ignition timing and performance

B8 S4
#1

Thought I’d start a separate discussion on this, look forward to hearing your comments.

So to review (and most of you know), you get maximum power (force pushing down on the piston) when the combustion completes at a known point after the piston reaches top dead center (TDC). Because the combustion process takes time we have to spark the mixture some number of degrees before (BTDC) while the piston is still traveling up. The number of degrees to advance the spark is largely dependent on RPM because a much faster moving piston means you have less time for this combustion to complete.

Ignite it too early and you get unproductive downward force on the piston while it is still traveling up and knock. Maximum power is obtained by advancing the timing to just below this knock threshold. The timing value of this knock threshold at a given RPM on the S4 engine will vary based on load/boost (how much air and fuel is being combusted), temperature, and of course octane, because a higher fuel octane can cause the combustion to burn in a slower more controlled fashion and complete later, so it allows for more advance. This creates even more power.

Now we get to the timing maps - for a given RPM/load/IAT etc. the tune selects a map value based on an estimated knock threshold for the engine under all those conditions.

If the map is too ‘optimistic’ the knock sensors go off and the timing will retard to below this knock threshold. How many degrees of correction you got simply reflects how far off the map’s guess was. One thing I see people say is “my car pulled timing because of heatsoak”. Technically no - the timing map already knows you’re heatsoaked from your IAT and should have a lower map value for that. If it knock corrected another 8 degrees on top of that, it means it simply means the map was overaggressive for the conditions.

Now the best tuners intentionally put in timing values a little over the known knock threshold for the conditions, so that the tune can adjust to max power. That’s a good thing. I remember on my B5 S4 I had an AMS tune that didn’t knock at all, not 1 degree. Is that a great running car? No it was a slow POS. The map timing values were so low for the conditions it never even touched the knock threshold. I got a GIAC tune and that changed. 8)

#2

No.

If you car is getting heatsoaked, the map will likely advance the timing. You can’t ignore the load part of timing correction - I feel like a broken record now telling you that.

Example: When you are approaching 80°C IAT, you will have 17% less load than at 20°C. The car will be advancing timing - because it can handle it - the cylinder charge is less.

From what I can see/have been told, the timing correction/protection for temperature for our cars don’t occur until at least 80°C IAT (70°C on stage III it seems). Some have said they didn’t even see any timing correction up to 85°C IAT.

#3

Can you post the APR 93 logs you were talking about from the other thread. I am genuinely curious to see the logs that approach 28 degrees timing on 93 fuel, and I would like to see the IAT and IM pressure, or MAF, to get an idea of the load.

I’m not saying I don’t believe it, I’m just want to see the log with my own eyes.

And please specify the mods - are these even pulley cars?

#4

Jspazz, thanks for responding and I 100% agree you can’t ignore load, and I acknowledge it does generally go down as IAT goes up. I used 85 deg C heatsoak as an example of protection mode in the timing map which I think you basically acknowledged yourself can exist. There’s an example map below for another engine. My main point though is the map selects a calibrated value for the conditions, and excessive knock correction is not a symptom of the conditions, but rather a symptom of the map not being calibrated for them.

http://i57.tinypic.com/29zbkmh.jpg

#5

I used to advance my timing curve (on Mazdas) just to the point of knock threshold and then back it off half a degree, then run some logs with the car heat soaked and see what number of KR I would get at specific boost temperatures. Then use the table that would cut the requested timing versus boost temps, filling in the numbers with tha amount of knock I was seeing.

No knock, was tuned to the max.

Saying that a car should see some KR for maximum power couldn’t be any more wrong, the ecu over compensates so you see less overall timing as well as dumping fuel into the cylinders so you run rich.

#6

I think the operative thing here is to focus on things you can easily or cheaply control. The first one is fuel - if you’re at the track use pump 100 octane every time, even if it’s a partial tank and you’re got a 96 mix for your 93 octane tune. The second one is cooling - if you’re going to run the car on a road course for 20+ minutes rather than a couple of 1/4 mile pulls, invest in an APR CPS or a supercharger cooling loop system of your choice.

Has anyone tried putting dry ice on top of the super charger? They sell it in those 12" x 12" x 2" tiles which would sort of fit nicely. Not sure if that’s so extreme it would warp the air screws out of tolerance, but I think I saw some OG E55 AMG supercharged guys doing it.

#7

You are on a different platform and had the ability to tailor a tune exactly for your car. Theoretically if one could calibrate all their timing tables/values to be just below the knock threshold in every load/IAT/etc/etc variation then I agree that is ideal, but that is not realistic on a generic tune for thousands of cars pumping various brands of fuel/octane/etc. You’ll see the most powerful B8 tunes out there have some correction. People have even seen correction on the stock Audis FWIW. I made this comment last week on AZ:

the theoretical sweet spot is probably 2-3 degrees of correction across the board to assure you are making max power with minimal correction.

And Sean@APR said “I couldn’t have said it better myself”.

#8

Thanks westwest, BTW I am catching up to you in rating.

I would add to your comments that if you see knock correction, keep adding/mixing octane (E85 or race) and re-logging until you see no knock correction at all. At that point your tune is making maximum power. i say that to Drob in fact because he can likely get another tenth/1mph out of that tune if he eliminates his remaining correction.

#9

Fourth time I’m asking: Can you or beaverdr point me in the right direction to see those logs? I looked in the AZ thread that was linked and no logs showing timing and load like you two mentioned.

I really am curious. I just can’t see a 93 tune at a reasonable load specifying 28 degrees of advance. Maybe if it was a weak air charge, ie. Stock tune motors can go to the 30s no problem. But show me something that is actually true.

#10

He graphed the logs. Timing and airflow below.

http://img11.hostingpics.net/pics/196920timing.jpg

http://img11.hostingpics.net/pics/255152boostPSI.jpg

Moogas is running APR 93 Stg 2, he added a little 100 to help his heavy timing correction and relogged - although I could barely get anything due the low resolution he had 1 line with the following values:

RPM=6565, MAF=1292 kg/hr, and ign angle=24.7 degrees with still a tiny bit of correction. Moogas is going to see a little higher load but I suspect his ignition angle will be not far under this curve at all with enough octane.

#11

Fifth time: can you post the log these graphs were made from? virgin?

#12

Karma for you again. I too wondered where these magic logs were that these guys kept talking about. Seems that drobs post and logs would be the ones we should be looking at now right. What am I missing there. Drob just kinda posted logs that shows none of this is an issue.

Also why is quoting sean@apr an official fact of anything. Not to sure on the pecking order over at APR but I dont think sean tunes the cars.

seems like lots of speculation. From the talking points and the information given it seems like there is correction on a stock car so why are we talking about correction. Is is ok that the stock tune has correction but similar correction on the APR tune is now bad

#13

AZT you said you did a review of the CW vs APR. Can you cut and paste that here in a new thread. Maybe I have this wrong but it seems like something is wrong and we know Bvdr is a CW fan and that appears to be part of his bias

#14

He didn’t post the raw logs, he posted a visual representation of the log data. If you want the raw logs I’m sure he’d provide them if you ask.

Why don’t you analyze moogas’s logs? He’s a trusted AR member and seeing the same thing. His timing appeared to be approaching mid-upper 20s it seems (with 100) if he gets better logs.

If you guys plan to imply this Mat guy on AZ with his ‘graphs’ is a fabrication in the grand AZT-bhvrdr conspiracy about APR’s B8.5 timing, well… drat you meddling kids, and we almost got away with it! Next on our agenda was that GIAC runs too rich. I had 3 people lined up on AZ with fake graphs showing 0.65 lambdas and stories of black smoke engulfing Prius owners. Watch out REVO you’re next.

#15

No one is saying there is a conspiracy. What there is here is a rush to conclusions about “APR is overtimed” when the issues could be something else. You guys say the root cause is aggressive ignition timing, I showed you my car was doing fine with 93 octane on a stg 2 93 file.

It helps when the logs under consideration are being presented by the owner of the car, since they can try and log different parameters, under different conditions, make changes and over time update the status. This is pretty simple problem solving 101. Scouring the internet for data that you didn’t collect to back a case without that person providing the context is exactly why we don’t trust you guys and want to see the actual logs.

Mike@cw has always used other users logs to try to support his claims to CW being safe and running fine…“well other guys with APR stg 2 are having problems, so anything that seems wrong on my car is also fine”. That was the fundamental undercurrent to this all. Everything is being mapped down to a simple transitive property based on “timing” and “knock”. To support an agenda (which is my opinion and one seemingly shared by many here).

Here is the post http://audirevolution.net/forum/index.php?topic=3310.msg98928#msg98928

I know Jran mentioned he is pulling 8 degrees as well. I dont think his car is going to blow up over that but it would be nice to see that tightened up a bit.
[/quote]

#16

For me the biggest issue is that AZT and Beaverdiver have this conspiracy theory about drob

they asked him to log his car WITHOUT e85 influencing the results. So…after a couple of days of driving, running down his tank, drob put in a full tank of 93 octane detroit gasoline, and logged his car

Same nice results he saw at the track.

AZT and beaverdiver have just COMPLETELY ignored those results. Why? because they don’t fit their theory. If a car can fill up with 93 from a normal gas station in a bankrupt city, and run withonly a couple of degrees to 4 degrees pulled (same as the 8-1 93/E85 results), it blows their theory out of the water.

How do they address that?

  1. they say they need to see more parameters
  2. they call drob a liar and say he added more E85 (or that he didn’t fill up and still has the same mix as Friday

This is why these guys are useless. It’s fun being right, but when you’re wrong, just say ‘ok let’s move on’.

#17

Well you never logged how high/aggressive your actual timing was now that you reduced the knock correction? I’d still like to see a log of that (in conjunction with the knock cylinder values) as i bet it will be quite impressive for Stage 2 pump cars.

But I’ll agree that one shouldn’t jump to conclusions about 1 person’s data just because it seems to fit a trend. I don’t think that was the intent, and then things got ugly, and I think each side just dug into their position. I say lets just go forward with open minds and analyze the data as we see it.

#18

Actually all I wanted to imply is that the timing you show in upper 28s on 93 tune - these guys are running the 100 tune, for whatever reason, thinking it’s 93. Not sure why or how.

I have compared what you sent on graphs in both these threads, and by all accounts it looks like 100 tune. On 93 tune stage II, there is no WAY the timing gets to the upper 28s unless there is a very weak air charge. Hence why I keep asking for the original logs.

I have many many logs from the last 3 years to compare to (but almost all stage II) on race, pump, APR, REVO, GIAC. I’d be interested to see what Tsivas thinks, he has a bigger log collection of different cars than me.

Give me any log like you show, I did just see one that I believe was moogas.

#19

J- upper 28s for timing advance on any pump File is not something I have seen in any stage 2 car I have logged. Something is not right there for sure.

Most of data I have for pump files has pump file timing just crack the 20 range. And we have probably the best testing conditions here.

25+ range is usually for any race file (100/104).

Lambda for a/f on most of these tunes @wot is in 0.75-0.8 range.

#20

When then maybe alot of APR 93 customers are getting 100 octane tunes by accident 8) If that is considered the 25+ range than Moogas is included (and I don’t know what it actually peaked at as a better resolution log is needed).

Also peak timing (redline) is not where the worst of the correction occurs in the data I’ve seen, but it is more the mid-range. I’m curious tsivas what typical timing do you see at 4500rpm for a 93 tune vs a 100 tune?

Anyone who wants to post some logs in here (of RPM, actual ignition angle, and knock correction in at least a few cylinders) is certainly welcome.