Fixit Algorithms

Let'sgoflying!

Let'sgoflying!

Touchdown! Greaser!
Joined
Feb 23, 2005
Messages
20,315
Location
west Texas
Display Name
Display name:
Dave Taylor
So I fixed another airplane thing today, and again got to thinking about how it all played out. The route to achieve a fix is what often interests me more than the fix itself.

We have all read about problems here and the multitude of offered solutions which follow. Usually we do not hear back to learn the result or what would have been the best approach to solve it after all was said and done.

This was a possible ignition or sensor problem. The engine (IO520K) was running fine.
The first symptom was there is a flashing yellow status light on my P1000 Digital Tach during the last flight. It would play a sort of morse code; ditdit, dit, ditditdit, daah, daah ditdit, but I never could see a pattern. Also the RPM indication would periodically roam. What do you think would be a reasonable first step to tackle this?
 

Attachments

  • P1000 Digital Tach.pdf
    224 KB · Views: 21
I'm not trying to bait anyone. This is about refining our approach to repairs to save money & reduce downtime. Avoiding shooting the Parts Cannon at a problem.
So far people have told me, or I originally thought it might be:
P1000 failure
P1000 wiring defect
Ignition switch
Ig switch connections
P-lead
P-lead terminals
various magneto parts
 
Was it a switch issue or the fine wire into the unit vibrating or broken in sheath?
 
The classic way(and the way I've found the fastest) is the binary system. You list all the possible failure points IN ORDER from input to output. Pick the center one. Then pick the one halfway between the center and the dead end. Keep dividing by two until you hit the problem.
 
It always had a normal mag drop during run-up.
I did a live mag test after one flight and it killed the engine. I jiggled the key and it did not intermittently come back to life.
So, it's not likely the mag or P-leads or ignition switch, right? Probably the P1000 and its wires/connections?
 
I prefer to list the options and start checking the easiest ones first, moving towards the more difficult. When I don't check them in decreasing order of probability . . .
 
i would call the engineering dept of the manufacturer
 
Why couldn't it be a p lead issue? If the engine quits on left mag, I thought it meant the p lead was grounded.
 
OzPilot said:
Why couldn't it be a p lead issue? If the engine quits on left mag, I thought it meant the p lead was grounded.
Click to expand...
Precisely. Or the left mag is dead internally.

Trivial to tell which. Pull the p-lead off the mag and start the engine. If the engine still runs with the left mag selected, it is the p-lead. Otherwise it is the left mag itself.

Jim
 
Let'sgoflying! said:
Where did you get this, please?
Click to expand...

From reading this below. If you mean it killed the engine as in the engine quit, if you were on left mag, most likely left p lead and vice versa for right side depending on which key position it quit on you during ground mag test

Let'sgoflying! said:
I did a live mag test after one flight and it killed the engine.
Click to expand...
 
Maybe we have different meanings to these terms.
When I said a Live Mag test, what I mean by that is to test if any mag is still hot or live, turn the key all the way off. If the engine keeps running I call that a failed Live Mag test. Sorry for the confusion.
The engine did quit when I turned the key all the way off.
Maybe it's my distant British Empire ancestry, here is a NZ description:
Live Mag Check or Test - direct pdf download
 
All ok. Well I am Aussie so CASA had us call this a dead cut which checks for a live mag.

So how did the engine perform when run on left mag only and then right mag only when at idle?
 
Last edited:
The engine ran flawlessly.
It had normal drops (~80rpm) on each side, and quit when turned off with the key.
The first time.
I flew it again. More status light flashing on the tach.
Being a suspicious person and not trusting only one test result, I performed the 'dead cut check' again and this time, the engine did NOT shut down as expected, I had to kill it with mixture.
Questions I asked myself: What items listed in post 2 can be eliminated, and specifically what components are very suspect now? Why did the dead cut checks have two very different outcomes? What is next?
More importantly to me, am I doing the right tests along the way (most efficient, least costly, most common sense).

 
Ignition switch
P-lead
P-lead terminal

Is where I would start... but that is where my ideas end
 
Lean and low idle, run left mag only, then right mag only and back to both. Then left mag only to off. If engine doesn't change. Left side first

If it doesn't cut, I'm pretty sure the p lead grounding would be checked first
 
That sounds like a good plan, I could easily have done it, thanks for bringing that up.
At the time however I was thinking, hm this is an intermittent problem as the 'dead cut check' showed the system killing the engine properly once, then it did not work properly the last time. I have had an ignition switch fail intermittently before so I decided to avoid too much action on the switch while I had it 'apparently not working' so I killed the engine with mixture and pulled the cowl.
I disconnected the p-leads from the noise suppressors (visually inspecting for looseness etc the entire time) and hooked up my vom to one lead at a time then turned the key to see the result. In each position, for each p-lead, the needle swung as expected.
(dang, if the switch is bad, it is intermittent once again I was thinking and if I am wrong then I might buy an unneeded ignition switch. I would hate to spend $150-500 on a switch then discover on the next flight I was wrong.) At some point, I also jiggled the wires on the back of the ig switch to ensure the needle did not flicker while on each mag, none seemed loose (and we are in dry country so not much corrosion on any of these connections)
Would you buy a switch? Or do some other test? Or what would you do now?
 
Last edited:
There is a principal of trouble shooting. Everyone uses it. Consider this.

Your alarm clock doesnt go off. You check to see if it has power. If it doesn't, you plug it in or whatever to fix the power. Point is, if it does have power you still don't know if it will work. But if it doesn't have power you know it WON'T work.

There are lots of tests like this. For instance if the engine passes its mag test, it could still have mistimed (or bad) mag.
 
Taking the position that the ig switch, p-leads to the suppressors were ok (and the p-1000, its wires, connections) were not the problem, that left me with the noise suppressors & their connection to the case of each mag.
I took the suppressor wires (they end in a 'plug' with a hex nut) off each points cover; the soldered tips:
-appeared fine.
-could be felt to contact the spring strip inside with adequate tension before the hex nut bottomed.

I opened the points covers and everything was clean and tight, positioned correctly in both mags.

What was still untested? The suppressors. They were only 2 years old, a previous one had the wire break off so I bought 2 new ones at the time.

I put the ohmmeter across each and both initially tested fine, needle swinging to indicate continuity.
However I then jiggled the terminal of the left one while testing it and noticed a needle flicker back to 'open'. Surely it was just me not making good contact?
I put it on the bench this time with alligators on the contacts and deflected the stem more vigorously, sure enough I had my culprit. The final proof was that all symptoms have disappeared since changing that suppressor out.

 
Once again, this thread is not about my simple ignition problem. I had this all done in less than an hour. All of us involved in maintaining a/c do this sort of thing on a regular basis, whether it be electrical, fuel, mechanical. It is about how it is done, what the mental process is to reach a conclusion.
For many problems in light a/c we do not have an algorithm, logical steps written out for each symptom. The ones in all light a/c maintenance manuals are sadly lacking. Yes most of the problems are simple BUT if that is true why do we continue to see lengthy threads on how someone did troubleshooting for low oil pressure, rough running engine, fuel pressure surge, you name it and every contributor has in their mind a different way to approach the problem? I have a friend who has an autopilot problem for the last year. Respected and trusted avionics shops have so far changed out the servo, overhauled the computer, and now they are changing the clutch and he is over 2000$ now. We have all heard the story of the electrical system issue that resulted in: a new battery, a new VR, sometimes an o/h to the alternator. Crazy!
It has always seemed to me that we could do better, reducing cost and downtime. How could I have arrived at my noise suppressor problem faster? Why did I have to do all those other time-wasting steps first? Maybe someone, even me would have needlessly bought a new ignition switch at one point there?
 
Absent any prior experience with this Tach, my first question; is it as accurate as it was when it was first installed (assuming was somehow tested on installation)? Is it accurate across normal operating range?

I work with a guy I sometimes call the "airplane whisperer," he contends the plane usually tells you what is wrong, if you learn to listen for it.
 
You must log in or register to reply here.
Back
Top