Fuel Injection Conversion - Part 1

 

A problem long in the tooth 

If you have ever bought a plane, especially one whose type you have never flown, you’ll know it takes a little while before getting comfortable with it, but eventually you start settling in and noticing patterns of things you like, and inevitably, things you don’t. Among the latter ones, some issues will show up so often that they’ll end up plaguing your ownership experience, all the while tugging at your pocketbook strings.

 

Getting home from work to "new airplane day"

Your subconscious will begin to estimate the amount of work it would take to correct the problem versus the improved ownership experience. Consciously, you will decide that while you might be able to live with some, you will just have to live with others. Occasionally though, some issues will smack you right in the face in ways you won’t be able to ignore. These are usually safety of flight items.

It is not unheard of for pilots to put up with problems for some time trying to figure out how bad they really are, and what to do about them, especially the ones that are hard to reproduce, the ones you hope to wish away. However, if one waits long enough, safety of flight items will scare the crap out of you one day if you are lucky, or kill you if are not. This puts a real premium on quickly diagnosing problems, and hopefully the datapoints one collects along the way trend in one unmistakable direction toward some actionable intelligence.

This is exactly what happened with my Ellison Throttle Body Fuel Injection (aka pressure carburetor).

The Ellison TBI

A simple and very effective device (most of the times)

 

From my first day of ownership, the inability to shut down the engine using the mixture control bothered me (origin datapoint). My mechanic at the time told me it was normal, and even though his rationalizations didn’t make any sense to me, I learned to shut down the engine by closing the mixture first, and then opening the throttle.

Imagine my surprise then, when a couple of years later I asked Chris how he shuts down his identical engine, and he replied he only uses the mixture control 🤔(more data).

If one pays close attention, the datapoints one collects over time might eventually paint the portrait of a very hard to diagnose problem, one that in hindsight seemed so obvious, but completely baffling at present time.

Another example of one was the black sooty condition of JT’s exhausts after long taxiing operations. JT was clearly running way too rich on the ground.

Evidence of sooty exhaust (depositing on an improperly clocked propeller)

 

Reading Mike Bush on Engines, I learned to lean very aggressively on the ground, and “lean of peak” in the air, and was able to make this problem go away. Unfortunately, previous owners hadn’t noticed this issue, hence the development of a condition known as “morning sickness” that plagued JT before I purchased it.

Morning sickness is an all too common Lycoming condition where an exhaust valve on a cold engine sticks in position due to buildup of carbon/lead deposits on the valve stem and inside the valve guide. 

 

Lead/carbon deposits on an exhaust valve (not JT)

The engine sounds like there’s a skip in the firing sequence, an occasional but recurrent misfire that lasts until the engine warms up. Sometimes morning sickness can result in engine damage right at startup if the valve sticks fully open and gets hit by the piston, or sticks fully closed resulting in the associated pushrod bending under pressure from the camshaft lobe.

Broken pushrod shroud, evidence of pushrod damage (not JT)

 

This valve did not retract fast enough and was introduced to Mr. Piston (not JT)

Once morning sickness develops, the only cure is to clean the valve stem and ream the valve guide. This operation can be done without removing the cylinder, and was performed on JT as a condition of my purchase.

Reaming the exhaust valve guide (not JT)

 

Back to the Ellison Fuel Injection (TBI), the design has quite a few benefits over a carburetor. It works in any attitude, is resistant to icing, and it is very compact and extremely simple. Because of this the only adjustment control on it is an idle mixture screw.

Minimal amount of components on the TBI

 

Now I’m not sure if all TBIs do this, but leaning the idle mixture to fix the rich taxiing problem also had the strange effect of leaning the takeoff mixture. 

 

 

Throttle closed, mixture cutoff.

Throttle open, mixture cut-off.

Throttle open, mixture full rich.

After some experimentation, it appeared that the takeoff fuel flow was effectively the sum of the idle flow and the full throttle flow. Unable to independently adjust both flows due to the lack of controls, I realized that any further adjustments needed to be done via new sets of internal jets after bench-flow testing by the manufacturer, who is unfortunately no longer in business.

 

In order to better understand its functioning, I pulled the 1985 Ellison patent from the web, and even partnered with the department of Aviation Maintenance of the University of Miramar for a few months trying to develop spare parts that worked, all to no avail.

Original drawing from the Ellison patent

 

Ultimately, tired of being the owner of a crippled aircraft, I started looking at the TBI more and more as a terminal disease, since it could no longer be fixed, and led to a slow inevitable demise. Swapping parts with another unit would also not necessarily fix anything, and might actually make things even worse. I’d say if you have one that works, use it, but when it starts developing problems… well, you see where this is going.

Good times about to end, quite suddenly.

 

So, there I was, months earlier, taking off from runway 27 in KVTA after visiting Terry Lamp and Mike Toomey, and as I made a left turn at 500’ AGL the engine quit!!!

F@#k me!!!😬

I was right over town when JT started doing the opposite of what it needed to do to keep me alive 💀☠️.

Of all the things I could have been doing while trying to get over my disbelief, I remember slowly pulling back on the throttle a bit, not really sure why, and feeling full power kicking right back in. I climbed, changed my diaper, played with the throttle some more while remaining within gliding range of the airport, but could not reproduce. Figuring that whatever had momentarily “clogged” the Ellison had gone through, I continued home with Chris on my wing.

All is well, continuing home.

 

A few days later I tore into the Ellison, replaced the diaphragm, cleaned the crap out of it, made sure all filters were spotless, ran it, then test flew it, and… nothing.

A bit dirty but nothing obviously wrong with it

After a good cleanup

"Move along, nothing to see here."

Original diaphragm

"New diaphragm reporting for duty!"

 

All was well again, and it would remain so for a few more months, until this same issue showed up again in an even more dramatic fashion at 200’ AGL 💩💩!! This time I immediately pulled the throttle back slightly, and the problem disappeared right away as before.

Time for another diaper change…

They were starting to stack up

 

To make a long story short, after many months of observation and testing, I realized that this problem would only occur below certain fuel flow values. If takeoff flow happened to be above 13.6 gph all was well, but if the value was lower things could get wild.

Part throttle fuel consumption Lycoming IO-320D

 

Obviously the TBI had an issue, but why pulling the power back solved it, I didn't understand. 

 

My best guess was that with airspeed, rpm, and power increasing down the runway, the engine sucked in more and more air, but the TBI didn’t keep up causing the mixture to become too lean. Closing the throttle slightly reduced the amount of available air making the fuel/air mixture richer once again.

More datapoints accumulating.

In hindsight the reason for the super-rich idle mixture setting (4.5 gph at 1000 rpm vs a more normal 1.7 after aggressively leaning on the ground) started to make more sense. The idle mixture had been bumped up to raise overall fuel flow and enable proper takeoff performance.

The only way to raise fuel flow on takeoff was to create a very high idle mixture fuel flow to begin with, and the price to pay for it was morning sickness, and the inability to shut the engine down using mixture control alone.

My operational fix had been to lean aggressively on the ground in order to prevent morning sickness while still enjoying proper fuel flow for takeoff. Unfortunately, this trick was no longer cutting it.

I was collecting all kinds of good data now in support of this theory, but eventually I realized I was spending more time looking at the fuel flow during the takeoff roll than at the airspeed indicator, and I had to admit to myself that I had a problem.

It got to the point that I no longer felt comfortable taking passengers, and eventually the frequency of these occurrences became too high to ignore. JT was talking… no, more like screaming at me, and one day she would end up failing if I didn’t pay her the attention she deserved right away.

Visiting fellow builder extraordinaire Wade Parton

 

One day, while leaving KMRH after visiting Wade Parton, JT’s engine quit again just as I was about to rotate with hardly enough runway left to stop. Again I pulled back on the throttle, and JT leaped into the air like the Space Shuttle, but that was IT for me, I had had enough! After landing, JT would be grounded for maintenance regardless how long it took to fix her, and might never even fly again unless my imagination, fabrication skills, and… pocketbook, proved worthy of the tasks ahead.

As Chris and I sadly stripped JT to the bare frame, I felt uneasy and unsure about what was in store for her. “I’d do her right” I told myself, while still unsure about what exactly I was intending to do.

Reaching the point of no return

 

"Fuel injection or bust!"

So… next time I will share with you the general plan of attack, and we will explore the intricacies behind modifying JT’s fuel system. 

 

This project quickly became a 10 hour day seven days a week struggle that lasted for 18 months. Luckily for me, the timing couldn't have been any better because, due to the ensuing pandemic, my airline was looking for volunteers to stay home at a reduced salary, so I took the offer and stayed home working on JT until she flew again. 

 

At the time I had no idea how much work this would entail, and the stress of making life altering decisions for JT was so high that I wanted to quit on a daily basis. However, with my back against the wall and no other real options, I just had to figure it out, or else. Force of desperation I suppose.

Now that JT is happily flying again, I will show you what worked, why I did it, how I did it, and what I have learned. There are many ways to “skin a cat” so to speak, and this is but one of them, perhaps not even the best one, but it’s mine and it works well, so I am happy and proud to share with you all.

Stay tuned for part #2