Joshua's RV-9A Project

Back out today, I decided I’d run the engine hard for awhile to take care of any break in, though it’s probably finished by now.  Because of all the troubleshooting stuff, there’s been a variety of low-power operation to test various things.  So, off to explore the perimeter of the phase I box.  About an hour’s worth of flying time took me around the edge, from Red Wing down to Winona, up to Menominee, Boyceville, and back home.  Overflights only this time, no landings…keeping the engine at high power.

When she’s up and going, it’s quite nice.  Still stumbling on final & taxi, but nothing noted otherwise.  I burned some fuel out of the left tank to see how heavy the left wing is; the amount of fuel it takes to correct the imbalance is about 9 gal.  A thread on VAF with the math suggested that ~8 gal would be the predicted amount to compensate for lack of passenger weight in a side-by-side RV.  Maybe it’s not that far out after all.  It does require a constant pressure on the stick, though, so I need to sort it out so I can get the roll trim installed.

After refueling, a second quick flight around Lake Pepin before sunset.  The interior lighting works nicely with the dying sun, though obviously I haven’t tried it in actual dark yet.  I need to adjust the wig-wag speed to be slower, as I want the lights alternating on downwind; they don’t need to go steady until on final.  It’s so neat to come into the airport with the field lights on, and to see people’s lights turning on as the sun goes down.

Off to the hangar this afternoon, armed with parts: a new nozzle line fitting to replace the leaky scratched one, and a 90° -4AN fitting to enable the installation of a shorter servo-to-spider fuel line.  A few hours later, everything was installed and secured.  A test run showed no leaks, so I put the cowling on and went to fly, a quick local flight to get things up to temp and test for changes.  Land, inspect, refuel, and off to fly some more…headed up to Menomonie and back, with a detour to follow a river valley and then up Lake Pepin.

On the up side: the nozzle line fuel leak is, as expected, gone.  Also, the main seal that was weeping a bit of oil seems to have sealed up; there’s no more oil being slung off the ring gear.  And, the EGT drop on #2 seems to have been taken care of by the test & rebuild of the fuel spider.  All 4 cylinders stay together now when pulled to idle.

On the down side, the thing still stumbles as much as it ever did; both subjectively and as shown in the data logs.  I pulled out 40 inches of -4 hose between the servo and spider, and replaced it with 29 inches, which is really about as short as it can be.  The line is routed as far as possible away from exhaust pipes.  I find the hottest place in the line to be the area where it rises up to the purge valve, between the #1 and #3 cylinders.  I was really hoping that the line change would produce some result, even if only a marginal improvement…that’d tell me that we’re on the right track.  The vapor explanation makes sense, given the symptoms and the data logs, but the FI folks seem to think we’re at the end of the road with this, that the hose was going to improve it, and there’s not much more to be done.

A thread popped up on VAF last week that dealt with this same problem, and it’s not unheard of.  Some people seem to accept it and move on.  However, feeling the engine stumble during descent to land and final approach is not confidence-inspiring, and I don’t know that it would cause happy feelings in passengers, either.  Not good, as far as I’m concerned.  Surely there is something that can be done…plenty of folks are running FI without these problems, and it’s not that hot outside right now.

So.  No progress on the biggest issue.  Perfect…I’ll start calling again on Monday and see what “they” want me to do next.  It’ll be two weekends out before I can get back to the problem, since next weekend is stuffed full at work.  (This does make it hard to work with…by the time I’ve been able to test something out on a weekend and report back, the technician has moved on to other problems and I have to reacquaint them with the entire story.)

I also did a ground run with the MP lines disconnected from the Pmags, to get ahead of the question train on ignition problems.  No change.  And, while it was stumbling during the ground run, I alternated ignitions on/off to see if one was different than the other, or causing the problem.  It ran equally lousy on either ignition.  Anything 1200 RPM or above still runs fine.

Headed into the Sunday-Monday “weekend” with high hopes…  one afternoon later, we’re no farther ahead than we were a week ago.

Reinstalled the fuel spider and did a ground run to check operations, then replaced the cowling and went for a test hop.  Something must have been flushed out during all the testing and inspecting, because the EGTs on all four cylinders tracked pretty closely together during the flight.  The hot, low-RPM stumble, however, is just as bad as ever.

Airflow Performance is of the opinion that the stumble is being caused by fuel vaporizing in the lines between the fuel servo and the injectors, most likely in the line from the servo to the purge valve.  They recommended shortening that line as much as possible, and removing the FF transducer for testing purposes.  So, I dug out the original hose that was sent with the engine, and set to work.  It turns out that hose is 2 or  inches too short to run from the servo to the purge valve…it measures 30″ end-to-end, and 32 or 33 would work nicely.

So, having failed on the “short as possible” avenue, I reinstalled the lines I had (which total 40″, split partway through to accommodate the FF transducer), but with the FF removed and a steel union, wrapped with ceramic felt & silicone tape, in its place.  A second test hop revealed that the stumble is still there in this configuration, so the removal of the FF did not help.  The stumble may have been reduced in severity — or it may have just been my wishful perception.  In either case, it is still present.  I suppose the final test would be to procure a shorter hose of the appropriate length to install and test.  The hottest-to-touch part of the hose is the vertical run where it heads up between the cylinders; there’s nothing else to be done there, though.

And, to top it all off, I found that I now have a fuel leak at the spider.  It appears that the nipple line fitting out of the spider for the #3 cyl somehow got a scratch across the seating face.  So I’ll have to order a new one of those, too.

Essentially, that means we’re back on the ground for another week and a half — at least — while I wait for more parts to arrive, family obligations requiring out-of-town travel for a weekend to pass, and evening work commitments to pause.

I’d expected to be at least half done with the test flight hours by now, given that I had a couple weekends with nothing else to do; as it is, I haven’t been out of sight of the airport, and the only flying I’ve done has been to test changes and modifications which have proven fruitless.  There were some disheartening, throw-it-all-away moments during the build; the feeling I have after all this testing (that has only created more problems, solving none) is that, times three.