Can I use 100 octane?

[dcstep]

I really do think that OBD software will reveal the source of the hp capping and decline. As PG and others have indicated, the software mapping apparently has a maximum advance rather than looking at octane/knock and say, "Oh, there no knock so I can advance further, no knock I can advance further" it does that to a limit and just stops advancing, where we peak. As higher and higher octane delays the ignition the power actually starts dropping.

I got a minute to look at some black market ECU software tables, which imply just that, there are tables with advance, AFR characteristics, etc. and not formula that constantly calculate optimal conditions. This software was in German and I only got a minute to see it, but there were clearly layers and layers of tables (mapping). Tuning is done my nudging the tables around resetting targets. Hopefully my dyno tuner will gain full access to the software and I'll have a better understanding of what's going on.

Dave

[lucid]

Yes, the ECU cannot be simply messsing with timing indefinitely based on sensor data. At some point, that doesn't make sense because a trade-off will be reached. That's what I am saying as well. What I am also saying is that BMW should know when messing with the timing based on sensor info will actually reduce output at the different octane levels, and can manage that by at least keeping the power output steady. If that is not happening, it might be because of what Lemans_Blue_M3 suggested--that the ECU is not "tuned" to leverage 100 octane in that regard. The ECU does not actually know the octane level of the fuel in the tank. It could possibly know under what conditions what octane level fuels will exhibit what type of detonation characteristics, which could very well be determined through experimentation.

[dcstep]

Originally Posted by lucid

Yes, the ECU cannot be simply messsing with timing indefinitely based on sensor data. At some point, that doesn't make sense because a trade-off will be reached. That's what I am saying as well. What I am also saying is that BMW should know when messing with the timing based on sensor info will actually reduce output at the different octane levels, and can manage that by at least keeping the power output steady. If that is not happening, it might be because of what Lemans_Blue_M3 suggested--that the ECU is not "tuned" to leverage 100 octane in that regard. ...

I think, in line with LBM's thinking, that BMW has elected not deal with what happens when you put fuel in excess of 100-octane in the car. They're indifferent if it loses power after say 100-octane because that fuel is outside the specs which they designed the car around. If someone's going to race the car in a series that allows 105-octane, for instance, then they'll have to develop their own ECU mapping to take advantage of the fuel.

BTW, my experience at altitude, driving with 105-octane for over a week, showed that excess octane doesn't alter the driving characteristics noticeably to the driver, other than power is lost.

Dave

[lucid]

Originally Posted by dcstep

I think, in line with LBM's thinking, that BMW has elected not deal with what happens when you put fuel in excess of 100-octane in the car. They're indifferent if it loses power after say 100-octane because that fuel is outside the specs which they designed the car around.

Yep, agreed on the interpretation of LBM's statement.

[bobbyd1961]

where do you get 100 octane ? in jersey 93 is the highest i find

[lucid]

Originally Posted by PencilGeek

Bavarian Technic -- $287.
http://www.bavariantechnic.com/

Thanks, I'll check it out.

Originally Posted by PencilGeek

Why not?

As far as I can see, if that was the only piece of information the ECU relied on to ingite, that would lead to strange ignition scenarios. I can only asssume other variables such as piston position and emissions data are evaluated in conjunction with knock sensor information. Do you know what is actually being measured through the spark plugs in our engines?

[dcstep]

Originally Posted by bobbyd1961

where do you get 100 octane ? in jersey 93 is the highest i find

Sunoco GT100 Locations in NJ

Dave

[d_dot]

Well I filled up with 100 oct and it almost fealt like the car was faster but not what I expected. So when I got to a quarter tank of 100 I filled up with 93 and in about 15 20 minutes I was like holy shit batman. The car came alive. So I guess it can Adapt to 96. But if u were to get a custom tune for 100 oct I bet it wud b insane.

[e36jakeo]

Originally Posted by dcstep

Then why'd you throw that lame curve ball into the fray?? There's no need to repeat that worthless discussion.

Let's just keep the discussion to octane in the current M3. That's complicated enough.

Dave

My point was to justify that I don't think the M3 makes a full claimed 414 HP on 91 octane (based on PG's baseline 91 octane dyno) and that 93 octane is probably needed to achieve the full 414 HP. Thus just two points of OCTANE in the M3 add hp, and perhaps these two points are as important, or more important,than going from 93 to 100. Somewhere between 93 and 100, as others have suggested, you probably reach a point of diminshing HP returns.

BTW, I have done 1/4 mile test with an acceleromete-based g-meter with 91 octane and then, on the same stretch of road and on the same night with ~100 octane and the trap speeds (best indicator of HP) were not much, if any, different (<1 mph). That was enough for me to stop spending extra dough on 100 octane!

[dcstep]

Originally Posted by e36jakeo

...

BTW, I have done 1/4 mile test with an acceleromete-based g-meter with 91 octane and then, on the same stretch of road and on the same night with ~100 octane and the trap speeds (best indicator of HP) were not much, if any, different (<1 mph). That was enough for me to stop spending extra dough on 100 octane!

You're incredibly confident about your ability to launch the car consistantly, making a decision like that based on only two runs, but not trusting the dyno. How were your launchs? This is NOT the best indicator of hp because there are two many variables. A very experienced drag racer wouldn't even be able to judge anything without more runs.

No one's suggesting using 100-octane 100% of the time. Any real drag racer could put a 10+% increase in hp to use at the track and would gladly spend an extra 50-bucks or so for a succesful day at the track.

BTW, has anyone got the formula to calculate what would be gained in a theoretical 1/4-mile going from 333 hp to 369 hp with a 3700 lb car, all other things being equal? I think we're only talking about 2 or 3 mph at the traps. Some people don't realize that's a big deal and can mean the difference between winning and elimination and translates into car lengths.

Dave

[stylinexpat]

I tried a tank of 100 Octane fuel from Union 76 today. Will have to wait and see how my car likes that fuel. Put in $90 today as it was $7.95 per gallon at the Union 76 off the 10 Freeway in West Covina. Driving back home there was too much traffic and it was too soon to tell any difference.

[Strax]

Originally Posted by PencilGeek

This implies that the ECU has the ability to detect octane rating. How does it do this? Without some type of chemical analysis, I don't think it knows anything about octane rating...let alone compensates for small shifts in it.


This makes more sense to me than anything else. If somebody told me the ECU is tuned for 93 octane (as has been said many times in many threads), I'd say they that is not possible because the ECU doesn't know what octane it is running. But if somebody told me that the ECU can only adjust timing between -10/+4 degrees, and anything above 96 octane required more than -10/+4 (such as -16/+4), then as a firmware designer myself, that's a story I could actually believe.

Again, this assumes the ECU knows how to detect octane rating. I'm very doubtful that it does. But if it does, I'm willing to admit in advance that I was 100% wrong about it.

The DME does not recognize octane rating. It uses its sophisticated knock sensor system (which were introduced with the V10 M5/M6 engine) for this purpose.

My experience with the DME so far shows that it works in the way you describe above. It has a base map and it applies its limited range knock adaptation values on top of it.

And here is a quick description of the new (not acoustic) knock system from the European Car magazine article of the M5.

"After three years spent on research and development, BMW makes its first introduction of a combustion charge ion analysis system first to be implemented on the S85. Similar to Saab's system of the 90's and based on theories developed in the 30's, this ion analysis takes the place of a conventional knock sensor by monitoring the combustion event and combustions pressures for each cylinder. Using the spark plug as a positive pole and the cylinder as a ground, the ion current system measures the conductance of the air fuel charge throughout the combustion process. As combustion takes place and the charge chemistry changes, so does the electrical conductance of the contents of the cylinder. Based on the conductance as a function of crank angle, each individual combustion event is evaluated to better refine and control the sequential fuel and spark maps on a per cylinder basis. This is basically like a very smart version of an automotive oscilloscope. Unlike previous attempts of combustion ion charge analysis, the BMW system also discerns misfires either from inadequate fuel or spark, something neither Saab's system or a knock sensor can do."

[foosh]

Originally Posted by Strax

The DME does not recognize octane rating. It uses its sophisticated knock sensor system (which were introduced with the V10 M5/M6 engine) for this purpose.

My experience with the DME so far shows that it works in the way you describe above. It has a base map and it applies its limited range knock adaptation values on top of it.

And here is a quick description of the new (not acoustic) knock system from the European Car magazine article of the M5.

"After three years spent on research and development, BMW makes its first introduction of a combustion charge ion analysis system first to be implemented on the S85. Similar to Saab's system of the 90's and based on theories developed in the 30's, this ion analysis takes the place of a conventional knock sensor by monitoring the combustion event and combustions pressures for each cylinder. Using the spark plug as a positive pole and the cylinder as a ground, the ion current system measures the conductance of the air fuel charge throughout the combustion process. As combustion takes place and the charge chemistry changes, so does the electrical conductance of the contents of the cylinder. Based on the conductance as a function of crank angle, each individual combustion event is evaluated to better refine and control the sequential fuel and spark maps on a per cylinder basis. This is basically like a very smart version of an automotive oscilloscope. Unlike previous attempts of combustion ion charge analysis, the BMW system also discerns misfires either from inadequate fuel or spark, something neither Saab's system or a knock sensor can do."

Thanks, very interesting!

[Lemans_Blue_M]

Just about to leave for work now, so no time to explain this in greater detail until later this morning...

The stock engine management computer DOES recognize the fuel octane.(sort of)

Every fuel has certain chemical characteristics that can be calculated and measured in regards to it's ability to suppress knock and or pre-ignition.

That is why the engineers spend so much time dyno tuning the engine on 93 octane unleaded gasoline. They actually CALIBRATE the tuning to use a given fuel octane and then write in small allowances for variations of this ideal combustion scenario.

So the stock ECU can detect what 93(US)/98(RON) octane is, because all the tables and ignition timing software was written around this particular fuel octane. Now can it distinguish between 93 and 100 octane ratings...NO. It can adapt to the 100 octane fuel...but only to a preset limit. If you want that limit expanded, then an aftermarket tuner can manipulate the code to get around these road blocks. (if he has the proper training and code disassembly tools)

BMW engineers have samples of all the various fuels available throughout the world. They request lab quality samples from Shell, BP, ESSO, etc.

Since there is no way they can control the availability of a given fuel in any geographical area, they must make allowances for this disparity in the their tuning parameters. (to a point)

Nearly 100% of BMW's engineering resources are spent trying to prevent knock on the lower end of the octane spectrum.

So that is why they allow some margin of adaptation within the ECU programming to 'cover' these slight fuel variations. (to protect the engine from severe detonation and possible damage)

They are frankly unconcerned with giving the ECU the ability to scale up the tuning parameters to take full advantage of very high octane fuels as a means to extract more power out of a given powerplant.

The above post by Strax is correct, and the ion flow technology is able to precisely measure the combustion chamber mixture which makes it even harder to use a higher rated octane fuel. (to extract more power)

I'll give more details later, I'm running late...

[dcstep]

Yes Strax, thank you very much for that very informative quote. To PGs question, "Does the engine sense octane", in a way it does. That on-the-fly analysis of the charge will yield different answers at different octane.

I'd been thinking in terms of acoustic knock sensors and wondering how even the fastest computer could stay ahead. This answers that question.

Now, given that high level of sophistication, why is the range of adjustment so limited? Is it to hold down table size?? I find that surprising. I think that they just limit the table because if someone needs a table for 105 or 110-octane, then BMW assumes that they'll develop their own.

Dave

[ta152]

Perfect info - I have a 2009 M3 - same question, high octane engine, what would high octane gas do ... ?

The info provided in this thread made a lot a sense. I am a software guy - and motor head, the insight (Lemans_Blue_M & PencilGeek) into the ECU is great stuff. I've got a similar question to BMW, but I think I'll get the standard "owner's manual" answer.

Thanks again!

[stylinexpat]

My Car ran great with 100 Octane fuel in it

[dcstep]

Originally Posted by stylinexpat

My Car ran great with 100 Octane fuel in it

That's to be expected, if you read the thread and live reasonably near sea level. The question is still open, would you gain more hp with 97-octane?

Dave