|
|
|
|
|
|
BMW Garage | BMW Meets | Register | Search | Today's Posts | Mark Forums Read |
|
BMW 3-Series (E90 E92) Forum
>
Vishnu Tech: Throttle Delay Fix Part II (timing control)
|
|
09-17-2009, 06:49 PM | #1 |
1733
Rep 17,960
Posts
Drives: A Lot
Join Date: Sep 2006
Location: SF Bay, CA
iTrader: (0)
Garage List 2018 Ducati Panigal ... [0.00]
2016 Mazda CX5 [0.00] 2017 Aprilia Tuono ... [0.00] 2019 BMW M2 Competi ... [0.00] 2015 BMW M5 Competi ... [10.00] 2016 Ducati XDiavel S [0.00] 2016 AMG GT S [0.00] 2011 Ferrari 458 It ... [0.00] 2017 Charger Hellcat [0.00] 2015 KTM Super Duke ... [0.00] 2016 KTM RC390 [0.00] |
Vishnu Tech: Throttle Delay Fix Part II (timing control)
who's there?
jk... As promised, here is a datalog from last night's 120.3mph run at Sac raceway: Running 17-18psi of boost on straight 100oct fuel. Which is pretty aggressive. Usually when you run too aggressive of a map, you get evidence of knock retard in the midrange to upper rpm range. But one thing I noticed was seeing knock retard in the datalogs right after the shift (and once in the middle of 3rd great at time stamp ~31s). The green circles show the knock retard points. The 1st at 3rd knock retard point can be considered knock retard because actual ignition advance at that point is just 2-3 degrees. After that event, timing stays reasonably low with a very gentle upwards slope. Turns out the reason for the post shift knock retard was caused by a very short term timing spike (see in red circles). If you look at the boost pressure (in the manifold... not shown) during these timing spikes, you'd see 10-11psi. And those who have spent a fair amount of timing logging these engines knows that 10-20 degrees of ignition advance at over 10psi of boost at 4500-5000rpm is a recipe for knock. Even on 100oct. And sure enough, reality mirrored theory. With most of my run hampered by knock retard activity, it's still surprising that it ran as fast as it did. But the next question is: How do we fix it? Sure, we can probably run MS109 and cool the charge down with meth. But that doesn't solve the problem when we are running on pump gas (on a lower boost map, of course!). So the real solution would be to tune out the timing spike with the PROcede's timing control/CPS offsetting feature. See next post |
09-17-2009, 06:49 PM | #2 |
1733
Rep 17,960
Posts
Drives: A Lot
Join Date: Sep 2006
Location: SF Bay, CA
iTrader: (0)
Garage List 2018 Ducati Panigal ... [0.00]
2016 Mazda CX5 [0.00] 2017 Aprilia Tuono ... [0.00] 2019 BMW M2 Competi ... [0.00] 2015 BMW M5 Competi ... [10.00] 2016 Ducati XDiavel S [0.00] 2016 AMG GT S [0.00] 2011 Ferrari 458 It ... [0.00] 2017 Charger Hellcat [0.00] 2015 KTM Super Duke ... [0.00] 2016 KTM RC390 [0.00] |
Here's another datalog from earlier today demonstrating the fix we are developing. What we are now doing is triggered a short-term ~10 degree ignition retard (in addition to the 2-4 degrees we normall pull out) during the Throttle Delay Fix event. What this does is two fold: 1) It reduces the magnitude of the timing spike that occurs during the shift event. and 2) It induces a transient ignition retard at the beginning of the next gear where knock retard would otherwise occur due to the sudden increase in load (from clutch going in to out). This shift-triggered ignition retard decays very quickly. Over the course of a fraction of a second. Not only does it "soften" the upshift slightly, it also helps to eliminate the upshift knock and, as a result, ensures that the rest of the gear remains at full advance. You can see in the logs that the timing rebounds quickly after the shift and stays, on average, a lot higher. We are still working on the fine tuning of this feature. Still experimenting with different retard amounts and positive and negative decay rates. But the results are spectacular. More performance and more safety. Hopefully this gives readers an idea of just how in-depth and thoroughly developed this feature will be once it sees full release. And it should also give you an idea of why are are able to effectively and safelysupport over 10psi at the shift without the bypass valves discharging Shiv Last edited by OpenFlash; 09-17-2009 at 07:08 PM.. |
Appreciate
0
|
09-17-2009, 07:24 PM | #3 | |
Major
23
Rep 1,179
Posts |
Quote:
thazt gotts to be the shietz !!!! props |
|
Appreciate
0
|
09-17-2009, 08:21 PM | #6 |
Lieutenant Colonel
95
Rep 1,659
Posts |
i don't even drive a manual and this is still getting me excited hahaha
yet pisses me off at the same time that the advantage us auto owners had is melting away... |
Appreciate
0
|
09-17-2009, 09:08 PM | #7 |
ASR Turbos+LSD+Meth= :D
73
Rep 1,517
Posts
Drives: E92 335i
Join Date: May 2007
Location: Miami, FL
|
awesome work, love the update... cant wait till release
__________________
2007 335i Coupe Mods: Check the Garage |
Appreciate
0
|
09-17-2009, 09:27 PM | #8 |
Colonel
181
Rep 2,840
Posts |
OK....call me skeptical - but wouldn't some of that knock retard activity have been eliminated in the second log due to lower boost pressure ?
Looks like the second log is running a couple of psi less boost. |
Appreciate
0
|
09-17-2009, 09:33 PM | #9 |
Lieutenant Colonel
293
Rep 1,512
Posts
Drives: Ever changing fleet
Join Date: May 2007
Location: Alabama
|
Auto boys, we're coming for you Impressive work Shiv.
__________________
23 iX M60, 24 GT3 RS Weissach, 22 Rivian R1T, 23 RS3, 13 E92 M3 Competition: Akra Evo, KW V3, etc
|
Appreciate
0
|
09-17-2009, 09:38 PM | #10 |
1733
Rep 17,960
Posts
Drives: A Lot
Join Date: Sep 2006
Location: SF Bay, CA
iTrader: (0)
Garage List 2018 Ducati Panigal ... [0.00]
2016 Mazda CX5 [0.00] 2017 Aprilia Tuono ... [0.00] 2019 BMW M2 Competi ... [0.00] 2015 BMW M5 Competi ... [10.00] 2016 Ducati XDiavel S [0.00] 2016 AMG GT S [0.00] 2011 Ferrari 458 It ... [0.00] 2017 Charger Hellcat [0.00] 2015 KTM Super Duke ... [0.00] 2016 KTM RC390 [0.00] |
It's also running 91oct, not 100. I'll post up more logs tomorrow.
|
Appreciate
0
|
09-17-2009, 09:56 PM | #11 | |
Colonel
181
Rep 2,840
Posts |
Quote:
I guess if you ran the same logs with 100 octane and higher boost, it would be more conclusive and I bet that's what you are going to do! |
|
Appreciate
0
|
09-18-2009, 07:08 AM | #15 |
Major General
307
Rep 5,175
Posts |
What is interesting is that you are now holding boost during shifts to speed up the shift, which is causing KR events, which you are using about 8 degrees timing offset to fix, which is still adding MPH to the run.
Regardless of which tune you use, it is interesting to watch as this issue is attacked and how. |
Appreciate
0
|
09-18-2009, 07:32 AM | #16 |
Lieutenant Colonel
293
Rep 1,512
Posts
Drives: Ever changing fleet
Join Date: May 2007
Location: Alabama
|
Same thing had to be done when tuning Evos as well with no lift to shift. Log where the ECU is floating during the NLTS event and yank a large deal of timing from those cells to prevent knock shifting into the next gear.
__________________
23 iX M60, 24 GT3 RS Weissach, 22 Rivian R1T, 23 RS3, 13 E92 M3 Competition: Akra Evo, KW V3, etc
|
Appreciate
0
|
09-18-2009, 08:38 AM | #17 |
Captain
49
Rep 704
Posts |
can you attach these data logs for those of us at work that can't see images unless they're attached
__________________
|
Appreciate
0
|
Bookmarks |
Thread Tools | Search this Thread |
|
|