So, What's the Deal, Sam?

I'm starting to get e-mails from people, wondering what I'm doing. I know, it's pathetic that I'm not flying.

My big hang-up has been getting my custom oil sump resolved. As you may know, I made a low profile sump from carbon fiber and it seemed to do the trick, except that it didn't dissipate enough heat. David Posey was kind enough to offer to build an aluminium one. I'm still fitting that one and coming up with a satisfactory dipstick.

So should that take 5 months? Actually, no.



I have also made some changes to the ignition sensor mount and I think I should have some success with it.




I am hoping to get the new sump installed this week. I need to get my ass in gear.

A Major Conundrum Solved

The big question of the day was "Why does the engine badly miss when I go over 2,900 RPM"?

It would be fine during takeoff, climb, cruise, etc., but the engine would badly miss when I went over 2,900. Was it the electronic ignition? A bad coil? Problems with fuel delivery? Since these systems are interwoven I wasn't sure where to start.

I thought I had already conquered the case of the vibrating sensor mounting bracket but in the back of my mind, the miss felt familiar. Since I had already beefed up the assembly, I didn't know what I could to do to make it better and I didn't want to spend a lot of effort there if I didn't know that was the problem.

I decided to shoot Lynn French an e-mail and ask for a gigantic favor. If I flew to Broken Bow Nebraska, would he consent to swapping parts - again? Fortunately, Lynn is very generous and he went along with the plan.

On the way to Broken Bow I stopped in Topeka to see my daughter and grand-kids, so I did get some great non-aviation time, too.

While I was en-route, I tried going to WOT a couple of times and each time the miss was back.

I arrived at the Broken Bow airport about 10:00, Lynn arranged a nice hangar to work in and parts started flying off the planes. We set them up nose-to nose and installed Lynn's bracket bracket and trigger wheel. Rob Martinson brought doughnuts from Colorado in his newly rebuilt Long-Eze, and took a look at our electronic ignition set-ups.

Here is Lynn's current set up:


Fortunately, it was a simple swap. I didn't have to change the timing or anything, and I used my sensors. We just changed the trigger wheel and mounting bracket. After a quick test on the ground, everything seemed well. I told Lynn that I would take it up and see how it was, that if it performed well I would just fly on home. (just kidding).

Here are Lynn's parts mounted in my buggy:

So, I went up for the test hop and what do you know? The engine performed just fine, smooth as a baby's butt. At WOT the engine was perfectly smooth and no hint of a miss. Problem diagnosed. It was somewhere in my trigger wheel mounting bracket setup and I'm pretty sure it was a vibration issue creating electronic noise in the sensors.

Got back on the ground and we put the parts in their proper places, and left the airport just as the sun was setting.

On top of all that, Lynn was nice enough to boot one of his boys out of his room and give me a bed to sleep in. Great guy.

One screwy thing, I tried another in-flight WOT on the way home and the miss was gone. This is not completely surprising, since it had previously behaved on occasion. I think my mounting system is just on the edge of performing/not performing, and when we swapped the parts I may have set it in a more reliable position.

Here is a photo of my view of Nebraska on the way back to Topeka:

Having said all that, I still need to work out a more reliable system. I have several options and I'll mull the whole thing over, before I come up with my final direction.

Thanks again, Lynn.

More chasing

Please take a look at these two sorrowful pages of sketches of today's oscilloscope output.

The engine always starts easily and idles very smoothly. Roughness starts around 1,800 RPM. Misfiring is very pronounced, really shakes the engine, and the exhaust is momentarily black.

Page 1.

First wave is pretty much what it looks like with both the 6" and 7 1/2" wheel, at idle. With the 7 1/2" wheel there is a little more of a hump in the middle of the trigger wave. The peak-to-peak amplitude is pretty even and is about 4 volts p-p.

2nd wave. This is what it looks around 2,000 RPM. Notice that the amplitude varies greatly, from about 3 V to 7.5 V p-p. We tried adjusting the air gap from .015 to .060 and didn't see a change here. We thought maybe the variations in amplitude might be from the prop bolts and flange lugs. Just speculation.
3rd wave. Another sketch at high throttle. Here, I tried to include the trigger wave. The trigger wave, at higher RPMs, always had a greater p-p voltage than any other part of the wave, while at idle, the trigger wave was the same.


Page 2.

There always is a consistent phase shift (skipping a tooth?) at any speed. In the sketch, I am attempting to show how it would bounce back and forth. We feel this is a key issue, but have no idea why it is occurring.

We also saw that the amplitude of the wave is pretty even at low RPMs, but at higher revs the amplitude varies a lot. Click on the sketches to better read the voltage ranges.


Other things we tried and noticed:

Changed from 1 1/2" to 6" wheel. no change seen on scope.
Removed the prop bolt closest to the trigger gap. - no change
Tried a smaller diameter sensor. - a little smoother.
Degaussed the crankshaft flange. - no change.
Ran on the single magneto only. - a little smoother.
An automotive timing light flashes consistently at low RPMs, but becomes intermittent, in sync with misfiring, at higher speeds.
Installed the 6" wheel and used the 7-1/2" sensor mount bracket. This caused the sensor to kind of be extended way out from the aluminum bracket. No effect.

I sent these sketches to Tracy Crook, at RWS and he thinks that the change in amplitude is the source of the problem and what I perceived as a "phase shift" is not an issue. Given that, I will be concentrating on eliminating all vibration in the sensor mount bracket.

Sam

Fowards and Backwards

Good news, bad news, good attitude.

The good news is the engine runs! Having said that, this is all extremely uncharted territory, and if it wasn't for Lynn French, I would be completely in the dark. Lynn is the character that first installed an electronic fuel injection and ignition, in his Tri-Q-200.

I used Bosch 24 lh/hr. fuel injectors, sometimes referred to as "blue tops". These are the same size that Lynn uses and it seemed like a good starting point. The ECU is programmable and you can adjust the fuel flow to your own particular application and injectors. When we first got it started, it was extremely rich. I mean rich, to the point where there was black smoke plainly coming out of the exhaust and now there are two major soot streaks running down the belly.

The first few times I ran it, it was kind of a sensory overload; all these new dials, buttons, and the like, to learn. And, it seemed like some features worked and some didn't. It was, and still is, a lot to sort out.

There is a fair amount of programming of the controller to do. The idea is to first match the basic flow of the injectors to the engine, then to map out the mixture tables.

Adjusting the mixture is interesting. First, instead of a mixture knob that you pull out, there is a mixture dial on the Programming Control Module (PCM). You use the dial for tweaking, but you use the Program Variable, Mode and Store controls to make your adjustments permanent. (Over simplification, really).

To adjust the mixture, you really can't use the conventional EGT. We welded a bung into one of the exhaust stacks and have installed a Bosch automotive oxygen sensor. This is connected to a gauge in the cockpit and it provides a digital presentation if the mixture is rich or lean. This gauge reacts much faster than a regular EGT and is easy to interpret. The first sensor I bought didn't work very well, so I bought a better one which has it's own heater.

The O2 sensor is temporary. They are designed for unleaded fuel and after a while 100LL is said to clog them up. I only need it though, for the initial mixture mapping.

All of this is so new to me and there is so much to learn. Sheesh!

After a while we (C-Rod and I), came to the conclusion that the programming button wasn't working. Now, we may be mistaken, but we think we have it right. Anyway, after lots of diagnostic e-mails, back and forth, with the controller designer/manufacturer, I pulled the unit and am returning it for evaluation.

I will say that the engine starts easily, and seems to run fairly well, considering. There are lots of unknowns here, I have an automotive racing throttle body, homemade induction system and lots of seat-of-the-pants engineering, so I am taking everything in stride.

I guess I'll work on my taxes while I wait the outcome of the controller.