LS3 Misfire

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
1,779
Location
Republic of Texas
No...not my plane.

But another T-51 owner has the newly installed LS3 engine. With a Link G4+ Xtreme ECU. The engine runs perfect, for about 15 minutes of flight. Then after getting warmed up, it misfires. No one seems to have a clue why.

My money is on:

1. Coil pack that fails when hot, so plug doesn't fire
2. HT lead that has a break and resistance increases when hot, so plug doesn't fire
3. Faulty injector

Anyone here care to take a stab at a diagnosis?

IMG_0533.JPG
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
13,061
Location
Memphis, TN
Do these ECUs have diagnostics? Codes? You could ohm each coil pack in hope you could see one off from the rest. If not you would either have to buy 8 coil setups and 8 plugs and change all at once or make 8 flights moving one new coil pack to each position. I would change all plugs. Plugs are cheap. Same process if you do injectors.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,085
Location
Rocky Mountains
Anyone here care to take a stab at a diagnosis?
Not me. Done too much of trying to diagnose intermittent problems.
You have 2 options that work. One is to just start replacing parts that have a high probability of being the cause until the problem goes away, as TFF mentioned. This works but can get expensive and time consuming .... or you can get lucky and hit it the first try with your SWAG.

The second method, and the one I prefer, is to gather data. Hook up a lap top and digital oscilloscope. Go fly and record the data real time. Note the time the event starts and then look at the data leading up to that point. You may still not be able to pinpoint the cause if it is something like a bad plug that can't be directly monitored. But you can eliminate a LOT of variables.

The second presumes that you have done the standard ground tests of the individual components, like ohms/volts/frequency, while shaking the wiring bundles and looking for intermittent signals. One can also bake things like coil packs and injectors in an oven to simulate operating temperatures.
 

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
3,749
Location
Calgary, Alberta, Canada
Just about any modern aftermarket ECU should have data logging capability which should quickly show you what's causing the issue. Almost certainly not an injector as they never have heat related issues fitting these symptoms.
 

narfi

Well-Known Member
Joined
Feb 23, 2016
Messages
837
Location
Alaska
I've never worked on auto conversions, but on traditional engines assuming you dont have individual cylinder monitoring,

Emmediatly after shutdown from running hot and rough, hit each exhaust flange at each cylinder with a laser temperature sensor to find the one not firing.
Alternatively run the back of a bic pen across the exhaust as close to each cylinder as possible. The good cylinder exhausts will be hot enough to melt the plastic and lubricate the pens path so it slides easily against the metal, a cold cylinders exhaust will not and the pen will scrape against it instead of sliding.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
1,779
Location
Republic of Texas
Just about any modern aftermarket ECU should have data logging capability which should quickly show you what's causing the issue. Almost certainly not an injector as they never have heat related issues fitting these symptoms.
Thanks. Yes, the ECU has many channels of data available, and should make a diagnosis fairly easy.
When we get ready to fly (paint shop this week), hopefully we will not have any of these issues.
 

Toobuilder

Well-Known Member
Log Member
Joined
Jan 19, 2010
Messages
4,777
Location
Mojave, Ca
So are we discussing your buddies airplane in this thread, or yours? I can't keep track of the problems you are trying to solve. And I'm with Ross... This problem should be a 5 minute job to solve with the data logging your buddy (or you) have. If this was an OX5 in a Jenny, that would be different.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
1,779
Location
Republic of Texas
So are we discussing your buddies airplane in this thread, or yours? I can't keep track of the problems you are trying to solve. And I'm with Ross... This problem should be a 5 minute job to solve with the data logging your buddy (or you) have. If this was an OX5 in a Jenny, that would be different.
Long story. Not my plane, but an LS3 engine. I am just trying to help Jim out, and get the plane flying soon.

Thanks to all of you for your input!
 

pictsidhe

Well-Known Member
Joined
Jul 15, 2014
Messages
7,552
Location
North Carolina
It could be a zillion things. Without looking at it, hard to tell. Somebody who knows engines needs to troubleshoot it. Or throw parts at it. If you can reproduce the problem on the ground, it will be easier.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
1,779
Location
Republic of Texas
Now I am really scratching my head...the folks at Titan claim the data log is absolutely no help in diagnosing the problem.

Does this make any sense to you? Seems odd to me.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
13,061
Location
Memphis, TN
My truck will tell me which cylinder. My wife’s car too. She had one recently on the way to her moms. She had it checked at an autozone and I told her to buy the part and I will come down and change it
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
13,061
Location
Memphis, TN
It went intermittent and was ok the rest of the trip, but when she got back I changed it. The information should at least tell the bank.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,085
Location
Rocky Mountains
Now I am really scratching my head...the folks at Titan claim the data log is absolutely no help in diagnosing the problem.

Does this make any sense to you? Seems odd to me.
Yes, it can be the case that the data log provides no clues. It depends on the detail and accuracy of the data log. A modern EFI using dual wide band O2 can almost give you a cylinder by cylinder real time data stream. I'm guessing that this aviation EFI doesn't even use O2 sensors? If so then the rest of the system is likely to be just as simple and the data stream may not capture the error.

in this case you are back to the old school (pre OBD II) days of using a digital multi-meter and a breakout wire harness.

If it's a Titan installed engine package - it's kind of their problem? Without detailed knowledge of the system architecture all we can do is guess.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
1,779
Location
Republic of Texas
Yes, it can be the case that the data log provides no clues. It depends on the detail and accuracy of the data log. A modern EFI using dual wide band O2 can almost give you a cylinder by cylinder real time data stream. I'm guessing that this aviation EFI doesn't even use O2 sensors? If so then the rest of the system is likely to be just as simple and the data stream may not capture the error.

in this case you are back to the old school (pre OBD II) days of using a digital multi-meter and a breakout wire harness.

If it's a Titan installed engine package - it's kind of their problem? Without detailed knowledge of the system architecture all we can do is guess.
Correct, O2 sensors not used, because the fuel is 100LL. Initial flight tests were done with the O2 sensors (on the test engine), then the fuel ladder was programmed.

On the surface it would seem that it is Titan's problem. But there is more to the story, and it is a long story. Not going into details here.
 

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
3,749
Location
Calgary, Alberta, Canada
Yes, it can be the case that the data log provides no clues. It depends on the detail and accuracy of the data log. A modern EFI using dual wide band O2 can almost give you a cylinder by cylinder real time data stream. I'm guessing that this aviation EFI doesn't even use O2 sensors? If so then the rest of the system is likely to be just as simple and the data stream may not capture the error.

in this case you are back to the old school (pre OBD II) days of using a digital multi-meter and a breakout wire harness.

If it's a Titan installed engine package - it's kind of their problem? Without detailed knowledge of the system architecture all we can do is guess.
Link is a high end automotive ECU and it has data logging capability.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,085
Location
Rocky Mountains
But there is more to the story, and it is a long story. Not going into details here.
Understood. Been there done that.

Get as much documentation as you can dig up on the FI system. Wiring diagrams and sensor specific specifications will help you reverse engineer the system. Unless you go down the parts swapping path you have to know the system details if you are to have any hope of doing proper diagnostics. You may even find it helpful to add some temporary stand alone sensors, like a wide band O2 sensor on each bank, just for the diagnostic value.

Edit: Just saw rv6ejguy's post. If it is indeed a well known system with data logging ability then the above probably is not relevant.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
1,779
Location
Republic of Texas
Link is a high end automotive ECU and it has data logging capability.
Without O2 sensors, will the data generated be adequate to diagnose and remedy issues like this?

I am considering running with O2 sensors if this is a big help.

The Link G4+ offers advanced closed loop system lambda strategy operation, and dual bank lambda control.

30 general purposes tables.

We hope to get efficient fuel flow, avoiding the use of O2 sensors if possible. This would dictate open loop operation.
 
Last edited:
Top