High EGT and high CHT

P

peter-h

Line Up and Wait
Joined
Sep 24, 2007
Messages
616
Location
UK
Display Name
peter-h
Hi All,

Engine: IO-540-C4D5D

I got GAMIs in 2003; worked fine. Always fly LOP in cruise, and climb with the constant-EGT method.

Over the past year, I have noticed the #3 temps *both* getting higher, and now it is hard (though possible) to keep below 400F during climb.

I did a quick temp v. fuel flow plot (like one does when ordering the injector set originally) and #3 peaks substantially after the others - about 1 USG/hr later. So no wonder that when LOP it is higher than the rest.

It appears to be a blocked injector, and I will get it cleaned.

I've examined the EDM700 downloads (have them all; 800+ hrs worth) and there is nothing that suddenly happened on a given flight. It isn't the probes either; if I fly say 100F ROP then I see the more usual spread, with the 12 temps fairly level and with 1,3,5 hotter than 2,4,6 presumably because the TB20 has the oil cooler on that side so a lot of air is leaking off through that.

I wrote to GAMI but I think my email went in their spam filter.

Any views on possible causes?

I might get all 6 injectors cleaned.
 
Have a look at Lycoming Service Instrcution 1275C:

http://www.lycoming.com/support/publications/service-instructions/pdfs/SI1275C.pdf

This is Lycoming's recommended way of cleaning injector nozzles. Gami might have something different to offer you on their nozzles, so I'd check with them as well.

If you installed your nozzles new in 2003 and haven't cleaned them since, you should definitely clean all of them. They will get clogged over time and do need the cleaning. I just cleaned the ones on my right engine (I knew there was a problem there) and it made a big difference in performance. During the cleaning process, I also ran the fuel pump for a bit with the nozzles off to blow whatever extra junk might be in the lines out. Probably didn't have much of an effect, but I figured it couldn't hurt.

You may have another problem, but 6 years is too long between cleanings, so that makes it a good place to start. It seems that most experienced technicians on these engines who I talk to recommend cleaning them at annual. I'd tend to be in agreement on that from what I've seen.
 
Check the plugs in that cylinder lately? Try swapping both plugs with another cylinder. Also, check the plug wires (harness) to make sure the plugs are getting spark.

400F CHT running LOP is pretty hot.

Let us know what you find, I'll be curious to know.
 
Last edited:
It's not a blocked injector. With a blocked injector fuel flow to the cylinder in question is reduced - it runs leaner than the other cylinders and reaches Peak EGT first. You have the opposite problem. Something else is going on. My immediate suspicion would be ignition, but that's just "suspicion". Have you done an in-flight LOP mag check? That might be telling. I'd also spend the £££'s to get GAMI on the phone and see what they say if you can't get them on email.

 
Interesting points... thank you.

The plugs are iridium plugs which get swapped top/bottom every 50hrs. They were last swapped about 20hrs ago. I do ~ 150hrs/year. The plugs look fine to me, and all look the same except bottom #2 plug which has a bit more oil in it, but is getting better. The high temp cylinder is #3.

I had the engine rebuilt 1 year ago in the USA, SB569A, by www.bpaengines.com. Very pleased with that. They did a lean test on the test bench and the peaks on that look well in line, and the gradual #3 problem has happened since then.

The mags were overhauled when the engine was done. A new ignition harness was fitted then. Mag drops are OK. I have not done a high altitude mag drop test.

My suspicion re the fuel system is because if I fly ROP, the problem is totally gone.

Maybe cylinders 1,2,4,5,6 have blocked nozzles and #3 is OK :)

Anyway I am getting the lot cleaned next week.
 
I've had some feedback, including GAMI's, and the consensus seems to be that a blocked injector would cause a cylinder to peak early (at a higher fuel flow). My #3 is peaking later than the others, as I reduce the fuel flow.

Suggestions are a worn camshaft, valve lifter not pumping up, or an induction leak.

Oil analysis (filter cut open plus oil sample analysis every 50hrs) does not show anything.

The induction leak is presumably only the pipe going from the sump to the cylinder - can it be anywhere else?

I will get the valve lift measured; this is not hard.
 
peter-h said:
I've had some feedback, including GAMI's, and the consensus seems to be that a blocked injector would cause a cylinder to peak early (at a higher fuel flow). My #3 is peaking later than the others, as I reduce the fuel flow.

Suggestions are a worn camshaft, valve lifter not pumping up, or an induction leak.

Oil analysis (filter cut open plus oil sample analysis every 50hrs) does not show anything.

The induction leak is presumably only the pipe going from the sump to the cylinder - can it be anywhere else?

I will get the valve lift measured; this is not hard.
Click to expand...

I'd expect that any valve lifter issues would result in less power on the affected cylinder and that should bring lower CHTs not higher. And since the cam lobes are shared between opposing cylinders a cam problem should affect both about equally.

Induction leaks usually occur in the rubber joint between the intake manifold and the cylinder but there is also the (very unlikely) possibility of a crack in the metal parts, as well as some potential for a leaking manifold drain (there are check valves to prevent this but they can get gummed up). But on a normally aspirated engine, a manifold leak will have virtually no effect on mixture (or anything else) when the throttle is wide open because the pressure inside the manifold is very close to the pressure outside it and the tiny bit of air leaking in will be hotter and less dense than the air coming through the induction system. It is possible for something inside the induction system to divert more than the 5/6ths of the airflow intended for one cylinder but aside from the generic uneven flow front to rear I doubt there's much chance of that either.

Probably the most likely possibility would be that someone put at least two of the injectors into the wrong cylinders at some point and you're just now noticing the effect. GAMI injectors are labeled with letters (usually A,B,C) indicating which cylinders they belong in but if the injector flow rates were ever "fine tuned" for your engine, two injectors with the same letter may have different flow rates.

And JOOC, why move the fine wire plugs around so often? Seems like just more opportunity to damage one. I've been running fine wire plugs for several years and only move them around at annual (150-180 hours) and haven't seen any issues with uneven electrode wear (they just don't seem to wear much at all).

I did have to replace one recently as it would fail to fire at high altitudes when LOP even though it always worked properly on the ground. I think it had over 2000 hours on it at the time but the electrodes looked like new.

One other tidbit: The air pressure in the vicinity of each injector has some effect on the mixture for that cylinder. I know of one or two cases where unexplained mixture variations were resolved by using turbo style injectors which have the atomizing vents plumbed together. Any chance that your baffling is creating a low or high pressure area near the affected cylinder?
 
Last edited:
possibility would be that someone put at least two of the injectors into the wrong cylinders at some point and you're just now noticing the effect
Click to expand...
I did wonder but the injectors were last removed in the engine shop in Tulsa and they did the lean test on the finally assembled engine after they put them back in - obviously :) Nobody has been taking them out since.

I admit I see zero wear on the plugs too - the changing around procedure is just a habit :)

I did wonder about the baffling. There probably are air leaks around it, and I plan to renew the whole lot as per this
http://www.peter2000.co.uk/aviation/tb20-experience/baffles/index.html

But... while a baffling defect would show up as a high CHT (and I know it doesn't take much of a gap to get +30F extra) would this result in a higher EGT? Does EGT correlate with CHT? The other way round (EGT -> CHT) it must do but CHT -> EGT?? I don't know.

The immediate post-GAMI data is here - from 2003. That shows #3 as the hottest then too...

I did a quick and rough lean test the other day. The data doesn't show #4 doing anything like what #3 is doing so I don't think the camshaft has a problem; also there is nothing in the oil analysis suggesting the engine is making extra metal. But #4 is slightly off too, in where it peaks, despite the fact that its temperature is not going high.
 
Last edited:
Peter,

I would still suspect the spark plugs. Check them for cracks or signs of arcing on the affected cylinder(s).

Do a mag check when operating at cruise power and LOP at a safe altitude. Test each mag separately and note any intermittent cylinders or a complete drop out of a cylinder. A normal Mag check will have all of the EGT's rise when running on one Mag.

In the rare event that a Mag has completely failed, the engine will quit when you attempt to run on the bad Mag. Mentally prepare yourself for this possibility and avoid the impulse to immediately change the Mag switch back to both until you have reduced the throttle position to idle to avoid a backfire in the engine due to large amounts of unburned fuel. If the engine was running before, it should resume when you go back to both and add back the throttle and adjust the mixture as appropriate.
 
Peter, try switching BOTH plugs with #3 & #1 cylinder see what the results are. You can't look at a plug and see that it is bad.
 
Geico266 said:
Peter, try switching BOTH plugs with #3 & #1 cylinder see what the results are. You can't look at a plug and see that it is bad.
Click to expand...

No, but the LOP check John suggested (at altitude) will show a failing plug long before it's visually apparent.

Swapping is another good way to isolate a problem, but if the problem is in the magneto the swap won't change much and it's a bit more work.
 
gismo said:
Swapping is another good way to isolate a problem, but if the problem is in the magneto the swap won't change much and it's a bit more work.
Click to expand...


This is true!

I remember several vehicles over the years where a bad plug was the problem. Sure would have been a lot easier just to replace the plugs rather than tear into the ignition systems. I dunno, just trying to help ol Pete.
 
I will try the mag check (at cruise power I assume?) next time I fly.

As for the plugs, they have been swapped around several times, though I can't say for sure they have been swapped between cylinders. Normally I just swap top/bottom on each cylinder.

The plugs were brand new too, April 2008.

I've just looked at an EDM700 download from 2006 and that too shows #3 dominating in both EGT and CHT. What I don't know is whether the temperature peak (as in the GAMI test) was shifted on #3 at that time too. The mags have had two overhauls since then, all new plugs, new ignition harness, and the engine rebuilt to new limits. Not a lot left!
 
Last edited:
I've done the mag check, 5000ft and LOP, and while all the EGTs rise substantially, none of them seems way out of line and certainly none of them falls.

The engine does sound a bit lumpy when doing this.
 
Peter,

Did you do a Gami check at two altitudes and compare results? This would point out an induction leak if there was one. If the number 3 cylinder is still a GPH richer than the rest, GAMI will replace the injector if you send them your data. If you think that your email is getting eliminated via SPAM control, send your email to me and I will forward it to them. Don't forget to include the engine information and your ship to address.

[email protected]
 
I agree that this sounds more like a plug issue than anything..many plug issues can be masked during ROP operations.could check the gapping, timing etc. Needless to say the ignition system has to be no less than perfect when running LOP. ..but on the otherhand exhaust leakage could cause your stated issue.
 
Last edited:
If the problem persists, you may want to post this in the "Maintenance Bay" forum. I am sure some of our excellent A&P's will chime in also. Good luck.
 
I should have posted this under Maintenance. Can the mods move it there, perhaps?

Anyway, today I repeated the GAMI test. The result is as the previous one - a mismatch on #3 and to a lesser degree #4.

Let me summarise the whole thing:

Here is the pre-gami spread (2003):
http://www.peter2000.co.uk/aviation/gami-before.gif

Here is the post-gami spread (2003):
http://www.peter2000.co.uk/aviation/gami-after.gif

Then, I didn't do any GAMI spread tests until very recently, when I noticed #3 EGT and CHT higher than the rest.

So I did a test a few days ago:
http://i101.photobucket.com/albums/m74/peterh337/gami/gami1.gif

and repeated it today (without noting down the fuel values but the X axis goes from 14GPH to 10GPH):
http://i101.photobucket.com/albums/m74/peterh337/gami/gami2.jpg

All were done around 5000ft. The last 2 GAMI tests show essentially the same result.

It's pretty obvious that #3 is peaking well after the others, but also #4 is peaking somewhat later than it should...

I have all the EDM700 data from 2002 to present day but unfortunately no other GAMI spread tests...

The engine was rebuilt to new limits 1 year ago. So we have a known point there. NO possibility of mechanical issues (e.g. stuck valves, worn cam, etc) at that point.

Anyway, looking back through the EDM data, I find the following:

1) #3 EGT has been the highest for the last few years. Before and after the engine rebuild.

2) #3 CHT has been among the highest but I can't really put a finger on a clear pattern in the way it has gradually gone up. It was high before the engine rebuild and is a bit higher since then.

3) There is no evidence that the higher #3 CHT is caused by baffling (lack of airflow) because when I climb full-rich, there are no variations across the six (other than 1,3,5 being hotter than 2,4,6 but that's always been there - presumably due to the oil cooler behind 1,3,5 stealing some air).

4) What is 100% clear is that #3 EGT, and to a lesser extent #4 EGT, have lost their original 2003 GAMI match. The big Q is WHY.

5) Re the possibility of injectors having been swapped around: the only opportunity for this to happen was the engine rebuild 1 year ago, and I have here a printout of the GAMI spread from the dyno test. This shows a close match in the peak EGTs in the dyno test but I have no x axis data and no power (MP) data; just the temperatures. So I don't think the injectors got swapped around. But I can check this easily enough.

6) Oil analysis does not show the engine making metal, which should eliminate the camshaft wearing away and causing the #3/#4 EGT loss of GAMI match.

7) Mag test, 5000ft, LOP, passes fine. All EGTs rise, on both halves of the mag test. The engine sounds a bit lumpy and is obviously delivering less power, but there is no extra vibration. Same incidentally when doing the LOP mag test on the ground. Did all this today.

8) Spark plugs were changed for new ones 1 year ago - iridium RHM38S.

9) Mags were overhauled 1 year ago, and also 1 year before that (not due to any observed defect). A new ignition harness was fitted 1 year ago. The engine has done ~ 150hrs in that year.

I would welcome any suggestions.

Things I can do easily:

Replace both #3 plugs. And #4 plugs too.
Check the markings on the injectors.

Things which are harder:

Check valve lift on #3 and #4.
Check for an induction leak.
Replace/overhaul the mags/harness yet again.

GAMI have suggested this:

Since that cylinder is peaking at a lower fuel flow than the others, that indicates it is richer. A blocked fuel injector would have the opposite effect.

Only a few things could cause a richer cylinder. Basically you are either getting more fuel or less air to that cylinder. A richer nozzle could give you more fuel, but the nozzles don't typically change over time. So that only leaves less air. That could be caused by an induction leak. It could also be caused by the intake valve not opening far enough. You can have your mechanics measure valve lift and duration. I've seen where a lifter isn't pumping up all the way and it made a mixture real rich in that cylinder. It was a real big difference, though. I've also seen where a worn cam lobe (Maybe a bad heat treat on the lobe) made that cylinder richer. I think those two middle cylinders share a lobe and that would explain why #4 is peaking later also.
 
The solution to this, if I ever find one, will be interesting....

The engine delivers exactly the same power as it ever did. 138kt IAS at 11.0GPH, peak or LOP. Before the SB569 rebuild and afterwards, and yesterday. Can't be all that much wrong with it mechanically, IMHO.

The "obvious" conclusion, from the data, is that #3 and #4 injectors have got bigger. Yet everybody tells me this is virtually impossible....

I will get #3 cylinder checked - at least valve lifts which is fairly quick.
 
An update on this...

It was discovered that the GAMI injectors are totally mixed up.

It's quite amazing because it happened without anybody having done it. They must have unscrewed themselves, had a walkabout, then screwed themselves back in, without anybody noticing.

No kidding.

They were also found to be very dirty.
 
So cleaning your nozzles solved the problem?
 
So far, what I have is that the injector numbering was all screwed up. GAMI injectors are numbered and have to go in the right holes. I will test fly it tomorrow.

Quite incredible and I don't know for sure when it happened.
 
Done a GAMI test and sure enough I am more or less back to when the GAMIs were originally fitted.

I have found out who mixed them up, too. No idea why they did it.

The engine runs LOP noticeably smoother. Time to do some vibration tests, like this

http://www.peter2000.co.uk/aviation/vibration/index.html

The spectrums in the above article were all captured, it now turns out, with the GAMIs all over the place.
 
It would be interesting to do the same spectrum analyses with the GAMIjectors in their correct cylinders.

Wondering: could the 1/2 subharmonic be due to load placed on the camshaft, causing a slight decelaration/acceleration event? Just pondering...
 
I think the half-speed harmonic is significant not because of the camshaft rpm but because there is an awful lot of metal being chucked around by the camshaft.

But who knows?

If the engine makers have ever analysed this stuff, they are not writing about it. Lycoming's line, for example, has always been that these engines rotate very slowly and the component weight matching (claimed as a big USP by the various small engine rebuilders) is not important. If they then published where the vibration (which any pilot will tell you IS significant) is coming from, they might do themselves a dis-service.

I will try to do another vib analysis in the next few days.
 
peter-h said:
I think the half-speed harmonic is significant not because of the camshaft rpm but because there is an awful lot of metal being chucked around by the camshaft.

But who knows?

If the engine makers have ever analysed this stuff, they are not writing about it. Lycoming's line, for example, has always been that these engines rotate very slowly and the component weight matching (claimed as a big USP by the various small engine rebuilders) is not important. If they then published where the vibration (which any pilot will tell you IS significant) is coming from, they might do themselves a dis-service.

I will try to do another vib analysis in the next few days.
Click to expand...

I'd be surprised if a reasonably balanced camshaft (i.e. one not missing any important pieces) or the valves themselves could cause a perceptible vibration. IIRC the centripetal force is proportional to the square of the product of RPM and the arm (distance from axis to cg) and both are pretty small
compared to the more likely source, an imbalance in opposing combustion events.
 
Interesting question whether the combustion is capable of producing a harmonic at crank_rpm/2.

On my IO-540 engine, I have 3 combustion events per revolution and indeed see this x3 harmonic.

But a x0.5 harmonic? It's there allright but I don't think it is combustion.

I would agree that there is no way a camshaft can produce any significant vibration.
 
peter-h said:
Interesting question whether the combustion is capable of producing a harmonic at crank_rpm/2.

On my IO-540 engine, I have 3 combustion events per revolution and indeed see this x3 harmonic.

But a x0.5 harmonic? It's there allright but I don't think it is combustion.

I would agree that there is no way a camshaft can produce any significant vibration.
Click to expand...

There are six different combustion events every 2 crank revolutions. If one of them were a bit weaker or stronger than the average, you'd have a 1/2 cycle source. Another possibility is that this is simply the result of the accelerometer placement at the front of the engine. Every time the front left cylinder fires there would be a transient lateral force in one direction followed by a similar force in the opposite direction when the right front cylinder fires. This left-right cycle would occur at half the crank RPM.
 
That's true. However, if this unbalance was present, it would be a second order effect and I would expect the first order effect to be massive, but there isn't anything there.

The accelerometer is placed in the cockpit, clamped to one of the RHS rudder pedal supports, with a G-clamp. I took care that it is sensing the metal airframe and not some piece of plastic trim, etc. This is very easy, safe, and it can hardly be illegal in any way.

It would be interesting to do this analysis on the engine itself, and it does get done (though not recorded in any useful manner) during the prop balancing, but to do this as an airborne test involves running the cable through the firewall, etc. This is one reason why AFAIK nobody here in the UK is willing (officially; I know it's been done) to do airborne prop balancing, whereas in the USA it is fairly commonplace.
 
I've done a new vibration measurement - the original URL is updated.

Not a big change but around a 2x drop in the linear amplitude of the highest peak of the vibration measured on the airframe.

However, I think that provided one has an engine in a reasonably good condition to start with, all these things make only a small difference.
 
You must log in or register to reply here.
Back
Top