High EGT on one cylinder (it's not the probe)

[Shadrach]

I gave your files a brief look. Have you looked at everything from the flow divider forward? I see no increase in #2 EGT when you lean. It just drops off LOP.

[Cruiser]

Kyle, 

Your EGTs seem to be in line when the FF is 8.0 gph and below. Only when the FF goes above that causes a problem. This is definitely a fuel flow problem, if a valve spring was the cause I do not think changes in FF would made the EGT react the way it is. 

 

Because the EGT only is high when FF is above 8.0 gph may be masking any troubleshooting you are doing at lower settings. You can see the EGT spike at around 10.0 gph and higher and it comes back in line at the lower FF. 

When you switched the injector nozzles, did you move the restrictors also? 

[Super Dave]

I gave your files a brief look. Have you looked at everything from the flow divider forward? I see no increase in #2 EGT when you lean. It just drops off LOP.

 

I think this was just a short test hop with no leaning (is that right Kyle?). The drop off in FF is just a power reduction for descent/landing.

[Cruiser]

snip...

 

I still don't buy that this is a fuel distribution issue. How would fuel distribution cause a cylinder that has always peaked in the high 1400ºs - low 1500ºs to now peak at 1600º+. Also, I don't understand how an engine with poor fuel distribution still has a Gami spread of .2-.3 GPH.

From what I see, I don't believe there is a .2-.3 gph GAMI spread at FF above 10 gph. In fact I don't see any change in EGT on #2 when FF is changed above that which tells me there is a problem with the fuel flow to that cylinder.

The fact that the EGT is slowly increasing would seem to point to an increasing restriction to the flow (above 10 gph) to that cylinder, be very careful now. You could easily get to a point where detonation may become an issue at high power.

[KyleKJ90]

Flight #563 was done by me. It was a short hop from a small grass strip to our home base and I was cleared for a straight in final. It was a tad windy and all kinds of adjustments were made. Our recording rate for the JPI is set to 6 second intervals and for that reason I'd consider the data up to the climb and just shortly after to be anything worth noting. I posted that flight to show the before and after effects of swapping the probes and thus ruling out the thought that it was a probe giving us the issue.

[Shadrach]

From what I see, I don't believe there is a .2-.3 gph GAMI spread at FF above 10 gph. In fact I don't see any change in EGT on #2 when FF is changed above that which tells me there is a problem with the fuel flow to that cylinder.

The fact that the EGT is slowly increasing would seem to point to an increasing restriction to the flow (above 10 gph) to that cylinder, be very careful now. You could easily get to a point where detonation may become an issue at high power.

It's harder to manipulate the files on an iPad. Are you seeing a commensurate spike in CHT?

[Shadrach]

Flight #563 was done by me. It was a short hop from a small grass strip to our home base and I was cleared for a straight in final. It was a tad windy and all kinds of adjustments were made. Our recording rate for the JPI is set to 6 second intervals and for that reason I'd consider the data up to the climb and just shortly after to be anything worth noting. I posted that flight to show the before and after effects of swapping the probes and thus ruling out the thought that it was a probe giving us the issue.

Here as well?

[KyleKJ90]

Here as well?

 

This would have been the same flight only reversed. Was too high on my first approach and then did a go around. Hence the two large spikes in fuel flow.

[Piloto]

Why would that effect only 1 cylinder? Other EGTs are fine as well as the CHTs. I would expect the voltage readout to be could be affected as well.

Different type of probes will be affected differently. Some probes have a grounded side but others not like on the UBG-16. The alternator current flows into the airframe through the engine itself and the + side cable. The current flow through the engine causes small voltage drops that can affect the probes reading. Typical thermocouple probes will put out 0.050 Volts at 1400F. An alternator 20amp current flow through an engine with a 0.01 ohms resistance will generate 0.2 Volts across the engine that can easily affect the probe readings. If one of the probes has a wire grounded or the engine ground resistance is too high it will affect the probe reading. The voltage differential is not constant across the engine.

The above can be checked by pulling out the alternator CB field out.

José

[KyleKJ90]

Dave here is the in flight induction leak checks etc. that we did. 

 

https://www.savvyanalysis.com/flight/909304/3faa364e-3150-4c9d-a3a1-6177ecf13711

[ArtVandelay]

That I'm not sure about. That particular flight took place when a different owner was flying. Are you talking about the two spikes/dips at 00:28 and 1:03 marks?

Exactly.

[KyleKJ90]

That I did. I physically did the labor and flew both of these flights so I can confirm that it's not a probe. =/

 

Here's the flight where we physically swapped the #2 and #4 probes but left the leads the same (#2 on the JPI still read #2 cylinder)

https://www.savvyanalysis.com/flight/909301/a0e24670-c5cb-404a-8acf-1dbfaf61f84b

 

And here's the flight afterwards where I left the probes hooked up to their respective channels (#4 on the JPI would be reading the #2 cylinder and vice versa)

https://www.savvyanalysis.com/flight/909302/30752cab-4741-417b-8da1-30353819efb9

I believe this would've ruled out your theory of it being a grounded probe. Also, since the JPI probes are shielded alumel and chromel thermocouples, wouldn't a ground/short give us an extremely low EGT indication?

[KyleKJ90]

Exactly.

 

I can ask him tomorrow but it would see odd for him to switch tanks twice like that. Then again, I don't know where he planned his fuel stops or his personal flying preferences/techniques.

[Super Dave]

Dave here is the in flight induction leak checks etc. that we did. 

 

https://www.savvyanalysis.com/flight/909304/3faa364e-3150-4c9d-a3a1-6177ecf13711

 

Perfect! Thanks Kyle. 

 

Engine data gurus please check this one out.

 

20:48 Induction leak test

 

22:54 back to full throttle

 

23:06 Big Mixture Pull

 

24:48 finding peak from the lean side (1655º for #2)

 

27:48 inflight LOP mag checks. #2 EGT reaches 1842º!

[Super Dave]

I can ask him tomorrow but it would see odd for him to switch tanks twice like that. Then again, I don't know where he planned his fuel stops or his personal flying preferences/techniques.

 

I think maybe he was just checking to see where peak was for a reference.

[Piloto]

I believe this would've ruled out your theory of it being a grounded probe. Also, since the JPI probes are shielded alumel and chromel thermocouples, wouldn't a ground/short give us an extremely low EGT indication?

Not if the wire that is grounded is the negative or closest to engine ground potential.

José

[KyleKJ90]

Not if the wire that is grounded is the negative or closest to engine ground potential.

José

I think I see what you're talking about. After giving what you said some thought, I took another look at the flight where we first saw the high EGT issue. I noticed the CHT's on cylinders 2 and 4 look rather "noisy" as they fluctuate way more than the CHT's on cylinders 1 and 3. I'm not sure the same can be said about the EGT's but it certainly does lead one to think you might be on to something José. The weather has been complete garbage the past two days and doesn't look to be getting better until later this week. I've bumped up the recording interval of the JPI and we'll test out your theory when the weather shapes up. One quick note though, the alternator CB is the recessed type and can't be pulled in flight. Could a guy pull the ALT FIELD circuit breaker to do this test instead?

[Piloto]

The ALT FIELD CB is perfect.

Note: I believe on the JPI EGT probes wires are grounded on one side. The engine ground path shares with the alternator also. On the EI UBG-16 none of the probes wires is grounded so they are not susceptible to current flow through the engine.

José

[Awful_Charlie]

JPI use ungrounded probes, unless you have a pre 1985 unit - https://www.jpinstruments.com/FAQCategory/probes/(point of interest, factory TIT probe on a Bravo is grounded, these are a PITA as (a). they wear out, and ( . they're expensive!)

 

At the next opportunity, can you set up at the lowest power cruise you can maintain altitude with whilst well ROP, let the temperatures stabilise, and then progressively open the throttle slowly getting to full power over a few minutes, allowing the monitor to record the data at something like 1/2" MP increments and then post the data back (if you put the EDM into 'Lean find' mode, it records at once per second regardless of the recording interval set in the 'Pilot programming' settings). What I'm looking for here is whether the split from the other EGTs occurs suddenly at a setting, or whether the difference gradually separates.

 

If there is a jump and the difference then remains similar, I would think that points more to the valve train, if the difference ramps up and gets progressively more, maybe a flow restriction to the injector (divider, pipe, nozzle). If it is the latter, then doing the same test whilst LOP then the difference should reduce as you add power

[RobertGary1]

Standard procedure for a high egt is to first rotate the plugs (you did that ) and then pull the injectors and soak them in MEK.

Pretty common actually.

-Robert

[Piloto]

From JPI Probes Q&A

How to tell if probes are grounded or ungrounded?

Disconnect the probe from the instrument and measure the resistance between the probe body and the red wire. Also measure between probe body and yellow wire. ‘Ungrounded’ probes will show open circuit. ‘Grounded’ will show 5 – 10 ohms.
Units built after 1985 are grounded.
SN’s less than 1400 are ungrounded.


From Electronics International

http://buy-ei.com/wp-content/uploads/2013/12/cht-catalog-web-ok.pdf

 

2. Grounded vs. Ungrounded Probes –

Grounded CHT probes will pick up ground voltages and noise at
the engine. This can cause shifts in temperature and jumpy readings. Some
instrument manufacturers require the instrument to be grounded at the engine
in order to reduce these problems. A 10mV signal at the engine, which is not
uncommon, can result in a 30 degree shift in the CHT readings.
Most of Electronics International’s probes are ungrounded and are not effected
by ground loops, noise or voltage shifts. In addition, Electronics International
incorporates common mode amps, differential inputs and filters to insure
accurate and stable temperature readings for grounded probes.

José

[Shadrach]

JPI use ungrounded probes, unless you have a pre 1985 unit - https://www.jpinstruments.com/FAQCategory/probes/(point of interest, factory TIT probe on a Bravo is grounded, these are a PITA as (a). they wear out, and ( . they're expensive!)

 

At the next opportunity, can you set up at the lowest power cruise you can maintain altitude with whilst well ROP, let the temperatures stabilise, and then progressively open the throttle slowly getting to full power over a few minutes, allowing the monitor to record the data at something like 1/2" MP increments and then post the data back (if you put the EDM into 'Lean find' mode, it records at once per second regardless of the recording interval set in the 'Pilot programming' settings). What I'm looking for here is whether the split from the other EGTs occurs suddenly at a setting, or whether the difference gradually separates.

 

If there is a jump and the difference then remains similar, I would think that points more to the valve train, if the difference ramps up and gets progressively more, maybe a flow restriction to the injector (divider, pipe, nozzle). If it is the latter, then doing the same test whilst LOP then the difference should reduce as you add power

 

 

You might be on to something.  What would a collapsed intake lifter do to EGT?  Since all intake lobes are shared, it's likely not a lobe.  I'm trying to think of what a wiped exhaust lobe would do to EGT.  Given this happened all at once.  A sticking valve at higher RPM could conceivably raise EGT, but I thought that they had already "rope tricked" the cylinder. 

[FloridaMan]

Try leaning and see if the cylinder goes cold and stops firing first. I would imagine that you could also have an obstruction between the spider and the injector as well. 

 

Also; is there any chance that the "#2" probe might not be on the #2 cylinder?

[carusoam]

Ross,

If I recall properly...

The Cam pushes the rod up, forcing the valve down into the top of cylinder, into the open position.

If that part is right...

A worn cam lobe or collapsed push rod, would fail to open the valve the whole way.

Seems to be the opposite of valve float. Where high rpm or weak springs would fail to close the valve.

Two real life valve failure modes. Failure to open and failure to close.

An intake valve left partially closed would not get the full charge of air and would show up as a higher ROP or colder EGT...

This is just a hypothesis, my memory is just not that strong...(yet.)

How does that sound?

Best regards,

-a-

[Shadrach]

I was thinking that if the intake valve was not opening fully, the mixture would get very rich thus slowing the flame front to the point that the combustion event is still happening during the exhaust stroke thereby causing the high EGT. I know this can happen when an engine is run very LOP. You can witness a second peak as combustion gets so slow that it continues in the exhaust. Not sure if it happens on the rich side. Someone ought to email Mike Busch.