Technical Ignition advance on E85
- Thread starter veer90
- Start date
-
- Tags
- e85 ignition advance timing tuning
good starting point is 8 degrees @ peak torque -> 11 deg @ peak power and adjust from there once you're on a dyno
Something I would like to point out to people on here. Commanded timing and actual timing are two different things. A lot of people have basic knowledge of the tuning tables and will be like OMG so and so is running 18 degrees of timing at X RPM. Two things that aren't being factored in are:
1) Are they actually hitting that part of the table? I could put 50* of timing at 8k RPM, probably won't be able to get to those cells though, right?
2) The knock sensors are pulling back timing. Say you are commanding 18 degrees but the knock sensors are pulling 10 degrees of it away. You are actually running 8* of timing, not the 18* you are telling people.
What's scary is that most of the people you guys are paying for tunes like to go overly aggressive on the timing and count on the knock sensors to pull it back in situations where it's right on the edge. This is super ghetto to me but is borderline acceptable since our ECU can respond so quickly.
When I tune people (LS world, not BMWs) I tell them that they should look at E85 this way: E85 doesn't let you add more timing, it allows you to takeaway less
1) Are they actually hitting that part of the table? I could put 50* of timing at 8k RPM, probably won't be able to get to those cells though, right?
2) The knock sensors are pulling back timing. Say you are commanding 18 degrees but the knock sensors are pulling 10 degrees of it away. You are actually running 8* of timing, not the 18* you are telling people.
What's scary is that most of the people you guys are paying for tunes like to go overly aggressive on the timing and count on the knock sensors to pull it back in situations where it's right on the edge. This is super ghetto to me but is borderline acceptable since our ECU can respond so quickly.
When I tune people (LS world, not BMWs) I tell them that they should look at E85 this way: E85 doesn't let you add more timing, it allows you to takeaway less
i have said this TIME and TIME again. that knock sensor adjustments are not meant to be done with blanket changes. But for some reason some vendors said publically its fine. lol. No it's not.
i say this too, along with the gaps people run. i run 18* timing on one of my other cars making 800whp, lol. And the plug gap is close to .028-.030
This is a bit lengthy but it should be known that ethanol has a faster flame speed than gasoline. There's a common misconception that you just add more timing with ethanol blends for more power, but if you understand what knock/pre-ignition is then it isn't hard to understand why that's flawed. If you classify an engine as being knock limited that would indicate that BEFORE we hit MBT timing our knock feedback loop started retarding the ignition timing to prevent or respond to knock. Like wise a non-knock limited engine would indicate we hit MBT timing before the knock feedback loop retarded the ignition timing. Now you need to understand what causes pre-ignition? Heat does, after all pre-ignition is the fuel igniting before the spark plug has fired. Ethanol's higher octane and great cooling abilities are able to both withstand higher temperatures before pre-igniting and also remove A LOT of temperature from the combustion process.
If we had an engine that was knock limited on gasoline, you can now easily see why this misconception of adding more timing for ethanol has come about. In this scenario the excellent knock resistant characteristics of ethanol may allow you to run more ignition timing and make more torque.
Conversely if we had an engine that was NOT knock limited (tuned to MBT on gasoline) you would actually want to remove timing as the faster flame speed of ethanol in-conjunction with the gasoline ignition timing would mean you're now past MBT timing. Going past MBT timing gains you nothing in torque (you'll actually start losing torque) and increases the loading on the rotating assembly.
Every engine responds a little bit differently but knowing the flame speed is faster is very important and this is exactly why you cannot and should not tune ignition timing without a dyno. Just because you're not seeing ignition timing pulled doesn't mean you should increase ignition timing, you may very well be already past MBT timing.
If we had an engine that was knock limited on gasoline, you can now easily see why this misconception of adding more timing for ethanol has come about. In this scenario the excellent knock resistant characteristics of ethanol may allow you to run more ignition timing and make more torque.
Conversely if we had an engine that was NOT knock limited (tuned to MBT on gasoline) you would actually want to remove timing as the faster flame speed of ethanol in-conjunction with the gasoline ignition timing would mean you're now past MBT timing. Going past MBT timing gains you nothing in torque (you'll actually start losing torque) and increases the loading on the rotating assembly.
Every engine responds a little bit differently but knowing the flame speed is faster is very important and this is exactly why you cannot and should not tune ignition timing without a dyno. Just because you're not seeing ignition timing pulled doesn't mean you should increase ignition timing, you may very well be already past MBT timing.
Dude, just put 30* across the board on e85. It's e85 brah, you won't detonate! Trust me, I'm a professional tuna...
I'm a TUNA!
I'm a TUNA!
bahn makes a valiant attempt to correct misconceptions, but I believe still hangs on to one "unicorn tears" property of ethanol.
The autoignition temperature reduces with increasing ethanol content. Most don't see this because of the extra cooling. However, if you push it too far with running lean, high boost, even without knock, you can blow impressive holes in things. Worthwhile having an EGT probe and aiming for a little lower than gasoline perhaps.
https://www.researchgate.net/public...of_Ethanol_and_E85_in_a_Spark_Ignition_Engine source of the above but only abstract available without paywall. E10 97 RON better than ethanol, better than 95RON.
https://www.sciencedirect.com/science/article/pii/S0360128516300764
On my travels around the literature, I also noticed that MON is misleading garbage for modern fuels in a turbocharged engine (where knock resistance has a positive correlation with sensitivity which is RON-MON). RON is better but still problematic.
The autoignition temperature reduces with increasing ethanol content. Most don't see this because of the extra cooling. However, if you push it too far with running lean, high boost, even without knock, you can blow impressive holes in things. Worthwhile having an EGT probe and aiming for a little lower than gasoline perhaps.
https://www.researchgate.net/public...of_Ethanol_and_E85_in_a_Spark_Ignition_Engine source of the above but only abstract available without paywall. E10 97 RON better than ethanol, better than 95RON.
https://www.sciencedirect.com/science/article/pii/S0360128516300764
On my travels around the literature, I also noticed that MON is misleading garbage for modern fuels in a turbocharged engine (where knock resistance has a positive correlation with sensitivity which is RON-MON). RON is better but still problematic.
Last edited:
Pretty interesting, @LamboLover .
All of this also suggests the importance of VANOS tuning as the amount of recirculated exhaust gas in the charge prior to spark should also have an effect on MBT and knocking, etc. How much overlap do we want? @V8bait’s spreadsheet was/is very informative.
And I also wonder how/whether the VANOS tables should be interpolated for FlexFuel operation as well.
All of this also suggests the importance of VANOS tuning as the amount of recirculated exhaust gas in the charge prior to spark should also have an effect on MBT and knocking, etc. How much overlap do we want? @V8bait’s spreadsheet was/is very informative.
And I also wonder how/whether the VANOS tables should be interpolated for FlexFuel operation as well.
Cheers, thanks for the information @LamboLover. I'm at work right now so I don't have a ton of time to research but if I recall the auto-ignition temp of E85 is very close to 91 RM/2 octane gasoline. The really important thing to note is the knock resistance is higher, whether that be from an increase of auto-ignition or cooling from additional fuel mass & evaporation. Thanks for the reply, we can all learn something.
There is not a "crank sensor calibration" issue. I have experienced the same and verified spark timing on the Syvecs with rpm.
If you check timing on a stock dme (which most dont) you will see that what's in the log is not what is actually happening. BMW obviously didnt get it "wrong", but simply there are comp tables in the background effecting final timing.
Finding peak power via a dyno is save with or without knock sensors (especially fuels with a large knock to peak power threshold). Common sense and experience tuning are used to identify "things that just aren't right".
Do you have an example of it in a log not being actual? The logged timing value should be after all compensations. I've traced a majority of the ignition code and am fairly certain the current logged value is what is sent to the base hardware to be realized as an ign angle.
I cant really send you a log of my timing light. Pretty easy to see though. Just pull #1 plug, find true tdc, and draw yourself some timing marks that you can see with a light. With a 2nd person operating the car, compare actual timing with the live displayed timing at different loads and rpm.
Timing from the dme is all over the place (compared the syvecs of the same test). Timing needed to make the same power is also much great on a standalone. This, are you mentioned, could be from an incorrect crank trigger (and related latch points on the syvevs). However, a simply check with the timing light verifies the the values entered are correct and stay correct regardless of rpm (ruling out incorrect trigger edge).
My guess (and just a guess) is that there are background timing corrections related to the dme's attempt to achieve its requested load and torque.
Only a few different loads and rpm. All head there to see a value.
Its obvious it's wrong. The same car on a standalone takes at least 8* (ish) more to make the same power and I know 100% that the timing I'm seeing there matches a timing light.
It is also obvious just looking at the values. You really think these engines only take 5-7* at 20psi? They take like 14+ on pump gas.
Not put a timing light on myself. It would be surprising if BMW were off the mark if the scaling and offset for ignition timing were correct and the item being logged is actual or final ignition and supposed to be in degrees BTDC, stable at the time of testing and logged fast enough to assess. Lots of ifs in either the logging of it or the assessment of it.
I ran some simulations last year and had to make what I thought were unrealistic combustion assumptions to match the performance we see on our timing curves. now in hindsight it makes more sense that the timing wasn't realistic.
The shop where I dyno'd my car builds a lot of high HP big turbo 2JZs. The tuner also said when tuning for MBT on E85 he usually ends up around 16*. Kept telling me to send it when he saw my logged timing was at 12.8*, lol
I don't want to make any conclusions based off purely anecdotal evidence but it looks like there may be a difference between what's in the main timing table and actual timing depending on conditions. But it shouldn't matter for a good tuner - you should be able to find MBT whether it's at 5* or 30* commanded timing.
I don't want to make any conclusions based off purely anecdotal evidence but it looks like there may be a difference between what's in the main timing table and actual timing depending on conditions. But it shouldn't matter for a good tuner - you should be able to find MBT whether it's at 5* or 30* commanded timing.
Did you stop where it was at or continue until you found MBT? Andy said it was 8ish degrees off, so that would mean you were close to 20*, unless the variance narrows at higher RPMS.
I've always understood it that a more efficient head would take less advance rather than more to make the most power, as the piston can be closer to TDC before igniting. Has anyone looked at NA DMEs? Do they have the same "low timing syndrome"?
