reamed, and then the top halves of the rods were completely machined with roughing and finishing passes using 1/8" cylindrical and spherical cutters. I then glued the perimeters of the rods to their workpieces before flipping them horizontally in the machining vise and running similar machining operations on the opposite sides. These bottom-side operations cut the rods free from their workpieces leaving only the cured epoxy gel to hold them in place. The parts were finally released free of their workpieces by heating them to about 175F with a heat gun.
Due to the fairly complex cross-section and filleting, the total machining time for all the rods was about 20 hours. The spindle speed using my Tormach Speeder was 13k rpm, and this allowed me to use feed-rates up to 25 ipm. Without the Speeder the machining time would have been close to 60 hours, and I likely would have opted to greatly simplify and (maybe) compromise) the rod design.
It's with this part that the enormity of this project has struck me. It's one thing to need 20-40 pieces of something requiring 10 or 15 minutes of machining time each but quite another when the required machining time for each part is an hour or more.
After the parts were free of their workpieces I made a simple fixture to support them while oil holes were drilled at each end. These holes actually have tapered lead-ins to encourage oil entry and so their sectioned profiles look like tiny funnels.
The resulting offset between the top and bottom finished sides ended up being nearly zero along the y axis but was about .002" along the y axis. I planned on leaving the rods 'as machined' since they are internal to the engine, and the offsets gave them a realistic 'cast' appearance. But, I couldn't leave well enough alone and eventually polished out the offsets as well as the machining marks.
The next step will to be machine the master rods. Fortunately, I need only two of these. - Terry

 

[petertha]

Your epoxy backfilling technique is neat!
PS, I noticed you used a disposable syringe for spooge injection. I've probably done about 40 lineal miles of fillets & joints for RC composite layup work. A good (cheap) alternative is a Ziplock baggy. Put the epoxy premix into a corner, snip the end to desired 'nozzle' size, purge any air, then squeeze a controlled amount, similar principle to those cake decorator thingy's.

I've been paying attention to how various radial engine link rods are configured on their ends. I noticed many don't use bushings, they seem to rely on the aluminum (lubricity?) itself & sometimes in conjunction with oil holes. Yet in a typical commercial RC engine, its rare to see anything but bushing inserts. Maybe its an RPM thing?

I'd like to hear your design comments on the oil passage hole because I've noticed engines seem to vary a lot between no holes, to a single axial hole, to 2 holes at 45 deg either side of centerline, sometimes staggered... and even the occasional slit vs. hole. Mind you, I've confined my mental data-basing to methanol fuel with pre-mix oil.

 

[mayhugh1]

Peter,
Thanks for the tip on the epoxy applicator.
Your question on the oil holes is a good one, and I have pretty much the same question myself. I think splash lubrication of the wrist pin end of the rod (the so-called small end inside the piston) is probably sufficient without adding any type of hole. I say this because in my past I restored several Mustangs for myself, and this included two engine re-builds. One was a 289c.i. 1966 fastback. The rods on that engine did not have bearings on the small end but did have an oil passage for oil. When I purchased replacement rods for the rebuild (bearings had spun on big end) I noticed that there were no oil holes on the replacement rods from Ford. When I asked about this I was told that Ford engineers had determined that the holes did nothing and so they dropped them from later production. I also cannibalized a lawn mower engine recently and there was no oil passage or hole there either.
I've seen examples of model engines that relied on splash lubrication for both ends of the rods -Steve Hucks' recent Demon V8 for example. I don't see any reason why a radial would be any different at the small end. On my two engines there is about .025" clearance between the small-end rod sides and the internal piston boss which means there is .050" of exposed wrist pin within the piston and this seems like plenty enough area to collect oil for capillary action to draw oil between the wrist pin and the rod.
On the other end of the rod, though, there is only a thousandth or so clearance between the sides of the slave rods and the carrier portion of the master rod in my two engines. I think an oil hole on this end would probably enhance its lubrication since it would probably take considerably longer for oil to wick into the interface between the rod and the shaft on this end compared with the wrist pin end. -Terry

 

[rcfreak177]

Terry,

Great tip with the glue on the connecting rods, I must have broken over a dozen end mills using this method without the glue to support the part, not to mention scrapped many parts as well.

Thanks heaps, I will try the method you use.

Baz.

 

[kvom]

Rather than a 20 hour 3D 2-sided mill job, it seems that a 2.5D job followed by turning the shafts on the lathe would have worked. But then we wouldn't see this technique, which is "cool".

 

[mu38&Bg#]

It's very common to find unbushed little ends in RC engines. The motion is an oscillation, so the requirements are different than the big end. Many engines also have unbushed big ends if they can get away with it. The bushing weakens the rod end and increases cost. Oil holes in engines is almost like religion, but several methods work. Seeing as both ends oscillate in a radial and the speed is low, as long as there is some oil present it will be fine. In a high speed big end fresh oil is important in cooling and lubrication because the the oil shear is very high. Even the clearance between journal and placement of oil holes can be critical in extrmely high performance engines.

I think I would have redesigned for an I beam rod if I was going to build this. I try to cut machining time down to a minimum, even if it's just the CNC turning the wheels. Because I don't have quick change tooling, I use corner radius endmills to do milling and contouring without tool changes. It's always interesting to see different methods. I might have to try the epoxy on an upcoming project.

Greg

 

[mayhugh1]

The machining of the master rods began with two 1" thick drops of 7075. I bored a hole in each one where the crank pin center will end up, and there I Loctited a slug of SAE 660 bearing bronze. The diameter of the slug was chosen to leave a minimum .065" of aluminum around its outer perimeter (for strength) after the clearances for the slave rod ends were milled.<BR>The scrap extrusions I used for the slave rods were remarkably flat (few tenths), and so I didn't need to face them in preparation for mounting in the mill vise. The faces of these drops, however, were out of parallel by several thousandths, and so I began by machining their sides parallel. While I was at it, I finished their thicknesses to those of the finished rod widths.
The next step was to spot, drill, and ream all the holes. The center of the crank pin bore was then used to reference the part for all subsequent operations. The top sides of both rods were roughed and finished to one-half the rod's thickness and then back-filled with the epoxy gel. I tried Peter's 'cake decorator' suggestion for dispensing the epoxy into the narrow slot around the semifinished part, but this particular gel was just too thick to flow under the hand pressure I was able to exert on the polyethylene bag.
After letting the epoxy cure overnight, I flipped the parts over and machined the opposite sides using programs that were similar to those I created for the tops. Due to the large size of these parts, though, the outer perimeters were initially cut to a depth that left a .015" thick aluminum web connecting the part to its workpiece. This web helped support the part under the subsequent heavy roughing cuts. The roughing and finish machining was completed before a final perimeter pass was made to cut the rods free of the workpieces. As before, a heat gun was used to cleanly release the parts from the epoxy.
The semi-finished rods were mounted to a horizontal rotary under the mill spindle, where the clearances for the slave rods were milled. Watching the rotary turn back and forth as corner material was removed at various acute angles inside the perimeter of the rod was a lot of fun but also nerve-wrecking, especially on the first part.
My method for retaining the slave rod pins within the master rod assembly is a pair of 2-56 set screws bearing against flats milled into either end of the pins. The holes for these screws were
drilled while the parts were on the rotary and later tapped by hand. Two screws each may be over-kill but that's just me.
For some inexplicable reason I used a known oversized reamer in my shop to ream the pin holes in both the master and slave rods. With the on-hand 3/16" drill rod I was planning to use for pins, my reamed holes are now .0025" oversize. The maximum I'll accept for this clearance is .001" with .0005" being ideal, and so the next step is to work on the pins. - Terry

 

[DICKEYBIRD]

For some inexplicable reason I used a known oversized reamer in my shop to ream the pin holes in both the master and slave rods.

Wow, I'll bet some salty new phrases were coined after that one! Whatever affliction causes that kinda problem is alive & well in my shop too and when it happens I literally have to force myself away from anything throwable.

I'm sure you'll come up with a fix that'll be good as gold!

 

[gadabout]

Terry, maybe you could go to 5mm pins, thats only 6.9 thou up from where you are at present

Mark

 

[petertha]

My method for retaining the slave rod pins within the master rod assembly is a pair of 2-56 set screws bearing against flats milled into either end of the pins. Terry

Will you access the set screws with a typical 90-deg allen key from the front & rear at the assembly stage once the link rods are on the pins, or through a cylnder hole in the crankcase, or?

What are your thoughts about locktite on the setscrews?

 

[mayhugh1]

Peter,
The set screws will be accessed radially through the cylinder holes in the crankcase. I probably won't use Loctite here as I don't think it's needed and will just make disassembly more difficult if it is ever required. The set screws are 1/8" long, resting on flats, and aren't under any significant moments trying to loosen them.

Mark,
Great suggestion.

Terry

 

[mayhugh1]

A seemingly simple solution to the pins for my over-sized holes is to just turn them to the correct diameter from over-size drill rod. My personal experience, though, is that to get the accuracy I'm looking for I would have to uniformly polish off the last thousandth over the entire length of each pin. I'm OK with doing that for a few parts but not for some nearly 40 pieces.
Mark's suggestion of switching to 5 mm drill rod and re-reaming the rods is an excellent solution that just didn't occur to me. I would have taken his advice if, by the time I had read his comment, I hadn't already started down a 'different' path.
While searching through the MSC website I discovered it is possible to buy individual gage pins in practically any diameter. Even more remarkable, at least to me, is the fact that they actually stock them and in large quantities. For a little over $25 including my expires-tomorrow free shipping coupon, I was able to buy enough gages of the exact diameter (.190") I needed to make the 34 pins I need plus spares.
I received the gages and found they were, indeed, a highly polished and perfect fit. The only problem, of course, was that they were harder than the back of Superman's head. After some experimenting, I found I could anneal them to a reasonable state for machining using a soak at 1475F for 1-1/2 hours followed by a slow furnace cool-down. After annealing, they machined somewhat like air hardening drill rod - maybe a bit harder - but they did tend to work-harden with timid feed rates.
I needed two types of rod pins: one for the master rod assembly end and one for the piston wrist pin. The pins at the master rod end have a pair of milled flats for the set screws and a shallow 2-56 tapped hole at one end for temporarily attaching a simple installation tool. The piston wrist pins are full floaters with soft aluminum buttons (rivets) pressed into either end. This wrist pin design is from the H9 plan set and is what I also used on my 9 cylinder model. After pressing in the rivets, the soft ends were machined with a radius matching that of the cylinder.
All lathe operations but one were done manually rather than with CNC due to the large number of set-ups involved in dealing with each pin. Machining the pins from 2" long gages instead of a single length of drill rod significantly increased the total lathe time. Since the ends of the gages were ground at an angle (probably to aid their entry into the hole under measurement) they had to be faced. The flats on the master rod assembly pins were manually milled together using a simple holding fixture. The radii on the soft ends of the wrist pins were turned using a simple program on my 9x20 lathe.
The wrist pins were the most tedious parts to make and were worse to deal with than the tappets. My 1/8" drill bits all drilled oversize and would not give me an acceptable press fit of the rivet, and so I had to wait several hours for a Loctite slip-fit adhesive to cure on each end before machining its radius. And, because the un-machined rivet heads started out with a diameter larger than that of the pins, the un-machined ends would not fit into the lathe chuck. So, only one end could be Loctited, cured, and then machined per session. The total machining time for each wrist pin averaged about 45 minutes and had to be spread over a couple days to allow for two curing cycles per pin.
Using the gage pins was an interesting exercise; but, in retrospect, Mark's suggestion was a more practical and much more economical solution. The annealed gages remained hard enough to dull two HSS drill bits, two HSS center drills, and a carbide parting insert over the course of making some 45 parts which included several rejects and a few spares. Looking for a positive, though, the finished parts didn't need to be heat-treated. Using the parts I've made so far I was able to sketch out an assembly process that I think will work with the way I've divided up the crankcase and crankshaft. I've had my fill, though, of these high quantity tiny tedious parts for a while and am looking forward to finally getting on with the distributors. - Terry

 

[Rivergypsy]

Very nice work ;D

 

[necchiom]

Amazing! With compliment!

 

[mayhugh1]

This distributor design is one that has evolved over my last three engines. My requirements for a distributor have always included not only a realistic appearance but a high level of reliability especially with respect to the Hall-effect trigger. In this design the Hall device is safely tucked away inside the bottom of the main housing and is shielded by a 'cage' created by the aluminum main housing and the brass trigger disk. For additional protection a dielectric spark shield is located between the Hall trigger and the lightning storm going on above it.
The Hall device can be rotated around the center axis of the distributor up to 30 degrees and then locked down even after the distributor is assembled and in place on the engine. This feature allows fine tuning of the ignition timing while maintaining a particular orientation of the distributor. Such capability is important on this engine since the distributor mounting flanges must be nearly perpendicular to the crankshaft axis in order to avoid interference with the intake pipes on the rear cylinder row. The limited number of teeth on the distributor gear allows a minimum increment of only 20 degrees for a rough adjustment of the timing. The distance between the Hall device and the center axis of the distributor will be determined later by the dwell requirements after some experimental measurements are performed using an actual trigger disk magnet.
This distributor design also allows a minimum gap between the rotor and tower electrodes without requiring tight control over the rotor's vertical clearance position. This gap should be as small as possible with respect to the .018" plug gap so that as much of the ignition energy as possible is available to ignite the air/fuel mixture instead of being wasted while burning up the rotor and tower electrodes. The fractional portion of the wasted ignition power is approximately given by the ratio: (rotor gap)/(rotor gap +plug gap).
Actual construction began with the distributor shaft. It was turned from drill rod and polished but was not hardened due to concerns about warping. The bottom of the distributor housing was turned from a 2-1/2" round aluminum drop, and the bore for the distributor neck was done in the same lathe set-up. The neck was turned and bored in a 4-jaw chuck separately from the main housing and then pressed into it. Stainless steel rather than aluminum was used for the neck for its durability against the set screw that will be used to lock the distributor position to the rear cover. A peripheral groove was cut around the neck to accommodate this 6-32 set screw. The neck was turned from 303 stainless after a phosphor bronze slug was pressed into its bottom end.
The distributor shaft is supported at its lower end by the bearing machined into the phosphor bronze slug and at its upper end by a 1/2" ball bearing. A stainless bushing, drilled and tapped for dual set screws, was turned and pressed into the the distributor gear. The machined face of the bronze slug acts as a thrust bearing in conjunction with the polished face of the gear bushing and limits the upward vertical travel of the distributor shaft. The downward travel is limited by the magnetic disk which is screwed onto the top of the shaft while bearing against the inner race of the ball bearing through a spacer. The final thrust clearance is set when the distributor gear is assembled onto the lower end of the distributor shaft.
In this engine the lower bearing of the distributor shaft is lubricated only by the air/fuel mixture which will be pump gasoline. I have never been entirely comfortable with this aspect of the radial design, but I am not aware of any spectacular failures. An alternative that some Hodgson builders have used for the lower bearing is oil-impregnated bronze, but I'm not sure that fuel would eventually wash away the oil allowing the softer bearing material to quickly erode. Phosphor bronze has significantly better wear characteristics, and I made the bearing length some 3/4" long to hopefully help
add to its longevity. - Terry

 

[ShedBoy]

Been a while since I checked in and WOW. Truly amazing work.

Brock

 

[mayhugh1]

Construction of the distributor continued with the removal of the chucking spigot from the top of the distributor housing. The interior of the housing was then lathe-bored using a four jaw chuck with the X-axis indicated off a long and snug-fitting pin gage inserted into the lower bronze bearing from the top of the distributor. This wasn't the easiest surface to indicate, but the distributor internals to be concentric with the distributor shaft even if the bore through the lower bearing is not perfectly concentric with the rest of the distributor. I've had trouble in the past with drills wondering off-center in the (CDA 544) phosphor bronze material I used for the lower bearing. A carbide drill usually solves the problem, but I didn't have one of the proper size.
The interior of the housing was bored followed by a pocket for the ball bearing, and then a shallow recess was cut in the floor of the housing to clear the ball bearing's inner race. The final lathe operation was a locating ring for the distributor cap. When completed, the rotor will end up concentric with the distributor cap so I can maintain a minimum rotor gap.
The shafts were trial-fitted in the distributor housings with the ball bearings in place. I discovered that one of my shafts was slightly bent with some .004" run-out, and so I made a replacement. Much of my drill rod has come from machine shop drops salvaged many years ago from a local scrap yard where it was badly mis-treated before receiving the TLC it rightfully deserves in my shop.
The brass trigger disks and their stainless spacers were then turned and assembled onto the tops of the distributor shafts using 2-56 SHCS's. The distributor sub-assemblies were inserted into their bores in the rear cover, and the mesh with the crankshaft drive gear in the rear crankcase section was verified. Both shafts turned smoothly with no sign of binding when the crankshaft was turned manually. The real test, though, was a backlash check, and it was much less than a degree.
The Hall retaining disk, the spark shield, and the rotor were turned from white Delrin because of its excellent dielectric properties. The diameter of the Hall disk is a few thousandths less than the i.d. of the housing so it can be easily rotated later when fine tuning the position of the Hall sensor. The spark shield is a snug fit to the distributor housing. When finally assembled, the spark shield is tightly sandwiched between distributor cap and the distributor housing.
I have not yet decided on the ignition I will use. I have the parts for a pair of TM-6 modules identical to those used on my H-9, but The only coils I have are the new and smaller style versions that are
currently sold through Jerry Howell's website. These coils come with an ominous warning about multi-cylinder engine use that leads me to believe they may not have the same dwell/current capacity of the larger and older style coil I used with my H-9. The trigger disk for that engine was sized for a 10 deg dwell which is only half of what it should have been, but the ignition did perform well with just under 2 mJ output at 5000 rpm. If I leave the dwell at 10 deg the inductance of the new coil may be too low to reliably fire the the plug at higher rpms. If I increase the dwell, the power dissipation in the smaller coil may cause it to overheat at low rpms.
Another issue is the fact that the TIM-6 is dc coupled, and so the ignition components are prone to overheating if the engine is stopped with the coil current ON. With my nine cylinder engine I had 40 degree (prop) wide safe zones, and rotating the prop while watching the trigger led to find a safe rest position was not at all difficult. With 18 cylinders the safe positions are only half as wide, and the rotary forces from cylinder compression may tend to nudge the engine into unsafe rest positions.
I spoke to Roy Sholl at S/S Engineering who supplies turnkey ignition modules to the model engine community. He supplies a CDI ignition module (essentially, no dwell requirement), but the spec for his standard module is only 12,000 sparks/min. If this is a literal spec, the ignition would limit the rpm of a nine cylinder engine to just over 2600 rpm. Roy mentioned he can also supply an 'off-menu' 30,000 sparks/min module which would boost the rpm capability to 6500 rpm. It uses a smaller discharge capacitor to gain rpms, and so I need to investigate its mJ output capacity.
For now I want to finish up the distributors, and so I've decided to set the magnets into the trigger disk for a compromise 15 deg dwell. I made some measurements on a rotary table using the .078" diameter x .042" thick magnets I have on hand, along with the pig-tailed Hall sensors I've been using from Roy. I came up with .9" for the diameter of the magnet array for 15 deg dwell. I'll come back to the ignition after the distributors are completed, but Roy's 30k spark/sec modules may be my best solution.
The trigger disks were drilled and reamed for the magnets which were Loctited into place. The pocket for the Hall sensor could then be milled into the Hall retaining disks, as well as the slots for the Hall sensor and locking screw in the floor of the main housing. Milling of the distributor cap mounting flanges and peripheral timing indicator marks finished the machining on the distributor housing. The green stuff in one of the distributor machining photos is modeling clay that I used to damp the vibration in the housing while its peripheral was being milled. A reference mark was drilled into the top of each housing to mark the tower to which the rotor will be pointing during the 'number 1' cylinder TDC for the distributor's associated cylinder row.
The rotor electrode is a strip of .003" phosphor bronze with an .030" thick brass tip soldered to it. The electrode is pressed into a slot in the Delrin rotor disk with some super glue for extra measure. Next on the list is the distributor - Terry

 

[Lesgsy]

I've never seen work this good before wow!

Les

 

[mayhugh1]

The final parts needed for the distributors are their caps. In the past, I've machined my distributor caps from clear polycarbonate due to its good machinability and resistance to cracking. A transparent distributor allows me to easily spot plug firing problems as crossfires are visible through the cap. For this engine, I decided to make a pair of transparent caps for troubleshooting purposes, but I also wanted something long term without the 'Visible V8' look. I had three short drops of 2-1/2" diameter blue Delrin rod in my scrap box left over from a project I was involved with several years ago. Each had enough volume to be machined into the cap I've designed for this engine. My plan was to also make three Delrin distributors and end up with a spare for that blown V8 I hope to make someday.
Construction started with mounting the rod drop in a flat-back 4-jaw chuck on my lathe. The workpiece was turned to the distributor's final o.d. and a cosmetic fillet was added just above a larger diameter pedestal what will eventually become the mounting flange. The chuck with the blank still in it was then moved to the mill where the more interesting machining took place. I filleted all the edges of the design in order to make the distributor look like a full-size extruded part. For once this didn't result in my typical outlandish machining times because Delrin is easy to machine, and beautiful surface finishes can be obtained even with aggressive feeds and d.o.c.'s. Flat-bottomed holes were plunged into each tower, using a cylindrical end mill, for bronze inserts that will be later pressed in and used as sockets for the plug wires. The total machining time for the top of each cap was about an hour.
The bulk of the excess spigot material was then parted off in the lathe, and the semi-finished cap was re-chucked with its bottom side up in a vertically mounted chuck on the mill table. Here, the remainder of the excess stock on the bottom of the cap, except for the last .005", was removed. An interpolated pocket which will become a clearance recess for the rotor's center tower spring contact was then milled. The dimensions of this recess aren't critical and don't even need to be perfectly circular, but cutting it at this time will provide a nice roughed-in i.d. for the critical boring that will be done later on the lathe. The real reason for semi-finishing the base before boring the interior of the distributor on the lathe is that the tower bronze inserts must be pressed in before the interior is bored, and a good flat base will help to get them in straight.
The inserts that I've made in the past were overly complex and machined with a shoulder but I decided to simplify these since the shoulder isn't really needed. The new insert is just a .620" length of 3/16" phosphor bronze. The under-size 3/16" end mill used to plunge the holes in the high voltage towers results in an interference fit with the bronze insert of just over .001". This is sufficient to prevent the insert from spinning while it is being cut during the internal boring operation. A hole was drilled in the top end of each contact before it was inserted into the distributor. This hole was sized to receive an electrical contact that will be soldered onto the plug wires during final assembly. For my last engines I came up with a process using several pieces of shrink tubing that will be built up to cover and strain relieve the contact on the wire and, in fact, end up looking like a pretty decent plug boot. I'll detail those construction steps later when I build up the plug wires.
The cap was then moved to a set-true chuck in the lathe and the i.d. was indicated. The .005" excess was skimmed from the bottom to re-true the part. The first boring operation opened up the i.d. for a snug fit to the distributor cap locating ring on the distributor housing. The final operation was boring the main i.d. for the rotor disk. The distributor housing with its rotor was trial fitted to the cap during both boring operations, and the last several thousandths were bored .001" at a time until the rotor spun freely with no rubbing. (It was during this last boring operation on my third Delrin part that I screwed up and trashed my spare cap.) At this point the housings became married to their particular fitted caps and the caps were temporarily marked accordingly.
The plug wires to the front row of cylinders will come from the engine's left-side distributor, and the wires to the rear row of cylinders will come from the engine's right-side distributor. The last machining operation on the caps was engraving the plug wire numbers onto the tops of the caps near their associated high voltage towers. When the crankshaft turns, the rotors spin in opposite directions and their directions were also engraved on the caps around the center towers.
The center tower inserts which rub against the rotor's spring contact were finally pressed in. The bottoms of these inserts have hemispherical ends for a better interface to the rotor's spring contact. - Terry

 

[cfellows]

You have to be proud of those, Terry. They really turned out nice!

Chuck