Ethanol or lower grade?

[Sonor]

So let's say you have a vehicle that wants 91 octane or higher while "high test" is 93 octane and your vehicle does not want 10% or higher ethanol. You pull into a station and notice the mid grade (89 octane) is without ethanol and ten cents cheaper than the 93 octane with 10% ethanol, which would you purchase?

[brewer90]

I would buy the real gas and some octane boost if the bike ran bad. I'd even buy the real gas if it was more expensive.

[Paul Mihalka]

Brewer90, welcome to this asylum. I get the impression you have firm opinions on things. I have firm opinions too, they may be different from yours, which opinions usually are. On my R1200GS I regularly run 87 octane, mostly 10% ethanol because that's what I find, and the bike is happy with it now with 160K miles.

[Ponch]

I had to put 87 in my bike on a trip last summer and it didn't run as well or get as good mileage. at the next fill, I was able to put 91 and the difference was readily apparent. I'd like to put in 93/94 to see how it runs, but 91 is the highest octane where I live.

[Dave_in_TX]

I've been running 89 octane with 10% ethanol for tens of thousands of miles in my 07 RT without problems.

[w2ge]

Paul, I'm really surprised that it runs okay for you on 87 octane... The RT runs (as you know) 12:1 compression (high). Is it possible your getting detonation and not hearing it? 160K miles is hard to argue with, but that IS high compression.

 

Does the RT have knock sensors, and then retards timing?

[dirtrider]

Does the RT have knock sensors, and then retards timing?

 

Morning w2ge

 

Yes, a knock sensor on each cylinder.

 

[Polo]

Just like Paul says, we all have at least one... opinion.

 

It may be related to temperatures, since MD is not as hot as TX, and knocking may happen at hotter temperatures more likely than at cooler ones, but I am yet to hear anything; maybe when I get my hearing aides I will change mu tune, but for now this is my jingle.

 

I usually fill up with premium which can be 91-93 and 10% ethanol which is the standard here.

 

I have filled up with 89 a few times when there was nothing else available in the summer, and in winter I regularly fill up with 89 or even 87 if it's my turn to buy breakfast.

 

I do not notice any adverse effects in performance, and I have pretty much given up worrying about mileage, but I think it's about the same.

[brewer90]

Hi Paul,

 

Thanks for the welcome. You are very perceptive

 

I am not a big fan of ethanol. My previous bike was carbed and I had a lot of issues if I let it sit plus there was a marked mpg difference. I've also had a lot of power tool issues due to ethanol. So I would pay more just to get "real" gas but that is definitely my opinion.

 

I'm only a month into ownership of my RT and all it has seen is premium with 10% ethanol so I have nothing to compare it too. Right now all is well.

[eddd]

I'm like Paul, pretty much run 87 with ethanol all the time and I'm at 168,000 miles on my RT. I've never noticed any decrease in performance.

[Selden]

Perhaps because of the recent repeal of the alcohol subsidy, alcohol-free gasoline seems to be getting easier to find. I have only 1 half tank experience with the stuff, and gas mileage may have been better, but 180 miles really isn't enough distance to reach a meaningful conclusion. Alcohol-free mileage should be better, but a lot of things that should be aren't in the real world.

[Ponch]

I don't know what to say. I could tell a difference in performance. It felt like the lower octane stuff chopped 20hp off the bike and when I filled up again with higher octane, it ran stronger. The bike was meant to run best on 98 RON, which is 93-94 MON. Considering it has a knock sensor, I would imagine retarding the timing could have an effect on performance.

[lkchris]

What's not debatable is if the knock sensors need to retard timing to compensate for knocking, the engine loses performance and efficiency.

 

It should also be noted that ethanol is an octane enhancer--octane meaning simply higher number = more resistance to knocking/pinging.

 

Independent of both is the fact ethanol contains less energy than gasoline and therefore fuel mileage suffers ... regardless of where the engine has self-tuned itself.

 

Seat-of-the-pants observations "I noticed" and "I didn't notice" are of little value.

[Dave_in_TX]

Paul, I'm really surprised that it runs okay for you on 87 octane... The RT runs (as you know) 12:1 compression (high). Is it possible your getting detonation and not hearing it? 160K miles is hard to argue with, but that IS high compression.

 

Does the RT have knock sensors, and then retards timing?

 

I'm not quite at Paul's mileage but my Rt has 134K on it.

[Buckster]

Phil, remember the 05-09 R1200GS (hexhead motor) runs 11.0:1 compression as compared to the RT with 12.0:1, as I read lower to withstand lower gas quality in less developed countries.

Still high compression but less prone to knocking...

Buck

[Peter Parts]

A lot of shouting in the dark.

 

1. Like a lot of things, subjective judgment can be good provided you do a good job of observing right. For example, gas from some filling stations might be poor even if it has the same nominal rating as other stations. Some riders never get the water out of their tanks. Some engines might have carbon or other issues that make them more sensitive to poor gas.

 

2. Some of the problems caused by inferior gas aren't easy to notice and/or happen only when you make extreme demands on the engine. So poor gas may be just fine for regular scoots and moderate rpms. Your bike can be pinging badly but might be hard to hear (wrong to think you can always hear it).

 

My machine runs honky-dory on so-so gas. But I like to lug and the higher the octane the lower the lug.

 

Ben

[Peter Parts]

What octane does is like changing your spark timing. Higher octane acts as if it retards firing a bit because it slows down the burn. That means a bit LESS power typically at all speeds. And, as in my previous post, lets you run the engine satisfactorily at lower rpm's (provided you don't get down into the damaging lugging speeds). No kidding.

 

The engine feels more "flexible" with higher octane. I like that feel - you can be lazier if you are just putting around town. So I use Sunoco 94 whenever I can find it. Feels good and costs just a bit more. For touring, middle-range octane is OK for me, if I want to save money.

 

So lower octane is like advancing the spark... but like with advancing the spark too much it can lead to detonation (AKA pinging) which is bad for your engine bearings and other parts.

 

I think you'll find this a helpful way to think about octane.

 

Can anybody say if that is the way to look at adding alcohol too? Or are inhibiting pinging and octane rating separate issues when alcohol is the vector?

 

Ben

[Ponch]

What octane does is like changing your spark timing. Higher octane acts as if it retards firing a bit because it slows down the burn. That means a bit LESS power typically at all speeds. And, as in my previous post, lets you run the engine satisfactorily at lower rpm's (provided you don't get down into the damaging lugging speeds). No kidding.

 

The engine feels more "flexible" with higher octane. I like that feel - you can be lazier if you are just putting around town. So I use Sunoco 94 whenever I can find it. Feels good and costs just a bit more. For touring, middle-range octane is OK for me, if I want to save money.

 

So lower octane is like advancing the spark... but like with advancing the spark too much it can lead to detonation (AKA pinging) which is bad for your engine bearings and other parts.

 

I think you'll find this a helpful way to think about octane.

 

Can anybody say if that is the way to look at adding alcohol too? Or are inhibiting pinging and octane rating separate issues when alcohol is the vector?

 

Ben

 

Yet higher octane allows more advance and lower octane less, in the context of the R1200 engine we are discussing. The result is more and less power respectively.

[Selden]

...Can anybody say if that is the way to look at adding alcohol too? Or are inhibiting pinging and octane rating separate issues when alcohol is the vector?...

Ethanol has a higher octane rating, but lower energy content, hence worse gas mileage when compared with a gasoline blended without ethanol. Methanol has an even higher octane rating (110), but is more corrosive, so it's used only in engines specifically designed for it.

[Mister Tee]

Here's what I notice - if I ran 87 octane, the knock sensors would kick in, and I'd take a hit on mileage but the bike would otherwise run fine. If I ran 89 octane, I would get the same mileage that I would if I ran 91, so presumably I didn't have a knocking problem with 89 so I ran 89.

[Peter Parts]

Ethanol has a higher octane rating, but lower energy content, hence worse gas mileage when compared with a gasoline blended without ethanol.

 

Selden - right. But does that mean ethanol does something to inhibit knocking or does it slow burning like refining higher octane into a gas? I've wondered a long time.

 

Ponch - Not quite right. We're not talking about somebody changing their spark timing according to their gas preference - although I would adjust my timing on an Airhead to taste from time to time! Not feasible with ECUs. Timing itself based on various factors, not just one dimension. The higher octane gas is LIKE changing timing because it slows burning.

 

Had a nice Maserati with two turbos and knock sensors. They are meant as safety limiters to protect an engine. Using your knock sensors as a routine way of controlling the engine while running poor gas is a bad and risky idea.

 

Ben

[Ponch]

Ethanol has a higher octane rating, but lower energy content, hence worse gas mileage when compared with a gasoline blended without ethanol.

 

Selden - right. But does that mean ethanol does something to inhibit knocking or does it slow burning like refining higher octane into a gas? I've wondered a long time.

 

Ponch - Not quite right. We're not talking about somebody changing their spark timing according to their gas preference - although I would adjust my timing on an Airhead to taste from time to time! Not feasible with ECUs. Timing itself based on various factors, not just one dimension. The higher octane gas is LIKE changing timing because it slows burning.

 

Had a nice Maserati with two turbos and knock sensors. They are meant as safety limiters to protect an engine. Using your knock sensors as a routine way of controlling the engine while running poor gas is a bad and risky idea.

 

Ben

 

The timing doesn't have to be changed at every tank. The computer does it if it detects knock. So given any fuel, if the computer uses a standard fuel/timing profile and it detects knock, it dials timing back so it no longer detects knock. In retarding the timing, the engine won't make as much power, but it will be able to run on the swill that was put in the tank. Again, we aren't talking about a 9.5:1 air cooled HD but an engine with a 12:1 compression ratio and I will still assert that my RT didn't have the same juice with 87 as it does with 91. That said, I wish I could put 94 in it to see how it runs.

[GRB60]

Higher octane is more resistant to detonation; not quite the same as slower burning. Because it's more resistant to detonation, it allows higher compression and/or more spark advance, which in turn will allow the engine to make more power.

[Selden]

Selden - right. But does that mean ethanol does something to inhibit knocking or does it slow burning like refining higher octane into a gas? I've wondered a long time.

I'm not a chemical engineer, but I'm not sure there is any meaningful real world difference. A fuel made from 100% iso-octane is the benchmark for a fuel's tendency to burn in a controlled manner, rather than exploding in an uncontrolled manner, under compression. Alcohol was used to raise the octane rating of motor fuels long before its use was mandated by the government.

 

Octane is not how fast a fuel burns

[Peter Parts]

 

The timing doesn't have to be changed at every tank. The computer does it if it detects knock.

 

Sorry to repeat myself, but the knock sensors are there for protecting the engine during the (hopefully brief) seconds when pinging occurs under conditions unsuitable for the gas, not for tuning.

 

Do you jump on your brakes full force and hope the ABS will steer you straight?

 

Wrong to attribute much smarts to the old Motronic or even successors with knock sensors.

 

Ben

[Peter Parts]

Selden - right. But does that mean ethanol does something to inhibit knocking or does it slow burning like refining higher octane into a gas? I've wondered a long time.

I'm not a chemical engineer, but I'm not sure there is any meaningful real world difference. A fuel made from 100% iso-octane is the benchmark for a fuel's tendency to burn in a controlled manner, rather than exploding in an uncontrolled manner, under compression. Alcohol was used to raise the octane rating of motor fuels long before its use was mandated by the government.

 

Octane is not how fast a fuel burns

 

Thanks.

 

If GRB60 is right, there just might not be a difference in practice, as you said, since you can advance the timing either way (because of slower burning or because of inherent knock resistance).

 

If the burn is slower, then you can lug more (which has nothing to do with knocking, I think). But if the ethnnol has anti-knock properties, it wouldn't help lugging.

 

But this is too deep for me to figure out.

 

Ben

[dirtrider]

 

The timing doesn't have to be changed at every tank. The computer does it if it detects knock.

 

Sorry to repeat myself, but the knock sensors are there for protecting the engine during the (hopefully brief) seconds when pinging occurs under conditions unsuitable for the gas, not for tuning.

 

Do you jump on your brakes full force and hope the ABS will steer you straight?

 

Wrong to attribute much smarts to the old Motronic or even successors with knock sensors.

 

Ben

 

Evening Ben

 

Actually the BMW 1200 hexhead is designed to operate on the knock sensors when needed.

 

From the riders manual-- under recommended fuel grade to use:

 

"Premium plus unleaded 98 ROZ/RON = (91 AKI)"

 

"Premium unleaded 95 ROZ/RON (fuel grade, usable with slight

power- and consumption related restrictions) 95 ROZ/RON = (89 AKI)"

 

Obviously the power reduction & increased consumption means retarded spark.

 

[Peter Parts]

Is that a quote from something possibly written and awkwardly translated by the Marketing Department?

 

Even if you accept that tenuous chain of reasoning that BMW is condoning (if not exactly recommending) piggy-backing on the safety sensor... I don't think it is a good idea any more than riding on any safety sensor is a good idea.

 

I offer this opinion based on general design principles and what I understood to be accepted practice before. Maybe the state of the art has advanced.

 

Ben

[mbelectric]

 

I offer this opinion based on general design principles and what I understood to be accepted practice before. Maybe the state of the art has advanced.

 

Ben

 

It has indeed changed my friend. On the race bike, tuning is more accomplished with a laptop and electronics than with the "old" screwdriver. Tuning to max efficiency is where it's at now. Not that I don't enjoy fiddling about with the DelOrto's on the 90S...sometimes.

 

Happy tuning.

 

MB>

[dirtrider]

Is that a quote from something possibly written and awkwardly translated by the Marketing Department?

 

Even if you accept that tenuous chain of reasoning that BMW is condoning (if not exactly recommending) piggy-backing on the safety sensor... I don't think it is a good idea any more than riding on any safety sensor is a good idea.

 

I offer this opinion based on general design principles and what I understood to be accepted practice before. Maybe the state of the art has advanced.

 

Ben

 

Morning Ben

 

Yes, that is from the riders manual, not quite exactly as printed as the riders manual has it printed in a column/box form.

 

Seeing as the original manual is written in German then translated to the other languages sometimes the translated versions use words that don't quite fit or sound a bit confusing. Plus Europe uses a different gasoline octane rating system.

 

Once a person gets used to reading translated German service & training manuals you kind of automatically skim over the poorly chosen words & look for the intended info.

 

Basically what that fuel usage is telling the U.S. rider in the usual confusing German way is that:

 

*Recommended gasoline is-- 98 RON-- or 91 AKI -- or on U.S. pump 91 (R+M)/2). (premium)

 

*CAN USE gasoline is-- 95 RON-- or 89 AKI -- or on U.S. pump 89 (R+M)/2). (mid grade)__ But using this gasoline will cause a slight MPG loss as well as a slight power loss.

 

To confuse it even more the conversion that the BMW riders manual uses is a bit strange as the actual German RON to U.S. /Canadian (R+M)/2) is closer to 4-5 octane numbers so in actuality the 98 RON should be 93 or 94 U.S. (R+M)/2) octane.

 

And the optional usage 95 German RON would equate closer to 90 91 U.S. (R+M)/2) octane.

 

Clear as mud Eh!

 

[Joe Frickin' Friday]

Higher octane acts as if it retards firing a bit because it slows down the burn. That means a bit LESS power typically at all speeds.

 

This is not correct.

 

There is no correlation between octane rating and laminar flame speed. As an example, here are four common motor fuels, their octane ratings, and their laminar flame speeds:

 

Propane: 105, 40 cm/s

Methanol: 99, 43 cm/s

Methane: 120, 38 cm/s

Iso-octane: 100, 34 cm/s

 

Higher octane fuels do not burn more slowly (or quickly) than lower octane fuels; they are simply better at resisting the transition from deflagration (a flame front traveling at the listed laminar flame speed) to detonation (a reaction triggered by, and reinforcing, a supersonic shock wave propagating through the mixture). Detonation is what high explosives do, and it is how you develop the localized extremely high temperatures and pressures that can wreck combustion chamber components and/or military targets.

 

Octane buys you time with which to burn the mixture via deflegration, which is what you want. Testing with laboratory rapid-compression devices shows that when you rapidly squeeze fuel-air mixtures up to a pressure/temperature high enough to cause autoignition (and then you hold the mixture at that final volume), there’s a short delay before combustion actually begins, and that delay is in fact longer if you’re using higher octane fuel. That delay is also shorter if you’re dealing with higher pressures/temperatures. So fuel with higher octane rating will better resist ignition via hot spots on the combustion chamber surface (pre-ignition), and it will also better resist detonation (knock).

 

If an engine is operating on a fixed ignition map (i.e. no knock-triggered spark retardation) without exhibiting knock, switching to a fuel with a higher octane rating will not reliably alter the engine’s performance. It may alter the engine’s performance if other properties of the fuel change. For example, if the octane rating is boosted by the addition of alcohol (as is often done), the energy content of the fuel will go down, and so will your fuel economy.

 

If a late-model engine designed around a high-octane fuel (e.g. the R1200 engine) is run with lower-octane fuel, the knock sensors will retard the spark timing, resulting in a loss of engine efficiency and reduced fuel economy.

 

The general advice is to use fuel with an octane rating that matches the vehicle manufacturer’s recommendation; running higher or lower octane is likely to result in lower fuel economy as described in the previous two paragraphs. (Note that at high altitudes, you can get away with slightly lower octane; the lower atmospheric pressure means that peak combustion temperatures/pressures will be lower reducing knock tendency. This is why gas stations in mountainous areas tend not to carry 93-octane gasoline.)

 

Can anybody say if that is the way to look at adding alcohol too? Or are inhibiting pinging and octane rating separate issues when alcohol is the vector?

 

Octane rating is a measure of knock resistance, regardless of fuel composition. Alcohols tend to have higher octane ratings than gasoline, and are often used to boost the octane rating of a gasoline blend.

 

[Ponch]

Is that a quote from something possibly written and awkwardly translated by the Marketing Department?

 

Even if you accept that tenuous chain of reasoning that BMW is condoning (if not exactly recommending) piggy-backing on the safety sensor... I don't think it is a good idea any more than riding on any safety sensor is a good idea.

 

I offer this opinion based on general design principles and what I understood to be accepted practice before. Maybe the state of the art has advanced.

 

Ben

 

Morning Ben

 

Yes, that is from the riders manual, not quite exactly as printed as the riders manual has it printed in a column/box form.

 

Seeing as the original manual is written in German then translated to the other languages sometimes the translated versions use words that don't quite fit or sound a bit confusing. Plus Europe uses a different gasoline octane rating system.

 

Once a person gets used to reading translated German service & training manuals you kind of automatically skim over the poorly chosen words & look for the intended info.

 

Basically what that fuel usage is telling the U.S. rider in the usual confusing German way is that:

 

*Recommended gasoline is-- 98 RON-- or 91 AKI -- or on U.S. pump 91 (R+M)/2). (premium)

 

*CAN USE gasoline is-- 95 RON-- or 89 AKI -- or on U.S. pump 89 (R+M)/2). (mid grade)__ But using this gasoline will cause a slight MPG loss as well as a slight power loss.

 

To confuse it even more the conversion that the BMW riders manual uses is a bit strange as the actual German RON to U.S. /Canadian (R+M)/2) is closer to 4-5 octane numbers so in actuality the 98 RON should be 93 or 94 U.S. (R+M)/2) octane.

 

And the optional usage 95 German RON would equate closer to 90 91 U.S. (R+M)/2) octane.

 

Clear as mud Eh!

 

If I could find 93/94 where I live, I'd give it a try, but 91 is the highest available with standard fuel. We have E85, but that verboten.

[lkchris]

Octane rating is a measure of knock resistance, regardless of fuel composition. Alcohols tend to have higher octane ratings than gasoline, and are often used to boost the octane rating of a gasoline blend.

 

A major point--beyond all the conspiracy stuff--in this seminal article on lead fuel additives. Quite a read.

 

The Secret History of Lead

[Peter Parts]

Well, great information from Joe Frickin' Friday. Worth a few careful readings. Many thanks.

 

But I am not sure if my rough characterization of octane in terms of burn speed really is as faulty as you say it is since you say, "there’s a short delay before combustion actually begins, and that delay is in fact longer if you’re using higher octane fuel..."

 

About BWM's manual and octane ratings, I think the dust is now settling: the knock sensor(s) are there to do their job if you have a tank of poor gas (but I'd still say don't "ride" the safety knock sensors). See note below.

 

And I do not appreciate being called a useless old tech fart working with screwdriver technology by mbelectric because I modestly admit to not knowing fine details of this year's crop of acoustic sensors (and lots more I don't know too). Nice of you to tell me those laptop-thingys are used for tuning.... I really have to tell my colleagues about that in the Computer Science Department where I was a professor.

 

Ben's tech note: knock sensors are microphones. Their signal gets processed by quite complicated sound-analysis software into a final "decision" about whether the sensor is "hearing" knocking or any of the many sounds engine make. Not an easy call to make and requires some number of repetitions of the knock (dozens, hundreds???), loud enough for the sensor, different enough from other sound wave-forms, and then the software goes to work to figure out if it really is knock. THEN the ECU has to react by shifting whatever it shifts when it feels like it. Finally, it takes a while for the cylinders to return to normal operating states. (mbelectric - did you understand all of that?)

 

Honestly now, does that sound like a kind of ECU sensor process you want to "ride" on or does it sound like an ultimate somewhat iffy engine protection on a bad day with poor gas once in a while?

 

Ben

 

[dirtrider]

Evening Ben

 

Most modern engine knock sensors are piezoelectric. They generate a voltage when a knock or vibration pressure excites them. Obviously they must be tuned to the correct frequency of the particular engine knock characteristics.

 

The auto companies have learned to tune them to recognize only combustion related knocking. Long gone are the days of spark retardation due to a noisy valve lifter or rod bearing thumping. In the old days of stand alone knock sensing systems you could thump the engine block with a long metal rod & with a timing light on the timing marks watch the spark instantly retard.

 

 

As for the response time, it is very fast in today's integrated electronic systems. Some of the automobiles I work with will imitate spark timing removal on the first sign of a combustion knock.

 

I'm sure there must be some other type of knock sensor systems either in production now or in the future. I'm only familiar with the systems I work with.

 

One of the novel ones I personally like is the Harley Davidson application. They place a high frequency across the spark plug gap (independent of the actual spark) then monitor that applied high frequency to detect the detrimental combustion knock as it disrupts the high frequency across the spark plug electrode gap. As far as I know that also requires using their specific spark plugs or at least plugs meeting their requirements.

 

[Peter Parts]

I think this little video Google brought up explains a lot about piezo-electric microphone sensors. All microphones are "tuned" to work in the frequency range of interest but my guess (and what I've read), that is not enough to clue the ECU at whether a click is knock or knot - some kind of sound analysis is still needed.

 

As in other debates in this forum, there are clever things that ECUs in cars do that the BMW ECUs do not do, esp. the older BMW ECUs. In as much as BMW/Bosch are very secretive... these debates will go on! We often come to a point of confusion when ECU learning or creative adaption (like for octane) is discussed.

 

 

The HD system is clever. I believe it detects ionization and stuff like that that is related to burning and detonation. Seems a whole lot better than the acoustic method.

 

Ben

[mbelectric]

And I do not appreciate being called a useless old tech fart working with screwdriver technology by mbelectric because I modestly admit to not knowing fine details of this year's crop of acoustic sensors (and lots more I don't know too). Nice of you to tell me those laptop-thingys are used for tuning.... I really have to tell my colleagues about that in the Computer Science Department where I was a professor.

 

(mbelectric - did you understand all of that?)

 

 

Ben

 

Whoa! Where'd we get off on the wrong foot? That's twice.

 

Don't know how to respond, or rather that I even should. Ouch!

 

But let's get to your reply and clear the air:

 

Never called you "useless" "techie" "old" or a "fart". You did that yourself. Don't read more into a reply (to your post) than you really should. No offense was intended. Just a comment on your statement "maybe the state of the art has advanced". Don't offer up a statement without expecting a comment.

 

Dirtriders last post contains some of what I was implying regarding the "tech" of modern day tuning and knock prevention.

Sorry it didn't come out as polished as his and was misinterpreted. My apologies. "Choose your words carefully" as in "user agreement" here....I'll need to try better.

 

In 40 years of playing with engines of all types, I've learned more in the last 15 years racing them than the previous 20 riding them. And my computer and engine control management systems have aided me. Still use a screwdriver though. And still ride an Airhead with no fuel injection or electronic engine management. I like both worlds, as I'm sure others do also.

 

Rubber side down.

 

And if ya want to talk further, let's take this PM.

 

MB>

 

 

 

 

[mbelectric]

And great thread folks, thanks to the OP.

 

As usual Joe FF does it up . Good job.

 

MB>

[Joe Frickin' Friday]

But I am not sure if my rough characterization of octane in terms of burn speed really is as faulty as you say it is since you say, "there's a short delay before combustion actually begins, and that delay is in fact longer if you're using higher octane fuel..."

 

This short delay is noticeable for reasonable temperatures and pressures, such as you might find when in contact with the hot surface of a combustion chamber at TDC. But that delay scales with the inverse of temperature: as things get hotter, the delay gets shorter. When you get to the temperatures encountered in the arc across a spark plug gap (several thousand degrees F), that delay is negligible - or more to the point, the difference in delay related to difference in octane number is negligible.

 

I think this little video Google brought up explains a lot about piezo-electric microphone sensors. All microphones are "tuned" to work in the frequency range of interest but my guess (and what I've read), that is not enough to clue the ECU at whether a click is knock or knot - some kind of sound analysis is still needed.

 

The "knock" we hear is the result of a shock wave (which is in turn the result of a detonation event) bouncing back and forth across the combustion chamber numerous times before it fades away. While it may sound like a "click" to our ears, it's actually more of a "chirp" with a frequency of a few thousand hertz; that frequency is a function of the time it takes for the shock wave to traverse the width of the combustion chamber. Early production knock-sensing technology was not so good - I've heard nightmare stories in the past of random things triggering the knock sensor, like a loose exhaust heat shield - but they've gotten a lot better at identifying exactly what knock sounds like, i.e. picking up the "signal" and rejecting the "noise." Part of it is better understanding, but the other part is better hardware: the ECU has far more computing power than it did 15-20 years ago, and can do a lot more real-time signal analysis. So if the knock sensor itself isn't tuned to detect the characteristic frequencies associated with knock, then the ECU will study the raw data itself and make a decision - before the next combustion cycle - about whether or not it needs to retard the spark.

 

Using the spark gap to detect knock is attractive for manufacturers for a few reasons:

 

• Yes, you've essentially got the sensor in the combustion chamber, as close as possible to the source, so it's more immune to extraneous noise. An ordinary piezo sensor can be set up to provide satisfactory results, but it takes a lot more R&D time on any given engine model to find the right place to put it on the engine block and identify what exactly to watch for in the signal.
   
  
• It's cheaper: no costly piezo sensor required, just a little more wiring and a couple of caps/diodes/resistors in a support circuit. see here for circuit diagram, combustion pressure traces, and explanation.
   
  
• It can also be used to detect misfires. This makes it a valuable tool for emissions control, which is attractive for car manufacturers. Cars are supposed to detect misfire events and throw an error code if they happen too often (I don't think bikes are required to do this yet).

[Selden]

Between you and Dirt Rider, this place can be very educational.

[Peter Parts]

+1

B.

[lkchris]

There is a pervasive human tendency to hear what we want to hear.

 

Especially if we hear something that sounds like it's cheap.

 

I'm pretty certain no automotive engineer anywhere thinks a 12-1 compression gasoline engine can run on low-octane fuel simply because of a knock sensor. That is, no engineer anywhere has designed one that works that way and no such sensor exists anywhere.

 

Let's just say sensors can affect minor--not major-changes.

 

Part of the rationalization behind "hearing what we want to hear" is denial, for example concluding the words in the owners manual are just suggestions and are not correct and are a conspiracy and on and on. Of course no matter which way so-called "marketing" people say it, if it disagrees with preconceived notions it's of course wrong. They just aren't as smart about their own creations as owners are.

 

Corrilary to this is the tyranny of culture.

 

In the USA our automotive tradition has been ever larger engines that do what they can do mostly because of their size as opposed to any particular efficiency. Run fine on rotgut fuel.

 

In Europe, where taxation is based on displacement and performance is actually legal, the "culture" is small, high-revving, efficient engines, rather dependent on high compression and high-octane/high quality fuels.

 

Applying 'merikun principles to something foreign is a bit like peeing into the wind, I'd say.

 

On another forum there's a guy that thinks a particular BMW design feature is "stupid" because special tools are required and "Craftsman" doesn't have them. This despite the fact that in German automotive "culture" there is zero/zip/nada engineering time spent on design for easy DIY. You try and find a "PepBoys" in Germany. There is no "Model T" in German history.

 

No, you buy German, you get German.

[dirtrider]

Afternoon Kent

 

If what you say is true:

"I'm pretty certain no automotive engineer anywhere thinks a 12-1 compression gasoline engine can run on low-octane fuel simply because of a knock sensor. That is, no engineer anywhere has designed one that works that way and no such sensor exists anywhere".

 

Then why are BMW's with those very same high compression engines sold to be used in countries with low grade gasoline. Seems BMW touts their BMW GS/GSA as a perfect around the world motorcycle even in parts of the world with non existent high test gasoline.

 

Actually the sensor does exist as there is one (actually 2) in every one of those 1200 hexhead high compression BMW's.

 

The sensor only sees the knock & doesn't have anything to do with "how much spark" is pulled out as the spark removal is the job of the fueling computer. It's no more difficult to have the computer remove 8° of total spark advance than it is to have it only remove 3° total advance. As long as the sensor is seeing & reporting the knock the fueling computer will remove more & more ign timing until the knock stops as long as the fueling computer has the latitude to do so. Obviously there are limits on how much spark can be removed & for how long due to engine heating & other issues.

 

Obviously it must work as I personally know a guy that has 2006 GS with well over 100K & all he has ever used is the cheapest regular gasoline he can find. And he is not a feather weight either. With his camping gear on & his butt in the saddle that combined must go well 450+ lbs.

 

 

High boost Turbocharged engines have been running on those knock sensors for years without problems.

 

[Paul Mihalka]

"Obviously it must work as I personally know a guy that has 2006 GS with well over 100K & all he has ever used is the cheapest regular gasoline he can find. And he is not a feather weight either. With his camping gear on & his butt in the saddle that combined must go well 450+ lbs."

 

D.R., we haven't met in person (I'd love to) but that almost could be me. '05 R1200GS, 160K miles. I'm not a lightweight but pack light - max. maybe 300lbs. All it's live gas 87, in Colorado etc. 85. Couple of times I filled it with premium and did not notice any difference. I rarely need full throttle, maybe passing on a two-lane, and absolutely no full throttle at low rpm. As you say, the bike is designed to run on any gas around the world. I would guess USA 87 is pretty good in that comparison.

 

[Joe Frickin' Friday]

Even if you accept that tenuous chain of reasoning that BMW is condoning (if not exactly recommending) piggy-backing on the safety sensor... I don't think it is a good idea any more than riding on any safety sensor is a good idea.

 

I offer this opinion based on general design principles and what I understood to be accepted practice before. Maybe the state of the art has advanced.

 

If the BMW hexhead owner's manual is not to be trusted, then consider these lines from the owner's manual of my car, a 2003 Nissan Maxima:

 

Use unleaded regular gasoline wiht an octane rating of at least 87 AKI (Anti-Knock Index) number (Research octane number 91).

 

For improved vehicle performance, NISSAN recommends the use of unleaded premium gasoline with an octane rating of at least 91 AKI number (Research octane number 96).

 

My car's engine is indeed equipped with a knock sensor (of the piezo type), and the manual pretty clearly permits drivers to rely on that sensor while using low-octane gasoline. Heck, low-octane is the first thing an owner sees when he turns to that page in the manual, and the rest is almost an afterthought: "use 87 octane gasoline. Oh, if you want better performance, I guess go ahead and buy the good stuff."

 

I don't see that it should be any different for BMW bikes, particularly given that the BMW manual says the same thing (if slightly less eloquently).

[GRB60]

I'm pretty sure that the GS uses lower compression because of the potential that only lower octane fuel is available. I think it's 11:1.

[dirtrider]

Morning Greg

 

I am no where near my shop so can't verify engine specs but didn't the 1200GS go to 12:1 compression somewhere around 2008?

 

The latest canhead 1200 GS's seem to show 12:1 compression.

 

[Peter Parts]

Parsing the exact meaning and the TRUE meaning of something from a vehicle manufacturer is tricky. Unless you are Kent and the text is from a German manufacturer in which case it must be pure and true and perfect.

 

In other words, silly to argue about it but esp. silly to make a "proof" on it. Nissan and BMW may have their own good or bad reasons for what they write and the message they mean to convey to various audiences (of which users are only one even if you think it is written just for little old you) and, unless you are Kent, German Engineering Purity is only one of those reasons.

 

No matter what BMW or Nissan says, it is plain wrong design or to drive your engine to the brink of destruction and expect the safety controls to save you over and over. Not Three Mile Island nuclear plant, you know.

 

Ben

 

 

[Witch_Doctor]

My understanding, and I could just be way off, but continued use of low octane fuel adds to carbon deposits more quickly, which leads to less efficiency. That being said, ethanol or not, if your not burning all the mixture, deposits are going to form. Its just a matter of how fast. With that being said, if you clean out a 100,000+ engine, your probably going to find it more effecient, while also finding hidden issue. I cannot speak to it on my beemer, because it hasn't hit 100,000 miles yet, but my 2006 Tacoma just hit 250,000 miles, and after doing a carbon cleanout I found a small leak in the lower seal which needed addressing.

[Peter Parts]

Here and in other threads, some people talk about the ECU "learning" stuff. True enough - for example the Motronic 2.4 needs to learn the highest and lowest voltages of the TPS and you "teach" it whenever power has been restored to the ECU after a break.

 

Lots of subtleties to learning. But ordinarily some kinds of criteria or guides are needed. Some scenarios are just wishful thinking. But in the case of "learning" that the octane is riskily low, that seems reasonable to me.

 

But then I wondered how does the ECU know when you've gone back to hi-test? Do you suppose it kind of sneaks an advanced spark in from time to time to see if knocking is still happening? Does it default to hi-test on every start-up (and then have to detect knocking every day or else)?

 

Sure, I can think of clever if complicated strategies like that. But anybody know the facts... as opposed to wishful thinking?

 

Ben