Paul Lee - SQ2000 Project, Pierre SD USA

Click on images for larger view.
Franklin 6A350 Engine (the first engine) & cowling.

Ah.. yes.. the engine. Here is a pile of parts, the way I got it from the seller.

The parts are said to be from a Franklin 6A-350-C2, 1430hrs, engine block with a rebuilt yellow tagged crank from Franklin Aircraft Engines, Inc in Fort Collins Co. for $5,500. I paid another $2,000 for new bearings/gaskets. I have a quote for about $1800 to rebuild the cylinders to new condition. The camshaft was to be reconditioned by Aircraft Specialties Services of Tulsa OK, but turned out to be a reject - see choices page for more comments. The oil pan was also rather strange. It fit but had a big open hole toward the front which I had to weld in with an aluminum disc. I suspect it was an older helicopter, or whatever, Franklin oil pan for shafts in front.

The basic block cost is about $9,900. A brand new engine is about $16,400. So the savings is about $6K. The accessories - starter, carb, etc. are all extra (the total with my choice of accessories by 12/2003 came to $16,700 - not bad for a rebuilt engine.)

With all the different parts and experimental accessories, this definitely is an "experimental" engine and not "certifiable". In my mind there is nothing wrong with it, just different.

Why a 6 cylinder engine? I personally believe giant four bangers VW engines produce too much vibration resulting in early failure of engine components and other fuselage parts. Vibration is the worst enemy of mechanical parts. There is no sense in aiding that with a giant VW 4 cyl. A 6 cylinder is naturally smooth. I am using an in-flight adjustable Ivo prop. Ivo Prop Corp will not sell props for Lyc. IO-360 because of vibration problems. The continental IO-360 is an alternative 6 but more expensive than the Franklin.

Postbuild Notes:

Engine make Note:
While I refer to the engine as Franklin 6A-350 and the basic block is a Franklin, it is an experimental engine that I assembled from a pile of parts and pieces, from different sources. It uses two Facet electric fuel pumps, dual electronic ignition, automotive alternator, velocity exhaust pipes, TBI fuel injection (2010), and 8.5:1 racing pistons (2010). The tag on the engine is "Paul Lee's experimental engine, serial 0007"
Engine choice and altitude ceiling considerations:
If you plan to fly in mountains you might want to consider a higher compression engine or turbocharging. I tested the Franklin (with 10.5:1 pistons) to 18,000 VFR legal limit - it still was doing 200ft/min. A friend of mine with a Lycoming O-320 in his experimental could not go beyond 13,300. Lycomings typically have a low compression ratio - 7.5.

The engine mount is somewhat patterned after Velocity Franklin engine mount that I have seen photos of with engine bed rectangular tubing.

The engine mount is shown with the welds completed with additional gusset reinforcements that I felt were needed. The small triangle gussets were cut from a 3/16 stainless steel plate piece that I got from the local welding shop. They do not stock 4130 steel that I ordered for the tubing and the rectangle. My stress calculations show that the mount should bear about 10,000lbs (the mount, not the firewall points) before breaking. Should be sufficient for a 350lb engine and a few g's. Hope that my welds are decent. I used a gas torch afterwards to smoothen out the mig weld beads and heated the weld joint areas uniformly to red hot to dissipate any built up internal welding stresses.

07/11/2010 Postbuild Note: The engine mount seems to have stood up well over six years now.

Purchased a 40A Nipon Denso alternator with internal regulator, from Niagara Air Parts for $225. At about 6 lbs it is considerably lighter than standard stock and about half price of similar unit sold at Aircraft Spruce. Had to weld an adapter bracket to the Franklin alternator mount. The brackets that come with it are only for Lycoming engines. I shortened the Franklin alternator mount bushing shaft (from front) by about 1/2" since it was too close to the engine mount arm which goes to the firewall. Vibration or the thrust of the propeller might push the engine close and weaken it.

2004 Postbuild note: The little alternator simply did not stand up when using more than about 15A. It quit several times on me - once I made a rush landing on batteries only. I purchased another same sized (new) alternator rated at 45A from eBay. That one just fried on me - the windings all turned black. My guess is that these small altenators are rated for temporary peak power and not for continuous duty - the heat under the cowling probably doesn't help any. I got a couple more alternators a 70A heavy duty unit off a Honda which was about 10 lbs, and another 60A unit from Van's Aircraft supply at about 7 lbs. I first mounted the 60A unit since it had similar mountings to what I already had. The 70A unit I still have but it has somewhat different mountings than my arrangement. So far I haven't had problems with the Van's 60A unit at any power demand.

I use a dual battery, single alternator system with a symmetric charging circuit - click here for details

I have since modified my alternator circuit. I placed normal alternator connections directly to the front larger battery and used one of the battery isolator diodes to charge the smaller rear battery from the large battery circuit. revised charging circuit - click here for details

The Van's 60A alternator came with a warning not to use overvoltage circuits, since it can damage the regulator. But my engine monitor has an overvoltage alarm and I can simply switch off the alternator in that case.

Got most of the engine parts in and now putting together the two crankase halves. A ring wrench part, socket wrench and a little welding help to make the necessary tools for the assembly - not quite as good looking as purchased tools but works ok and eliminates waiting for a special tool order to arrive.
And the engine is growing. Did not have an engine work stand, so might as well use the mount and prevent having to hoist it in position. The exhaust "tail" pipe was a forward extension in the exhaust and then curved down and back slightly. But it did not fit my setup and it is shown cuttoff from the exhaust. The stainless steel exhaust was purchased from Velocity for the Franklin engine - $750. I purchased two additional SS U tubes ($110) and cut them to shape and had them welded to the exhaust at a local welding shop ($110) - see result below. That brought the exhaust cost close to near $1000. Later, after reading EAA articles, I found that cheaper mild steel exhaust can last relatively long. Oh... well... hope this one lasts the lifetime and its a bit lighter too.
And growing... The tail pipes will be cut to size after the cowling is built.
01/25/2005 Postbuild note: This is of interest to Franklin engine owners. The standard Franklin 6A-350 alternator drive adapter gear rubs against the adapter collar depending on engine oil splash to keep the contact surfaces lubricated. I have found this to be inadequate showing premature wear with the gear/collar contact wearing away faster and the gear working itself more and more toward the outside. To solve that I have removed the collar with a lathe and had a larger collar welded in place and machined a hole for a larger ball bearing to hold the gear from moving outward (outward = away from the starter gear). The ball bearing replaces the inner roller bearing, with the outer roller bearing staying where it is. This gives me a spare roller bearing - which is hard to get. In extreme case even the outer bearing could be replaced with a ball bearing in a machined new hole.

The ball bearing was purchased from Baumgardner and Sons, Cleveland, OH 44110, 216-531-2800 - part number 6204LL 7/8 (0.875 ID) (mfg part ref: 5204-11-7/8)

07/14/2005 Postbuild note: The previous mod lasted five years - 180 hours. On a recent inspection I found the outer roller bearing failed and I re-designed the unit with two ball bearings. I made the ball bearings removable rather than pressed in like previous time.

Often stock parts orientation and connection directions - like this oil separator - are not suitable and need changes.

Postbuild Note: The oil separator did not seem to do much good. I removed it. I eventually tried a more expensive one - RMJ-Aero Air-Oil separator - which I hooked up to drain into a valve cover. The combination actually increased my oil consumption. So now (2006+) I do not use an oil separator. Not sure whether its my particular setup or just that oil separators simply are not what they claim to be.

Got a carb airbox for the Franklin engine from Velocity. The box is simple enough but does not provide air filtration. Decided to use the NACA scoop under the fusel. for carb filter and intake since I am not using it for anything else. The problems of not having an air filter is carburetor contamination especially on runways. Up in the air, cleaner air is expected. I intend to make a somewhat complex flap combo that will use the air filter when taking off/landing and on ground; and open to straight ram air when up in the air. At this point I plan to use a servo connected to landing gear signal switches to close the ram air flap (use air filter) when landing gear down and open the flap with landing gear up to get ram air into carb. The picture shows some 3" PVC sewer elbows used for a carb curved intake mold. The intake bend will be epoxied to the carb flange made from alum. flat - the 3.25 hole cut using a hole saw.

Update: The aluminum flange did not work out and I made one from several layers of glass and bonded it to the elbow.

The LightSpeed Engineering (LSE) dual ignition system direct crank sensor circuit plate does not fit my configuration - too large and would interfere with the IVO prop electric control. The factory plate was originally designed to fit the Lycoming engine and not the Franklin setup inspite of their assurance that it can fit the Franklin. If I knew this beforehand I might have gone with the ElectroAir ignition system which uses the magneto gearing hole to trigger the spark. However LSE did send me the Hall effect sensor circuit elements to fabricate my own. Here is my attempt at it. I made two smaller plates out of fiberglass: the front one (3 layers of 900 E glass) on which the sensors and wires go and the back one (5 layers of 900 E glass) for more support. The two will then sandwich a layer of flox which should be pretty strong.

In third photo, the cured plate is shown trimmed. Loose (OD but exact 1/4 ID) steel bushings were put into the mounting holes floxed in, allowing exact axial and radial adjustment for the sensor plate. Radial distance and axial distance to the hall effect sensors must be equal all around. The thicker plate (right) side (impossible to make perfect) is clamped down with a washer over the flox while the axial distance to the left side is adjusted (using wood slat) until the axial distance to hall effect sensors (little black dots - hard to see) is equal all around.
If you do this, don't forget to cover the engine seal retaining plates and the washer with mold release (I use clear plastic packing tape).

The fourth photo shows the final plate assembly with the magnet holder. It is about 6 inch diameter, 3/8" thick, made from 3 + 5 layers of E-glass sandwiching a layer of flox encasing the connecting wires ("circuit"). The aluminum magnet plate is welded to a 180 degree of round tubing which is clamped to the crankshaft. The sliding freedom and "fingers" allow for more precise adjustment of magnet positions. The two flat head machine screws are inside steel machine bushing floxed/cured more precisely into place. The extra tabs on the plate in second photo were intended for additional rigidity support but turned out not to be needed and were cutoff. (The green original factory plate is shown on the side for size comparison.)

Unless LSE comes up with a smaller sensor circuit plate and magnet holder for the Franklin to not interfere with other propeller controls, I would not recommend the LSE EI. This was tremendous amount of work. But I already paid "non-refundable" $1880 for the unit through Aircraft Spruce.

I have indicated my setup to LSE manager who was concerned about the magnet holder and the clamp holding it. The cheapest SS clamp I have can hold over 500 lbs of force. At 2800 rpm the centripetal force on the 1.4 oz holder is only about 30 lbs. But as a safety measure I used HI-TORQUE - Heavy Duty Breeze clamps having about 2000 lbs of clamping force capacity. That should be sufficient safety factor.

08/18/05 Postbuild note: At about 75 hours the engine developed some roughness, especially at lower RPM being harder to start. The double ignitions were uneven, with more roughness when the front battery ignition switch was turned off. The plugs were almost new. I contacted Klaus of LSE who suggested to look into plug gapping and mixture. However I traced it down to a shorted out coax ignition cable. The RG-58 cable that came with the original kit did not stand heat too well. I noticed the newer LSE ignitions come with teflon RG-400 type instead of RG-58. I replaced the cables with RG-400.
That was an obvious defect in the original kit. It would have been good to at least notify LSE kit users about the problem or even post something on LSE website. But perhaps it avoids an expectation of free part replacement.
03/09/11 Postbuild note: LSE is offering a newer LSE mini crank sensor to reduce the size for Franklin and other engines. Hoever the new setup is still huge bulky compared to my own custom build and would interfere with prop controls like IVO.
The carb air duct system is growing but is taking a long time. With no pre-mold every piece/shape has to be glassed/fabricated individually. Shown upside down it will be installed in the bottom NACA scoop - since I have nothing else to use the scoop for. The duct system shows the hot air control flap and will have a flap to bypass the filter at altitudes away from ground level dust.

11/27/03 second late photo shows the installed duct system from the bottom. The front part was cutout since it was not needed. A properly shaped cover encloses the system so that the air is directed from outside to inside of air filter to inside of the carb duct system. The filter is FRAM AC148 for a 300ci 6cyl Ford truck engine. The Franklin is 350ci. But keep in mind that an automotive engine revs up to about 5000 rpm, while the top on the Franklin is about 2800rpm. So comparatively the filter should have about ( (5000 x 300) / (2800 x 350) ) x 100% = 150% air design volume flow requirement for the Franklin. In view of that I did not bother with the complex flap system to bypass filter at altitudes. The filter in the photo is held in place by a short glassed ring surrounding the filter "top". The third and fourth photo show side view of setup and the inside of the bottom cover.

The original engine baffling cover purchased from Velocity was too large. I decided to split it into two halves as shown. This method is already used in some Jabiru engine systems.
There are two heat muffs, one for carb heat (LS) and one for cabin heat (RS) seen on top and using the overhead console tunnel. After trying standard suppliers route I eventually had to weld my own using a few pieces of SS straight and elbow tubing from Woolf Aircraft Supply for about $84. Before that I tried a muff from Velocity supposedly made specifically for the Franklin engine but turned out to be the aluminum Cesna type muff that Velocity got from Aircraft Spruce and simply marked up $20. I got a local welder to weld an alumnum elbow (I am still not good at welding thin aluminum) but welded it at wrong angle - the whole thing was near $100 and did not work. The stainless steel muffs I welded toghether worked out fairly good with the added benefit that it was attached to the SS exhaust with just a few spot welds - no clamps needed.
AND.... the engine is running. Kind of rough, especially at idle. Need to adjust carb. The Odyssey PC680, is an amazing little battery, only 3x7x5H. Has no problem cranking over and starting this high compression (10.5) engine.
The best price I found for Odessey batteries is at

I found some oil leaks after running the engine. The main one, I fixed later, was around the oil temperature sensor which I mounted in a too small pipe thread shaft. I fixed that by putting some J-B Weld on the threads. The other were static leaks near front of engine oil pan and some on bottom where the plate caps were. But these added up to about 1 drop a day so, while annoying, I did not think it was worth taking the engine out to fix. I put a small hole near bottom of the carb cover back of the cowling so that the drops will drip through one spot.

The factory cowling does not fit the Franklin engine - a little higher than Lycoming. Starting to make cowling. I hate fussing with foam molds (mess and all) and have used a metal sheet for making glare shield. Here an alumninum sheet is curved to the bottom shape of the cowling and used for bottom piece mold.
And here is another idea for making the upper cowling part. Make a thin single 9oz S2 layer and let it cure. After curing it is very flexible but strong enough to stay in shape that you support it in and will hold the next layer of glass which will stiffen into shape. No mess with foam forms. (The bottom part has already set and removed).
With this new thin glass sheet form method, shown are the rudimentaries of engine cooling ducts. The half moon scoops were made using plastic salad bowls as forms. This method takes one piece at a time - not sure if it takes longer than foam carving but sure less mess.
Finally getting the air ram ducts to what they should be, now working on the rest of cowling. The rear top side skirting shown curing over a mold made from a piece of tin and a curved dryer went quarter round pipe.
Finally the engine is surrounded laterally and now I am making cowling wing extension. For a form, a simple tin piece easily curved to shape and hot glued in position (underneath a series of small wood blocks hot glued to hold the sheet up). The tin is covered with plastic for release agent.
The Franklin engine has external oil connections on the right side so the empty right cowling wing extension is a good place for the oil filter and cooler. The air enters from a wing bottom separate scoop into backward angled cooler and comes out on top through a exhaust port ducting air near prop. Keeping the oil cooler circuit away from the engine helps keep the engine compartment temperature down.

06/09/04 Post build update:
1. The 4211 Positech oil cooler recommended for Lycoming O-320/360 was not sufficient size for the hotter running Franklin engine. It requires a much bigger oil cooler - 4219 or equivalent.
08/24/04 Turns out that the second considerably larger 13 row 8000215 AeroClassic oil cooler I installed (typically suitable for Lycoming I0-540) was just marginal for the US built Franklin 220. They require even larger coolers and I had just got a larger Stewart Warner 10614R unit I will have to install again.
2. The Peterson brand oil filter adapters with 10 AN fittings ( detail photo , part numbers: 09-0313 right-to-left or 09-0311 for left-to-right) are available from Peterson Fluid Systems for about $65. These "FORD" adapters are machined from solid aluminum billet and fit the common aircraft oil filters like CH48108/109. They looks better and are cheaper than a similar units from Aircraft Spruce.

And now the left side cowling wing extensions. Much faster since I know how to do it and less complications without the oil filter/cooler assembly. The right side shows the top exit port for the oil cooler. The bottom entry being a small NACA scoop. The factory prototype has a large under-wing scoop for the oil cooler. I am counting that the exit port being closer to the prop will generate a lower pressure.

I'll have to do something to taper the end of the cowling a little more near the prop. It looks a little too boxy.

The engine is finally closed in. But a lot of details remain. Its not looking so boxy anymore. Its surprising how just plain epoxied glass draped over the edge gap can be curved by hand - especially as it gets stiffer with time. The surface isn't perfect like a mold, but the imperfections can be filled in with glass bubble epoxy and filed/sanded to desired shape.
I am using glass bubble / epoxy mix to fill in the surface irregularities on cowling halves. The mix is light and after its sanded it will be covered with a final layer of 9 oz of 7781 E-Glass which is very dense weave ("satin weave") and smooth. The small layer of glass bubble mix helps to stiffen the cowling since it acts like a sandwich - similar to factory cowling made up of foam & glass.

Just a diversion here regarding the 7781 E-Glass. I got some 7781 E-Glass from to try it out. It is somewhat comparable in strength to the KLS popular 9 oz S2 glass. But the stuff is much more dense and about half the thickness of the S2. What that means is it absorbs much less epoxy for approximately same strength - the strength is more in the fiber than epoxy. This is important for a beginner builder who tends to put more epoxy than necessary and add to weight. I wish I found out sooner about such cloths and could have used them more to lighten the construction.
04/02/05 Note: 7781 E-Glass is now available from both Wicks and Spruce in wider widths and cheaper than
I have done a simple experiment. Took a piece of the cloth and weighed it before and after applying epoxy (and peel ply one side). The result was 40% epoxy and 60% cloth. Thats pretty good ratio for simple hand application. Beginner builders often get 70%/30% ratios but this cloth type will dramatically improve (lower) the epoxy ratio for lighter resultant layups.

Here is what the IVO prop looked like at a fly-in.
April 2010 update. I sold the IVO prop in 2009 trying to experiment eliminating a persistent vibration I had. I purchased a two blade Felix prop from Fred Felix and also ordered a more expensive three blade Catto. Plan to use the Catto for normal use and the Felix as a spare. The Felix prop has a good climb with 2500 static RPM but maxes out at about 155 kts. It is a wood prop and can easily be dinged with grit so I had it covered in carbon fibre by Lonnie Prince of Prince Aircraft. Lonnie does a variety of prop services beside manufacturing his "P-TIP Propeller" design.

The Catto prop is a strong fiberglass/carbon-fibre prop which initially gave me only about 2100 static RPM. I returned the prop and Craig removed about 1.5" from the diameter to give 2400 static RPM.

This year I gave close attention to the vibration problem and tried enlarging exhaust pipe cowling outlets. I also put more effort into propeller balancing. In 2009 I splurged $1500 and purchased a DynaVibe electronic balancer. A few years before I paid $350 for the IVO prop dynamic balancing and they did not improve anything. So I figured it would just take a few balancing jobs to justify the purchase and I can probably resell the balancer someday. Below you can see the mounted Catto prop, Felix prop with the carbon fiber cover and the result of recent balancing effort on the Catto. I managed to get 0.00 ips adjustment with the DynaVibe on the Catto.

For some reason IVO props are not as easily balanced as fixed props.

In 2009 I modified the cowling to open end. It was also part of an effort to locate the vibration - possibly some "aerodynamic effect". Turns out it did not make a difference in vibrations but made a signifficant difference in engine cooling. The cylinders are running about 20C cooler now.

Return to main page