What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Get the lead out! Not so fast!

While the article may be accurate, I'd appreciate seeing the data behind its claims. Aviation is not a faith based endeavor. ;-)
 
Interesting. GAMI is/was developing unleaded avgas. Are they just pre-warning us of this downside? I don't buy the point about the 96 octane could not be sold because people believed they needed 100. If the FAA approves a fuel as a direct substitute, most folks will use it.

It's well known that many supercharged radials had to be de-rated when switched from 115/145 or 100/130 to 100LL. This impacted maybe 5% of aircraft at best 20 years ago. Today, even fewer of those aircraft are in service.

Most naturally aspirated opposed engines will not have to worry about the octane ratings of the newer unleaded avgas coming along since they can run on 91 mogas already with some care. Cruise performance is not impacted at all with well developed IE systems.

I believe if valve recession is really an issue, the aero engine makers and aftermarket will offer hardened valves and seats. I think the unleaded fuels are coming regardless because of the environmental concerns. Swift already has their low octane fuel on the market. Maybe the fuel makers can add something to mitigate this concern as well?

This could all "spark" some new SI engine development or hasten the switch to diesels too.
 
Last edited:
All hail Mr. Braly! He is brilliant for sure.

He wrote, "Unfortunately, if some guy buys an engine that has been run on a dyno with 100LL and then gets a little 100LL during the life of the engine run mainly on unleaded mogas, and then goes to full TBO, he can write an article on the internet that claims valve recession is a myth, and people believe him.", in his article that Bill linked us to.

Now, I think many of us know this guy who has pretty much done exactly this and has passed 3,200 hours successfully.

I am not saying that Mr. Braly is wrong or in any way missing the mark. He is always spot on.

What I am confirming is that his statement about derating or detuning the performance of these engines will aid in reducing valve recession.

Our un-named 3,200 hour case study primarily operates at 60 percent OR LESS power most of the time.

Excellent article! Bill, thank you for bringing it to our attention!

:) CJ
 
Chicken Little lives.

I am particularly amused by the author relating "knock complaints" in the switch from 100/130 to 100/130LL, and warning that these complaints will rise when we make the fuel switch ("But the remaining 10% to 15% or so may notice pinging or knocking under some conditions. This will necessitate detuning or de-rating the aircraft.")

Exactly nobody can hear pinging in an air-cooled aircraft engine.

Only the folks with individual cylinder monitoring can detect it otherwise. So how many 450 Pratts have individual CHT/EGT on the panel? And if your NA Lyclone showed signs of detonation, I'd like to think owners of such instrumentation have enough knowledge to understand what their expensive EFIS is telling them...and how to make it stop.

Ok, I'll concede maybe the guys with 10:1 pistons might decide to detune. And my B-36 isn't going to get flown as much ;)
 
GAMI seems to be backtracking on what they said in 2014 with their G100UL fuel being run in Embry-Riddle training 172s:

"No issues were found," Roehl reports. "We just lack a couple of other tests to complete our first STC on the 172."

Now they say valve recession may be an issue. They've been running tests on this fuel since 2010 I believe. They need to update some of their website info.

Significantly, GAMI's G100UL fuel is not part of the FAA PAFI program where Swift's fuel is. Swift 94UL is on the market now and their UL102 is being testing under PAFI. Sour grapes?
 
Last edited:
The high output musclecar world went through similar consternation with the disappearance of leaded fuel decades ago. Valve seat recession was fixed with the addition of hardened exhaust seats. Done. Fixed. Easy. Any reason why we can't do that here if the fleet starts loosing valve seats?

Detonation? Based upon what we now know about ignition advance, that can be largely fixed with an appropriate advance curve at high power settings. I predict the other EI manufacturers are going to have to follow Ross' lead and allow people to set their own slope, or Ross is going to be selling a bunch of CPI ignitions.
 
Last edited:
De-rating for lower octane as applied to forced induction engines usually imposes a lower manifold pressure limit.

For atmo (naturally aspirated) engines, usually ignition timing would be reduced to maintain safe detonation margins. For Experimentals, this is easy to do with user programmable timing on the SDS CPI or EM-5 systems. You'd just add a little more retard at higher manifold pressures which takes about 30 seconds.

This generally would not affect cruise power or economy since cylinder pressure is a lot lower at cruise power/ MAP and you can run optimal advance.
 
Last edited:
Chicken Little lives.
...

Exactly nobody can hear pinging in an air-cooled aircraft engine.

...
That's exactly what I thought when I read this article.

...

Detonation? Based upon what we now know about ignition advance, that can be largely fixed with an appropriate advance curve at high power settings. I predict the other EI manufacturers are going to have to follow Ross' lead and allow people to set their own slope, or Ross is going to be selling a bunch of CPI ignitions.

Or the right set of ignitions coupled with the right monitor and some knock sensors. ;)
 
That's exactly what I thought when I read this article.



Or the right set of ignitions coupled with the right monitor and some knock sensors. ;)

Knock sensors are hard to implement on an independent cylinder layout such as the Lycoming without individual ones for each cylinder. If one wanted to experiment though, the CPI and EM-5 already have knock sensor capability. Maybe screw them into the primer ports?
 
Knock sensors are hard to implement on an independent cylinder layout such as the Lycoming without individual ones for each cylinder. If one wanted to experiment though, the CPI and EM-5 already have knock sensor capability. Maybe screw them into the primer ports?

Correct, except we have a different location in mind that seems to work very well. The real trick is using...

I will give more details when we are ready. We have been working on them for more than a year.
 
If you can figure out a reliable knock sensor that will be a big step foreward.

...but it should not be a crutch for an overly aggressive advance curve, it should be a fine tuning aid.
 
Knock sensors are hard to implement on an independent cylinder layout such as the Lycoming without individual ones for each cylinder. If one wanted to experiment though, the CPI and EM-5 already have knock sensor capability. Maybe screw them into the primer ports?

We experimented with this in the old Porsche air cooled boxer engines (very similar in case/cylinder design to the Lyc). It is doable but requires a lot of work. It's not just placement, it is as much or more about finding the right frequency. Most EFI systems look for a contact closure or a voltage signal to indicate knocking. Therefore you need to find the frequency associated with detonation and install a sensor designed to pick up that frequency. OEM Engineers spend a lot of lab/dyno time figuring this out for every engine

I was able to wire up a DIY sensor that went into a microphone, amplifier and headphones. I used a location that others had found to be workable on the case (didn't need to go on the individual cylinders) and was able to hear the detonation, but could not identify the frequency and therefore could never rig it up to my EFI system. I don't know if that case location would have produced a strong enough representation of that frequency to trigger a sensor to fire. A big challenge is competing noises. You had to find the location that would best increase the amplitude of the detonation and reduce the others.

Larry
 
Last edited:
If you can figure out a reliable knock sensor that will be a big step foreward.

...but it should not be a crutch for an overly aggressive advance curve,

This is how the auto OEMs do it. They keep advancing timing (within some pre-defined limits-they won't go past the lab proven point where > advance = >power) until they get a knock signal than back down. Still knocking back down more, all in real time. This is pretty standard and how they deliver varying performance with varying octane level fuels available across the globe. They, however, have well designed and proven knock sensor setups that can be relied upon to pick up very early stage detonation, before it becomes dangerous.

Larry
 
Last edited:
We've done knock sensing for nearly 20 years and done a large number of Porsches but the factory has already placed the sensors well, with a sensor for each bank and an acoustic bridge between cylinders.

Key on any installation is placement and being able to filter out extraneous noise which can be mistaken for knock.

Not all OEMs use full on knock sensing at all times. That went very wrong for Subaru on US spec STis and they had a lot of engines to replace in 2008-2009MY because of an overly aggressive advance strategy and slow retard strategy. At WOT, all OEM ECUs I've ever evaluated through the OBDII immediately retard timing a bunch as soon as MAP goes past a certain point since you know you're going to knock there with 45 degrees of timing in and there must be a reasonable safety mechanism in place if a knock sensor fails. Open loop mapping is still important as a fall back. What happens when a knock sensor fails or the knock wire breaks and you have no backup open loop strategy? The engine blows up.

We've never recommended to use knock as the primary timing tool, preferring to have realistic timing curves mapped and use knock sensing to modify those in case of bad fuel, wastegate failure etc. Performance, racing and aircraft applications are much different than OEM automotive though.

Most OEM ECUs for a number of years have used DSP to track which cylinder was just fired and a single or twin knock sensors to track which cylinder knocked. They can retard timing in that cylinder only or any which knock. Very sophisticated stuff running millions of lines of code and processors more powerful than many PCs.

I'll be running our Continental O-200 on the test stand next month and plan to do some tests with knock sensing on that to see how feasible it is with a single or twin sensors on a pushrod, air cooled, individual cylinder design. Will put some 87 octane mogas in it and see how far we can push it. With the EM-5 PC data logging we can observe knock activity and map the best strategy to deal with it. Should prove interesting. If we learn something new there, that will be incorporated in the next CPI and EM-5 software releases so we're ready for that "low octane" unleaded avgas.
 
Last edited:
measuring combustion pressure

I've read that the holy grail is to have a reliable, cheap way to measure combustion pressure. I've spoken to some of the suppliers in Europe, and these babies are $3000 and up - each. And they are fragile. Until these combustion pressure sensors become cheap and robust enough, I think the solution will be a system like Ross's and Robert's where they use these expensive sensors on their test platform to find an indirect way of calculating combustion pressure and can ensure that the engine computer optimizes timing and the air/fuel ratio for every MP, RPM and engine load.

https://www.avl.com/pressure-sensors-for-combustion-analysis1

https://www.kistler.com/au/en/appli...ombustion-analysis-cylinder-pressure-testing/
 
How about.......program a specific algorithm into the ECM to see CHT speed of temp ramp up and EGT to identify Knock, a lot of us already have these sensors in place. Then the program a learning curve for the MAP?
 
The article is an interesting read regarding lead in our fuel. Remove lead and valve recession rate will go up.

But no mention of higher than normal EGT's running LOP (another issue, granted) and exhaust valve failure due to continuous high temperature environment.

It seems the farther we go from original design parameters of 75 years ago, the more problems come up with these old technology engines.

The real cure would be for Lycoming and Continental to come up with a clean sheet engine that would meet the needs of today and not rely on band aid fixes to be aligned with the world as it is today.

They could do it if they put their minds to it.
 
Through a couple of different abnormal operating situations, I've experienced events that each resulted in rapidly rising cht. Detonation; likely. My point for relating this is, once the condition begins and temps get high, you need to act fast. Like immediately. CHTs can rise 2, 3 degs a second. Point being, you should be very well instrumented, and know what you're doing to get out of the condition, if you are reaching out into knock territory. Things can get exiting really quickly. And no, you can't hear it in the plane.
 
I am again amazed at the collective knowledge available via this website. All I know about knock sensors is that when someone knocks at the door a dog barks. :)
 
How about.......program a specific algorithm into the ECM to see CHT speed of temp ramp up and EGT to identify Knock, a lot of us already have these sensors in place. Then the program a learning curve for the MAP?

Almost anything can be done in software if you develop a sound strategy based on lots of actual testing. That's the time consuming and expensive part, not writing the code. We always have to look at failure detection and corrective action as well which is often more time consuming than the base control code. We'd need some more inputs to handle the extra thermocouples as well which would mean a new ECU design.

Eventually, we'll see timing maps based on CHT, AFR, IAT inputs as well as rpm and MAP. This may get us a couple more percent efficiency. Right now, we're being conservative so we never get to that detonation corner. Lycomings don't suffer as much loss as some other engines with retarded timing which is a good thing.

A clean sheet design would be nice but the old stuff is so entrenched and widespread, I don't see that happening in any numbers for at least a decade, if ever. Maybe the aftermarket can put out some newer heads with knock sensor bosses. We already have heads with 14mm plug holes to take automotive plugs and accessory covers with no provision for mags. These are both in response to the way EI have changed the Experimental ignition landscape.
 
Last edited:
...But no mention of higher than normal EGT's running LOP (another issue, granted) and exhaust valve failure due to continuous high temperature environment...

A few points:

Lycoming has approved running continuously at peak EGT for decades now. LOP ops is cooler than this, by definition. As a bonus, CHT comes down with LOP too.

With few exceptions, Lycoming does not specify a maximum EGT. The primary exceptions being the turbo models. For most of us, this means you can run whatever the combustion process will produce without fear. 1600? No problem.

Valve seat recession is a reasonable expectation, but again, that was solved decades ago in the automotive world with a simple retrofit of hardened valve seats into the 75 year old technology (in some cases) car engines. I'd bet Lycoming, Continental and the cylinder rebuild shops will soon be installing hardened seats for all cylinder work and the "problem" will go away just as it did in the classic car world.
 
A few points:

Lycoming has approved running continuously at peak EGT for decades now. LOP ops is cooler than this, by definition. As a bonus, CHT comes down with LOP too.

.

Your information on Lycoming and LOP may be dated. I have Lycoming Tech Tips, picked it up at OSH not long ago, says on page 43 relative to LOP with FI engines and carb engines, "In any case, leaning past the peak is not recommended".

The part about leaning past the peak is in bold print.

Or maybe my info is dated, but it is a recent publication.
 
Last edited:
Lycoming rep

I just talked to a Lycoming rep at the AOPA flyin at Camarillo. He reiterated that Lycoming does not recommend running lean of peak due to decrease detonation margin. Be is aware that several O320 (carb) guys routinely do this. But he acknowledge that if an injected engine has full analyzer suite and tuned injectors, that lean of peak might be OK but to remember one is decreasing detonation margin. He again expressed concern with the carb O320 guys running lean of peak. Last thing he said was that lean of peak with an injected 320 needs to have careful baffling to avoid a hot cylinder
 
Last edited:
I just talked to a Lycoming rep at the AOPA flyin at Camarillo. He reiterated that Lycoming does not recommend running lean of peak due to decrease detonation margin. Be is aware that several O320 (carb) guys routinely do this. But he acknowledge that if an injected engine has full analyzer suite and tuned injectors, that lean of peak might be OK but to remember one is decreasing detonation margin. He again expressed concern with the carb O320 guys running lean of peak. Last thing he said was that lean of peak with an injected 320 needs to have careful baffling to avoid a hot cylinder

Counter to the what the science says. Amazing in 2017, Lycoming continues to contradict the facts about detonation margins and CHTs running LOP. Agree you need proper engine monitoring instrumentation though.
 
Your information on Lycoming and LOP may be dated. I have Lycoming Tech Tips, picked it up at OSH not long ago, says on page 43 relative to LOP with FI engines and carb engines, "In any case, leaning past the peak is not recommended".

The part about leaning past the peak is in bold print.

Or maybe my info is dated, but it is a recent publication.

I think you may be missing the point. Peak EGT is the hottest it can be. Leaning past peak is cooler.

Lycoming says it is OK to run the EGT at peak (the hottest possible temperature) indefinitely. So if exhaust valve temperature is the LIMFAC as you imply in your earlier post, then going LOP and cooler should be better, right?
 
I just talked to a Lycoming rep at the AOPA flyin at Camarillo...

This is contrary to actual experience. The detonation charts are available and slightly rich of peak is a greater detonation risk than slightly LOP. That and the detonation range drops off sharply LOP and quickly becomes chemically impossible for all practical purposes after (IIRC) about 50 LOP. At any rate, 50 ROP is much more in the danger zone than 50 LOP.

The comment about cooling baffles leads me to believe that this rep has never actually flown LOP and seen how the CHT will plummet. It is so effective that many of us use LOP as a cooling aid on a long hot climb. Cools things off a lot quicker than dumping a bunch of fuel in there.

I'm shocked that in this day and age these OWT live on. Shame on Lycoming for promoting such disimformation.
 
This is contrary to actual experience. The detonation charts are available and slightly rich of peak is a greater detonation risk than slightly LOP. That and the detonation range drops off sharply LOP and quickly becomes chemically impossible for all practical purposes after (IIRC) about 50 LOP. At any rate, 50 ROP is much more in the danger zone than 50 LOP.

The comment about cooling baffles leads me to believe that this rep has never actually flown LOP and seen how the CHT will plummet. It is so effective that many of us use LOP as a cooling aid on a long hot climb. Cools things off a lot quicker than dumping a bunch of fuel in there.

I'm shocked that in this day and age these OWT live on. Shame on Lycoming for promoting such disimformation.

Definately waaaay off topic from the original post, but I would like to add that Lycoming has said this for many decades and sees no reason to alter their position on the subject of LOP operations. They know that we do it and it causes no problems.

Meh, it really doesn't change the way we operate.

:cool: CJ
 
How about.......program a specific algorithm into the ECM to see CHT speed of temp ramp up and EGT to identify Knock, a lot of us already have these sensors in place. Then the program a learning curve for the MAP?

That's a very reasonable idea. Not sure you need an EGT alarm, and it might be hard to make it work. The CHT alarm, however, would just be an extension of a feature already included with many engine monitors...the "shock cooling" function. Just as shock cooling monitors cooling rate, we want a little bit of code to monitor heating rate.

There's plenty of headroom. Detonation generally doesn't hurt anything if CHT is not allowed to get high. When it starts, the CHT digits on the EM will click upward like the digits on a gas pump, so the code needs to trip an alarm when rise rate exceeds (spitballing here) 30~40 degrees per minute.
 
Last edited:
We've done knock sensing for nearly 20 years and done a large number of Porsches but the factory has already placed the sensors well, with a sensor for each bank and an acoustic bridge between cylinders.

Key on any installation is placement and being able to filter out extraneous noise which can be mistaken for knock.

Not all OEMs use full on knock sensing at all times. That went very wrong for Subaru on US spec STis and they had a lot of engines to replace in 2008-2009MY because of an overly aggressive advance strategy and slow retard strategy. At WOT, all OEM ECUs I've ever evaluated through the OBDII immediately retard timing a bunch as soon as MAP goes past a certain point since you know you're going to knock there with 45 degrees of timing in and there must be a reasonable safety mechanism in place if a knock sensor fails. Open loop mapping is still important as a fall back.

We've never recommended to use knock as the primary timing tool, preferring to have realistic timing curves mapped and use knock sensing to modify those in case of bad fuel, wastegate failure etc. Performance, racing and aircraft applications are much different than OEM automotive though.

Most OEM ECUs for a number of years have used DSP to track which cylinder was just fired and a single or twin knock sensors to track which cylinder knocked. They can retard timing in that cylinder only or any which knock. Very sophisticated stuff running millions of lines of code and processors more powerful than many PCs.

I'll be running our Continental O-200 on the test stand next month and plan to do some tests with knock sensing on that to see how feasible it is with a single or twin sensors on a pushrod, air cooled, individual cylinder design. Will put some 87 octane mogas in it and see how far we can push it. With the EM-5 PC data logging we can observe knock activity and map the best strategy to deal with it. Should prove interesting. If we learn something new there, that will be incorporated in the next CPI and EM-5 software releases so we're ready for that "low octane" unleaded avgas.
Thanks for the info Ross.

I'll give you a call in a couple of days to discuss our strategy.
 
We've done knock sensing for nearly 20 years and done a large number of Porsches but the factory has already placed the sensors well, with a sensor for each bank and an acoustic bridge between cylinders.

Key on any installation is placement and being able to filter out extraneous noise which can be mistaken for knock.

Not all OEMs use full on knock sensing at all times. That went very wrong for Subaru on US spec STis and they had a lot of engines to replace in 2008-2009MY because of an overly aggressive advance strategy and slow retard strategy. At WOT, all OEM ECUs I've ever evaluated through the OBDII immediately retard timing a bunch as soon as MAP goes past a certain point since you know you're going to knock there with 45 degrees of timing in and there must be a reasonable safety mechanism in place if a knock sensor fails. Open loop mapping is still important as a fall back.

We've never recommended to use knock as the primary timing tool, preferring to have realistic timing curves mapped and use knock sensing to modify those in case of bad fuel, wastegate failure etc. Performance, racing and aircraft applications are much different than OEM automotive though.

Most OEM ECUs for a number of years have used DSP to track which cylinder was just fired and a single or twin knock sensors to track which cylinder knocked. They can retard timing in that cylinder only or any which knock. Very sophisticated stuff running millions of lines of code and processors more powerful than many PCs.

I'll be running our Continental O-200 on the test stand next month and plan to do some tests with knock sensing on that to see how feasible it is with a single or twin sensors on a pushrod, air cooled, individual cylinder design. Will put some 87 octane mogas in it and see how far we can push it. With the EM-5 PC data logging we can observe knock activity and map the best strategy to deal with it. Should prove interesting. If we learn something new there, that will be incorporated in the next CPI and EM-5 software releases so we're ready for that "low octane" unleaded avgas.


I hope to see a write up / video on this!
 
I hope to see a write up / video on this!

We plan to do a series of EM-5 tuning videos showing how to address each parameter using the test stand engine and hopefully show the data logging runs as an aid to that. Once those are complete, I'll get around to some knock testing. Unfortunately the O-200, does not have many suitable places to install knock sensors but we'll experiment and try to learn something even if we have to use a VC screw temporarily to mount them.
 
I think you may be missing the point. Peak EGT is the hottest it can be. Leaning past peak is cooler.

Lycoming says it is OK to run the EGT at peak (the hottest possible temperature) indefinitely. So if exhaust valve temperature is the LIMFAC as you imply in your earlier post, then going LOP and cooler should be better, right?

I don't know.

What I do know, when attempt LOP with this Barrett Superior IO360 180HP engine with AFP FI system, it runs smooth until LOP. Not very far on that side engine begins to run rough so I don't do it. What for?

Is it detonating or just not enough fuel to make engine happy, I don't know.

Engine definitely runs better on rich side of peak where it usually is.
 
I don't know.

What I do know, when attempt LOP with this Barrett Superior IO360 180HP engine with AFP FI system, it runs smooth until LOP. Not very far on that side engine begins to run rough so I don't do it. What for?

Is it detonating or just not enough fuel to make engine happy, I don't know.

Engine definitely runs better on rich side of peak where it usually is.

That rough running is characteristic of ignition system's inability to light off the lean mixture, not detonation. If you were detonation prone, you would be seeing it in the main "danger zone" which is slightly ROP. I'm guessing you are running magnetos? If so, then a switch to ANY electronic ignition will make that problem go away. Well, it will still stumble as you go leaner, but it will happen well past any practical mixture setting. As an example, the RV-8 I fly would just barely get to peak EGT without stumble on the Slicks. As soon as I installed the Pmags, it would go to as much as 200 degrees LOP and remain smooth as a sewing machine. Of course, 200 LOP is hardly making any power at all and the CHT was so cold it was hard to tell if it was still running, but it does illustrate the difference between magnetos and EI.

And if you cant run LOP smoothly, then I dont blame you for not doing it... I wouldnt either. Throw an EI on one side and that will change your whole outlook.
 
Detonation

Before I had "learned" any of theleanng stuff from VAF, I had operated a number of engines for over 2000 hours. One was a Piper Lance PA32R with a 300 hp IO540. We ran this 100? ROP in climb and @ peak for cruise. The engine went over 2200 hrs. with no undue wear and no evidence of detonation. Main bearings showed some wear. There was some attrition of the exhaust valves. Nothing major. The overhaul was done because of slightly increased oil consumption and total time. An earlier Comanche PA 24 260 was run in the same fashion and it went beyond 2300 hrs. with nothing in the engine showing much except significant main and cam bearing wear. CHT's were kept below 380, probably because we usually cruised at 65% or less. The Pitts S2-A with a 200 hp IO-360 was run about 1500 hours with same leaning goals, except that we did serious acro @ 130-150 ROP, full throttle and 2500-2700 RPM. Incidentally, all of these had a single cylinder EGT probe. The Lance later got a six cylinder monitor. All climbs were either at full throttle or 2500 square 100 ROP. Serious question.
Why weren't these engines destroyed?
 
Serious question.
Why weren't these engines destroyed?

They shouldn't. They didn't. Nothing you describe is pro-detonation, assuming a small GAMI spread.

Personally I think you were running them exactly right, assuming you want best power for the given MP and RPM. For best economy, add LOP operation to the menu.
 
Balancing the injectors would be an easier less expensive solution and, IMHO, should be done anyway for a smooth running engine. Don Rivera at Airflow Performance can help you with that.

Leaner mixtures do require more spark advance. So even a balanced engine (meaning narrow GAMi spread, not physical engine balance) will run rough just a little bit lean of peak on mags. Run the timing up, and smooth as glass.
 
Leaner mixtures do require more spark advance. So even a balanced engine (meaning narrow GAMi spread, not physical engine balance) will run rough just a little bit lean of peak on mags. Run the timing up, and smooth as glass.

My experience with my IO-540 using standard timing with Slick mags is that, after balancing the injectors to peak within .4 GPH, I could go significantly lean of peak. I don't have the degrees lean, but I could get ~8 GPH (8000' running 2300 RPM and 21" MP) before it got rough. Peak range was 11.7 - 12.1 GPH.
 
Well, there is smooth, and then there is more power. A strong set of magnetos can go LOP and actually light the mixture, but you need advanced timing to get the much slower burning flame front to deliver peak cylinder pressure at the right time. This is graphically demonstrated with the advance switch function of the CPI ignition. Flipping that switch once LOP gives me 3 more knots. No other change, just a slight ignition advance. Also keep in mind that this is just an additional advance that is already in place due to the EI's map... the difference would be even more dramatic if starting from a magneto's fixed timing. A mag may run LOP "fine", but there is no doubt that it is leaving a lot of performance on the table compared to a properly configured EI.
 
Last edited:
Article "Quotes"

No where in the article can I find where the quotes below regarding potential for valve recession are actually attributed to George Braly. I think the GAMI folks have always stated that lead in fuel provides no benefit to valves other than by increasing detonation margin.

Skylor


All hail Mr. Braly! He is brilliant for sure.

He wrote, "Unfortunately, if some guy buys an engine that has been run on a dyno with 100LL and then gets a little 100LL during the life of the engine run mainly on unleaded mogas, and then goes to full TBO, he can write an article on the internet that claims valve recession is a myth, and people believe him.", in his article that Bill linked us to.

...

GAMI seems to be backtracking on what they said in 2014 with their G100UL fuel being run in Embry-Riddle training 172s:

"No issues were found," Roehl reports. "We just lack a couple of other tests to complete our first STC on the 172."

Now they say valve recession may be an issue. They've been running tests on this fuel since 2010 I believe. They need to update some of their website info.


Significantly, GAMI's G100UL fuel is not part of the FAA PAFI program where Swift's fuel is. Swift 94UL is on the market now and their UL102 is being testing under PAFI. Sour grapes?
 
Last edited:
Thanks for your reply Dan. I guess I was concerned about the rather slow and gradual leaning process that I have had to use define peak. This left EGT in the red box area for a while every time I flew. This was the reason for my concern as well as the high temps associated with peak EGT ops. I'm pleased that you think that's a safe approach. Much of the time years ago I had only a single probe EGT so I have no idea what the Gami spread was. Your response is appreciated...
 
Thanks for your reply Dan. I guess I was concerned about the rather slow and gradual leaning process that I have had to use define peak. This left EGT in the red box area for a while every time I flew. This was the reason for my concern as well as the high temps associated with peak EGT ops. I'm pleased that you think that's a safe approach. Much of the time years ago I had only a single probe EGT so I have no idea what the Gami spread was. Your response is appreciated...

A slow and gradual pull from rich to peak isn't necessary. Instead, do a three-second pull from rich to LOP...until it stumbles, or you know from fuel flow that it's quite lean. Then sneak up on peak from the lean side.
 
Leaner mixtures do require more spark advance. So even a balanced engine (meaning narrow GAMi spread, not physical engine balance) will run rough just a little bit lean of peak on mags. Run the timing up, and smooth as glass.

I fully agree that leaner mixtures require more advance for optimum power production. However, my experiences don't match yours on the roughness issue. I have one EI and one mag and my GAMI spread is .1 GPH.

The other day I had a problem with my EI and needed to get a flight done for a trip. I pulled the EI and put the old mag back in for the trip. In cruise I run 30-40* LOP at all times. I noticed no additional roughness with the mag. My speed was down, but it was no rougher.

Larry
 
Thanks Dan. Will try this - never had courage to do it before ("big pull"). My original question still stands. That is, since I spent a lot of time leaning very slowly in the Redbox area, why did my engines do so well?
 
Because the "red box" does not mean "detonation is assured". All those other times that you were lollygagging in the danger zone of the red box you were probably NOT with the added pro detonation factors of elevated CHT, high engine load, bad fuel, etc.

Also keep in mind that the red box scenario is most critical in turbo engines, which really is the crowd that Mr. Busch is talking to. We can and do detonate the NA Lycomings, but it's relatively rare if it happens under normal temps/loads.
 
Back
Top