Not wanting to be the next Cirrus Engine failure

SixPapaCharlie

SixPapaCharlie

May the force be with you
Joined
Aug 8, 2013
Messages
16,020
Display Name
Display name:
Sixer
1.1 hr flight today. Everything going as smooth as can be. and suddenly for a about a second, the engine stumbles as if the key were turned to off. Just that bit and right back to running perfectly.

Just had the mags rebuilt, cylinders checked, fuel lines, pump, etc and everything checks out.
This is the 2nd time since that work was done that I have experienced this stumbling.
Putting this thread here in the maintenance section with my stats from today's flight. I downloaded them off the avedyne.

The stumbling occurred somewhere near the last 10-15 minutes of the flight.
I see a dip at one point but that might have just been me bumping the throttle down a bit and over doing it. Several of the lines dip at that point.

One that I am curious about is the AMPB. I don't know what that is and google is not telling me.
But it appears to go crazy near the end of the flight.

I am giving these to my mechanic tomorrow but maybe someone here can see something and go "Dude you are going to make the news" or "nothing to see here"

258pieg.jpg
 

Attachments

  • EngineStats.jpg
    EngineStats.jpg
    94.2 KB · Views: 69
Sorry. the forum appears to have squashed the image. 5 from the bottom is the one I am most curious about (AMPB)
 
What was the ampb high value? -99 amp draw would lead me to believe that there is a shunt issue with the reporting of the amperage (i.e., it is not plausible that this is the case). This would not effect engine performance.
 
Subsea said:
What was the ampb high value? -99 amp draw would lead me to believe that there is a shunt issue with the reporting of the amperage (i.e., it is not plausible that this is the case). This would not effect engine performance.
Click to expand...


60 was the highest
 
Just looking over the other items, anything that starts with AMP seems to be amperage... maybe? AMP2 having spikes to 50 seems wrong.

That AMPB chart looks (from a gut feel only) a lot like some charts of some other telemetry projects I've done when the sensor itself is loose and or going out... it looks way too out of whack to be sane.

See some other oddities like your 28VDC bus having some sort of heavy draw that's pulling the voltage down by a volt or so near takeoff and landing (fuel pump? mega lighting? dunno...) and I have no clue what that AMP2 is really showing, but 50+ amps is a lot of draw even if you had to crank for a bit and had to recharge the battery.

But as far as engine stumbling goes, I'm clueless. In my carb'd airplane I would have said you probably induced carb ice and didn't realize it with the low fuel flows in the decent toward the end and the low CHTs... engine got cooled off pretty good there, about 100C difference between the "big hump" before landing, and landing.

Tried uploading the data here to see if they have better titles on those data points, if you can't find it in the Avidyne manuals? Maybe they have plain english labels for those data fields in the CSV built into their website display.

http://cirrusreports.com/
 
Oh god, here is where I admit something stupid.
I did leave the master on while I ate. I had to have the plane jumped prior to takeoff.
Not sure if that would change the characteristics of the voltages throughout flight.
 
SixPapaCharlie said:
I did leave the master on while I ate. I had to have the plane jumped prior to takeoff.
Not sure if that would change the characteristics of the voltages throughout flight.
Click to expand...
That explains AMP1 amperage slowly tapering off with time (charging the battery(s) back). I would like to see voltages on each bus, and the corresponding alternator amps. Amp2 didn't look like it was putting out (or another reporting problem). Did/do you monitor alternator and battery output during the flight?

Not really enough information for me unless I understand your electrical system (one alternator per battery, independent buses, one emergency bus, ect.)
 
Subsea said:
Did/do you monitor alternator and battery output during the flight?
Click to expand...

I check them every 30 min or so (switch from alt 1 to alt 2 and back to confirm they are charging the batteries)
 
It almost looks as if Alt2 came back online around the time you felt the hesitation (albeit the data is highly unreliable). My plane has an alternator for each battery and they are separate from each other. If one alternator fails, you can cross the other alternator over manually to supply both buses with current. Not sure how yours is wired. Interesting that the master bus is a full volt lower than the essential, and the master bus doesn't "waver" like the essential, but I'm not an Instrument Tech, A&P, nor have I spent any time near a Cirrus.
 
Subsea said:
It almost looks as if Alt2 came back online around the time you felt the hesitation (albeit the data is highly unreliable). My plane has an alternator for each battery and they are separate from each other. If one alternator fails, you can cross the other alternator over manually to supply both buses with current. Not sure how yours is wired. Interesting that the master bus is a full volt lower than the essential, and the master bus doesn't "waver" like the essential, but I'm not an Instrument Tech, A&P, nor have I spent any time near a Cirrus.
Click to expand...

My will fail over automatically.
One alt powers batt1 and main buss and the other powers just the essentials

 
Subsea said:
It almost looks as if Alt2 came back online around the time you felt the hesitation (albeit the data is highly unreliable). My plane has an alternator for each battery and they are separate from each other. If one alternator fails, you can cross the other alternator over manually to supply both buses with current. Not sure how yours is wired. Interesting that the master bus is a full volt lower than the essential, and the master bus doesn't "waver" like the essential, but I'm not an Instrument Tech, A&P, nor have I spent any time near a Cirrus.
Click to expand...

And thank you for all the feedback by the way.
 
That explains the essential buss being a volt higher than the main buss. Learned something new today!
 
Subsea said:
That explains the essential buss being a volt higher than the main buss. Learned something new today!
Click to expand...

Same here, I don't know Cirrus systems... but I don't like how that battery voltage sensor is all over the place toward the end of the flight. I think you have something loose or something wrong with the battery... seriously... I'd be worried about it arcing looking like that. Or the sensor went totally out to lunch.

Now that I can see that's Alternator 1 & 2 Amps and Battery Amps, I don't like what I'm seeing from that battery sensor at all.

That could cause weirdness with the alternators but the numbers for the alternators don't match up with the battery weirdness. So here's hoping we're just chasing an instrumentation problem that's unrelated to your engine sputtering thing.
 
By the way, big thank you to @TangoWhiskey. I didn't realize I could download this data from the plane and upload it to those websites.
I called him today and he goes "Cant you do such and such w/ that plane?"
Me: "I dunno, it has never come up"

Off I go to the hangar with my flash card and wholla!
 
Yeah most of the "advanced avionics" airplanes have a data dump capability these days. It started with the engine monitors doing it and when folks started making integrated avionics systems for GA, thankfully the manufacturers kept doing it. G1000 was a little late to the game but a firmware update eventually added it.
 
What is the resolution of those data points? Meaning how many per second/minute?

We might be better off looking for an issue in the raw data versus a compressed graph. I'd be looking to see an abnormal brief change in egt or fuel pressure. It wouldn't be for much length of time.

The only reason the engine is going to completely stop running is because it lost fuel or spark resulting in no more bangs in the cylinders. An immediate result in egt would have to happen. It may or may not have been measured depending on the frequency.

Hopefully there are at least a few points per second.
 
jesse said:
What is the resolution of those data points? Meaning how many per second/minute?

We might be better off looking for an issue in the raw data versus a compressed graph. I'd be looking to see an abnormal brief change in egt or fuel pressure. It wouldn't be for much length of time.

The only reason the engine is going to completely stop running is because it lost fuel or spark resulting in no more bangs in the cylinders. An immediate result in egt would have to happen. It may or may not have been measured depending on the frequency.

Hopefully there are at least a few points per second.
Click to expand...

I have to double check but I believe it is 1 data point per 6 seconds.
Given that, I know it is impossible to figure out what may have happened in a half a second for sure.

But possible any anomalies in the graphs might provide possible clues as to what would lead to the stumble.
Long shot, I know.
 
Did it occur before or after that power reduction?
 
Realizing that egt's really are relative to basically every airplane.....I'm curious, how were you leaned? At what altitudes were you roughly at each major point in time?
 
jesse said:
Realizing that egt's really are relative to basically every airplane.....I'm curious, how were you leaned? At what altitudes were you roughly at each major point in time?
Click to expand...

sure.
4500. LOP per the Cirrus Lean assist for best economy.
We were just about to begin our descent when it stumbled. Maybe 5 minutes from descent.
 
I'm glad @jesse asked because I was working backward in time from engine shutdown and forgot that some of that was taxi. My assumption, and a bad one, was that you were already in the descent when it happened.
 
I just noticed something interesting. Right around 51 minutes your EGT, CHT, and Fuel Flow climbed up a bit and then remained relatively flat until 59 minutes.

However your manifold pressure and RPM stayed basically the same (rised towards the end of the time, but were dead steady at 51 minutes). So you didn't likely touch the throttle or thrust or warp lever or whatever the hell a cirrus has for a gas pedal.

The rise in fuel flow suggests that at 51 minutes you took yourself out of LOP operations...The mixture was pushed in...or fuel boost was turned on or something.. Fuel flow went up, CHT went up, and EGT went up. All indications that you were LOP and you shoved more fuel into the Cirrus warp drive.

Now the question is..why did you make that change at 51 minutes? Did the engine stumble and that caused you to push the mixture in a bit? If so then 51 minutes is when that stumble happened.

Perhaps it stumbled at 51 minutes because maybe you were leaner then the engine could run reliably and it missed? That's a relatively common thing to happen when if you get too lean while LOP with a big continental... You then shoved the mixture in a bit and took yourself ROP...and that's all there is to the story?

In my opinion, it is unlikely that your engine is about to come from together (your battery, after the abuse you put it through, perhaps).
 
Last edited:
jesse said:
I just noticed something interesting. Right around 51 minutes your EGT, CHT, and Fuel Flow climbed up a bit and then remained relatively flat until 59 minutes.

However your manifold pressure and RPM stayed basically the same (rised towards the end of the time, but were dead steady at 51 minutes). So you didn't likely touch the throttle or thrust or warp lever or whatever the hell a cirrus has for a gas pedal.

The rise in fuel flow suggests that at 51 minutes you took yourself out of LOP operations...The mixture was pushed in...or fuel boost was turned on or something.. Fuel flow went up, CHT went up, and EGT went up. All indications that you were LOP and you shoved more fuel into the Cirrus warp drive.

Now the question is..why did you make that change at 51 minutes? Did the engine stumble and that caused you to push the mixture in a bit? If so then 51 minutes is when that stumble happened.

Perhaps it stumbled at 51 minutes because maybe you were leaner then the engine could run? You then shoved the mixture in a bit and took yourself ROP...and that's all there is to the story?
Click to expand...

So it stumbled and recovered before I did anything.

Lets say the hiccup was 1 second (in plane time that is enough to get your attention)
It stumbled, and went back to running smoothly.

I had 2 other pilots on board and I said "That was really weird"
Pilot #2 (more advanced than me) asked How are your nerves or something to that extent.

I did bump the mixture at that point. and confirmed which tank I was on, and basically scanned everything.

If I were leaner than the engine could run, I would expect it to not recover.
NO question, after the fact, I am giving it a bit of extra fuel to rule out it being too lean.
 
but this is good. in looking at the chart, I think you are right.
I bet the hiccup happened and then we see the EGT / CHT increase briefly as I reacted.
 
Well, I can say that I've seen the very same thing from the very same Continentals while LOP. I've never seen them just quit...and not come back. They miss and pop and give you some warning. If you're right on the "line" it can just happen and then continue to be fine at that setting for another 30 minutes before it happens again.

Fits the data, fits the story, I'd fly in it :)
 
Last edited:
Still go figure out WTH is up with that battery voltage sensor.

You may have fried it with the jump start. That line going all over the place ain't right at all toward the end of the flight. I think you broke it.

That or they knocked something loose during the jump start. Sensor wire, or worse.

Straight up/down square sawtooth patterns on a graph like that aren't good.
 
Next time it happens shoot your eyes to fuel pressure.

If this wasn't my above theory (perfectly normal miss from being lop), it really only can be two things:
1. The mag switch is totally jacked and is intermittently killing both mags. Or somehow both mags p leads are getting grounded out. Unlikely really.

2. Fuel system, pressure is dropping, something is wonky in the fuel controller, air is getting into the fuel line, water, buggers, etc.
 
jesse said:
Well, I can say that I've seen the very same thing from the very same Continentals while LOP. I've never seen them just quit...and not come back. They miss and pop and give you some warning. If you're right on the "line" it can just happen and then continue to be fine at that setting for another 30 minutes before it happens again.

Fits the data, fits the story, I'd fly in it :)
Click to expand...

I have seen the same in cruise flight...and my feeble mind attributes it to the heat soak of the fuel pump. My IO550N will not idle below 1,000 rpm after a flight, although it idles fine during preflight. The boost pump must be running in low speed to cool the pump. Once the electric pump is cooling the fuel pump, the engine will idle down to 600rpm. Heat soak could happen 50 minutes into the flight, and running LOP, there might not be enough flow through it to keep the fuel charge cool.
 
Subsea said:
I have seen the same in cruise flight...and my feeble mind attributes it to the heat soak of the fuel pump. My IO550N will not idle below 1,000 rpm after a flight, although it idles fine during preflight. The boost pump must be running in low speed to cool the pump. Once the electric pump is cooling the fuel pump, the engine will idle down to 600rpm. Heat soak could happen 50 minutes into the flight, and running LOP, there might not be enough flow through it to keep the fuel charge cool.
Click to expand...
We used to have problems with a TSIO-550 quitting pretty much right about the time you'd turn off a runway after landing. Didn't really matter what you did with the mixture. If you "bumped" the fuel prime the moment it started to quit it would continue running and you'd be "fixed" until the next landing. That problem went away on the last annual, not sure what was changed, but theboyz probably knows. The occasional miss while running LOP still happens from time to time. I don't worry myself about it.
 
Right before you bumped the mixture in (which raised temps and fuel flow), #3 cylinder EGT drops. Stuck exhaust valve on cylinder #3 would be my guess on the stumble. Can you correlate a similar drop on #3 when you had the stumble on the Wichita Falls flight you told me about? Was that data able to be downloaded?

Screen Shot 03-07-16 at 11.43 AM.PNG
 
SixPapaCharlie said:
sure.
4500. LOP per the Cirrus Lean assist for best economy.
We were just about to begin our descent when it stumbled. Maybe 5 minutes from descent.
Click to expand...

This can happen if a bit of crud or maybe a spot of water made its way to the fuel spider. Has it happened again?
 
Mark57 said:
This can happen if a bit of crud or maybe a spot of water made its way to the fuel spider. Has it happened again?
Click to expand...
Haven't flown since. Hoping to get up for a bit at some point next week
 
SixPapaCharlie said:
Haven't flown since. Hoping to get up for a bit at some point next week
Click to expand...

OK. I tried interpolating your RPM, MAP and FF from the POH for a normally aspirated SR22. Your posted engine data shows a full rich mixture (I presume for about 5 minutes in the climb). I don't know where you flew from, but I'm guessing you were around 4,000-6,000 MSL. It looks like your running with the mixture setting somewhere on the lean of peak side in cruise flight.

When running a lean of peak mixture setting, I've experienced engine stumble when the spark plugs are nearing the end their useful life and I'm trying to run very lean.

You can try running a LOP mag check under cruise power. That's what I'd check first. Some recommend running them on every flight before you begin descent, as it's a better diagnostic than the mag check we do on the ground. Besides a bad/weak plug, you'll also see if there's more noticeable roughness on a particular mag. And the engine data will show whether the mags are mistimed. Be careful running this lean. It's normal to run slightly rough on one mag lean of peak. But, if it runs sh*t your pants rough on one mag, you're to lean. Start over and try running the test a little richer. About a minute on each mag will give you enough data.

If you identify a culprit plug, I'd be curious whether its Champion or Tempest, whether it had a cracked insulator, massive or fine wire, and its resistance.

I ran Tempest massives before switching to Tempest fine wires. I don't recall exact hrs I had on the massives, but it was well under the 400 flight hrs I'd expected out of them. The center electrode on massive plugs erodes over time and develops a football shaped cross section. I gather it's harder for them to ignite lean mixtures when they get like that.
 
Plane is in annual They are "concerned about compression on 2 of the cylinders" He said they are in the 30s which I find puzzling as 3 weeks ago they said everything was in great shape.
We'll see... If there is an issue, I hope they can overhaul versus replace otherwise this is going to do bad things to my airplane fund.

airplanes are stupid.
 
You must log in or register to reply here.
Back
Top