Mystery cylinder problem: Cylinder dies in flight, then comes back online.

John Richardson

Pre-Flight
Joined
Sep 12, 2019
Messages
40
Display Name

Display name:
John Richardson
I have a 1979 Piper Dakota with a Lycoming O-540. While turning final today, the engine suddenly developed a mild vibration, and the EDM indicated the #4 cylinder (CHT and EGT) had very low temperature reading, as if the cylinder was no longer firing. All the other cylinders read normal.

After about 5 or 6 seconds, the vibration stopped and the temperature for #4 slowly came back up to normal on the EDM. We did a mag check upon landing and everything was normal. Does this sound like an electrical problem?

* We had the throttle pulled all the way back since we were landing high.
* We had just completed a leisurely 2-hour flight, with no problems.
* The engine is at TBO, around 2000 hours.
* At annual 2 months ago, everything checked out OK.
* Compressions are around 72 on all 6 cylinders.
 
Last edited:
electrical is the only way the cyl could read zero, assuming an electric instrument.
 
You have three electrical connections involved at the cylinder. Two spark plugs and wires, and one CHT probe, and one at the exhaust manifold. None seem likely because they wouldn’t affect two readings at once like that other than perhaps a bad plug causing rich operation and that seems iffy.

Does the EDM have a shared plug on the back of the instrument for both cylinder’s probes that came loose? One connector per cylinder perhaps?

Improper cable run with zip ties cut through probe wires at the same tie point? Shorted?

Just some logical thoughts to think about.
 
electrical is the only way the cyl could read zero, assuming an electric instrument.
It didn't literally read zero, it's a digital bar, not a number. It was so low it didn't register on the monitor. (I editing my post to reflect this)
 
You have three electrical connections involved at the cylinder. Two spark plugs and wires, and one CHT probe, and one at the exhaust manifold. None seem likely because they wouldn’t affect two readings at once like that other than perhaps a bad plug causing rich operation and that seems iffy.

Does the EDM have a shared plug on the back of the instrument for both cylinder’s probes that came loose? One connector per cylinder perhaps?

Improper cable run with zip ties cut through probe wires at the same tie point? Shorted?

Just some logical thoughts to think about.

I too thought perhaps a faulty sensor reading, but the engine was vibrating abnormally for about 5-6 seconds. So, obviously some sort of actually problem, not just a sensor.
 
Can a stuck valve suddenly "unstick" and return to normal?

Yes. Not sure about that engine but really common when valve guides get crudded up on the old engines in the Cessna 150 for example.

You can look up YouTube videos on “the rope trick” to replace parts on some aviation engines without pulling the cylinder.

Often caused by engines that were build for lower lead content have dodgy valve designs, and aren’t leaned super aggressively.

As long as they’re not an interference type and the valve can be stuck open without hitting anything you just lose all compression or most of it on a cylinder and it will misfire. Basically producing no power on that cylinder.

Lots of arguments about proper fixes for it, including debate about the rope trick. Others can start the argument anew here or can just search the same term here.
 
How lean did you have the mixture when this occurred? I had an intake leak for a while and when I pulled the power back toward idle the cylinder with the leak would go lean and quit. Soon as I brought the power up, it would come back to life.

Sent from my Pixel 2 XL using Tapatalk
 
How lean did you have the mixture when this occurred? I had an intake leak for a while and when I pulled the power back toward idle the cylinder with the leak would go lean and quit. Soon as I brought the power up, it would come back to life.

Sent from my Pixel 2 XL using Tapatalk
It was full rich.
 
In that case, my vote is for occasionally sticking valve.

Sent from my Pixel 2 XL using Tapatalk
 
One event isn't enough to diagnose anything positively. If your CHT dropped off? Instrument problem.
 
Ever had a "wobble test", or pulled the valve springs and checked for free movement of valves in the guides? Drop in temp coupled with roughness sure sounds like a sticky valve to me.
 
How lean did you have the mixture when this occurred? I had an intake leak for a while and when I pulled the power back toward idle the cylinder with the leak would go lean and quit. Soon as I brought the power up, it would come back to life.

Sent from my Pixel 2 XL using Tapatalk

It was full rich.

I was thinking along these lines as well. But, your mixture was full-rich. What density altitude? Maybe it was too rich for that cylinder at idle power?
At any rate, induction leak, or valve problem sounds way more likely than an electrical issue of some kind.
 
Last edited:
Maybe one spark plug wire is already bad and the other is going? Would explain it, but unlikely. But you should see this on a mag check.
 
An immediate in-air mag check after an event like this can be helpful. Did both CHTs and EGTs drop? It might be helpful to download your data and take a look at it.
 
I would get the valves in your #4 checked before starting the engine again. If it is a stuck valve, you only get one warning. It will be more likely in a cold engine and you could bend a push rod causing the case damage.
 
Sounds like a stuck valve to me. My Dakota did that and it started fine after cooling down but ultimately needed the Lycoming Service procedure to clean the valve guide.
 
@John Richardson
Whatever the cause, please realize it is incumbent upon you to return here within a month with a full report on the cause of your mystery...too often we hear a problem, throw out suggested causes/diagnostics/corrections....and never hear a word again! Very frustrating to lose that learning opportunity.
(First posted Jan 11, will be watching for your deadline of Feb 11.)
 
I had a stuck valve years ago on a Conti A-65. The A&P said put a quart of Bardahl in the oil. Problem went away and did not come back at least until we sold the plane a few hundred hours later. So maybe some Marvel Mystery OIl or another solvent. I never used MMO in a plane but it has freed up a stuck valve or lifter in an old car (the one in the the picture).
 
I had a stuck valve years ago on a Conti A-65. The A&P said put a quart of Bardahl in the oil. Problem went away and did not come back at least until we sold the plane a few hundred hours later. So maybe some Marvel Mystery OIl or another solvent. I never used MMO in a plane but it has freed up a stuck valve or lifter in an old car (the one in the the picture).

It’s always nice when the mechanic says to do EXACTLY what the manufacturer says NOT to do — linked in this thread MULTIPLE times. Sigh.
 
My local engine shops advise to add MMO to the oil for morning sickness. Not so much as a solution but as a diagnostic tool. But that’s for morning sickness. This topic is about an unusually random event.
 
Especially if it actually works. Of course manufactures know everything which is why we never have ADs.

Or they’ve seen all the damage caused by not investigating properly and repairing valve guides that are junk that broke other stuff. LOL.

We wrote “don’t do that” documents for a reason when I worked in support of a manufacturer. If people did it anyway we always laughed at them after quoting them the repair price to do it right.

Dumping some petroleum distillates in the fuel tank is just lazy. Lots of lazy people get away with it, some don’t. Doesn’t actually fix the problem but most change their behavior with the red knob along with getting lucky they unstuck the crud in the already worn guides.

Oh well. Problem will return until repaired correctly.

WD-40 fixes everything too out in my garage when I want to be cheap. LOL. And duct tape.

One buddy claimed MMO was great until the third time he lost a cylinder in the takeoff run. Then the rope trick, proper reaming, and parts replacement was done. Guess what actually fixes bent, worn, and such metal things? Not sewing machine oil. LOL.

Just sayin’. It lasts for a while. Spin the wheel and hope it doesn’t stick at 100 AGL. Lots prefer that way. One armed bandit, give a pull.

In this case there’s still other things to check but the point here was how many “professional” mechanics flat ignore the manufacturer on this one.

The AD comment is bogus. Most of these older power plants have very few of them and they’re well known. This crap is OLD. The majority of ADs are defects like the place that casted all the crankshafts completely wrong. Not worn valve guides and bent valves with crud in them from too much lead and too much red knob forward.
 
As I clearly said I never have used MMO in an airplane but it DID STOP valve/liter noise in the pictured auto. And Bardahl did free up what my A&P said was a stuck valve on my A-65 whether anyone, including a manufacturer, like this result or not.

Of course one should never dare question the judgement of a manufacturer. Et tu Boeing?
 
As I clearly said I never have used MMO in an airplane but it DID STOP valve/liter noise in the pictured auto. And Bardahl did free up what my A&P said was a stuck valve on my A-65 whether anyone, including a manufacturer, like this result or not.

Of course one should never dare question the judgement of a manufacturer. Et tu Boeing?
Correlation doesn't guarantee causation. Still, thinning out the goo can certainly unstick a valve, or make ticking lifters quiet down (if they are hydraulic). But I'd rather attack the root cause.
 
@John Richardson
Whatever the cause, please realize it is incumbent upon you to return here within a month with a full report on the cause of your mystery...too often we hear a problem, throw out suggested causes/diagnostics/corrections....and never hear a word again! Very frustrating to lose that learning opportunity.
(First posted Jan 11, will be watching for your deadline of Feb 11.)
So true. This site has proven to be a wealth of knowledge for me. I’d love to give back, but unfortunately because the engine is already at TBO and I vowed to myself that I would perform an overhaul at the slightest hint of mechanical trouble, I made the decision to get a remanufactured engine. So, I won’t be able to investigate the cause further. Thanks for all the input.
 
Correlation doesn't guarantee causation. Still, thinning out the goo can certainly unstick a valve, or make ticking lifters quiet down (if they are hydraulic). But I'd rather attack the root cause.
Me too. I have little hesitation on pulling a cylinder which I do not see as a big job but then I work for free. However, I might try a solvent first since I doubt if it would result in damage and might just work..
 
Of course one should never dare question the judgement of a manufacturer. Et tu Boeing?
Of course a manufacturer should be questioned in certain cases. But the questions must come from people who actually have wide experience and knowledge and more than a little engineering skill.

ADs usually arise when Service Difficulty Reports (SDRs) are filed by mechanics or Directors of Maintenance when unusual defects or wear shows up. Those reports are forwarded to the manufacturer, who has engineers that examine them and look at the possibility that something is indeed wrong, and with the FAA they write up an AD to force repairs before someone dies due to the problem. More SDRs are a sign that the problem is not isolated but widespread and needs addressing. I filed a lot of SDRs in my time.

The ignorant folklore that abounds in the piloting community is just amazing, and that folklore is often responsible for damaging practices and accidents. Pouring unapproved stuff into your tanks or crankcase makes you a test pilot, and even if it doesn't cause an accident it might eat your fuel hoses or carb/injector seals or engine internals. Adding strong stuff to a crankcase could loosen a whole bunch of varnish and sludge that soon overwhelms the filter and plugs it solid, or gets into places it has no business being, like the oil pump, which can get chewed up.
 
Back
Top