03-18-2011, 06:34 PM #1
Cobb AP - Stage 1 - 93 Octane on 335 xi
I drive a 2009 335xi as a daily driver and occasional track day ride when my track car is not available. My background is in mechanical engineering, instrumentation, and control systems. Driving style ranges from safe, but assertive on the street to smooth, but all-out on the track. With that background, I'll offer some initial impressions on my first evening with the Cobb AP - Stage 1, 93 octane - no bolt-ons.
As expected, torque and power are substantially increased from the OE calibration! The car feels much more eager and willing to rev. I found myself glancing at the speedo and making a mental note to be careful where I open the throttle, as the car builds speed much more rapidly than it did before!
In stop and go traffic, and at WOT the Cobb calibration (downloaded Mar 16th) is very driveable, but at partial throttle roll on I experienced an interesting response. In a lowish gear at modest rpm, say 2200 in 2nd gear, if I roll on about 1/3 - 1/2 throttle I feel a substantial surge of torque and the resulting accelleration, followed by a quick drop off of the torque curve. It's almost as if the throttle is commanding a speed setpoint. (I know it's not, but it gave that impression initially.) I took a peek at the live data stream while this was going on and it appears that my partial throttle application is resulting in a fairly rapid ramp of req'd TP to 81% that holds until the turbos spool and boost hits something close to peak. Then the TPS reports a rapid closing to about 30-40%.
The data matches up well with what my 'butt dyno' is saying and I can imagine the logic of opening the throttle to get the turbos spooled, but the end result is way more torque than I want with a partial throttle opening and an abrupt change in torque delivery for a constant throttle position.
Personnal, I would value a more linear throttle response. Commanding WOT, even briefly, in response to 1/3 to 1/2 throttle pedal deflection is disconcerting. It makes the car feel 'fast', 'strong', 'responsive', (which is why I suppose the OE calibration is like this) but in my assessment, it also feels 'twitchy', 'jumpy', 'difficult to drive smoothly'.
For what it's worth, if I want the throttle to open all the way, I'll push the pedal to the floor.
I thought I'd try some other throttle positions and saw similar responses, even rolling on to WOT! (The TPS indicated a fairly quick - 2 sec? - ramp back up to maximum after the initial closing event.) I asked Cobb if 81% is really WOT and they confirm that it is. Also puzzling was an indicated TPS reduction to 40-45% after turbos spool and boost comes on while my right foot is commanding WOT.
They report that this a BMW OEM boost management technique and that they will continue to use this strategy as well. I suspect that closing the throttle is a response to managing boost in excess of what's being commanded. I read somewhere else that people were speculating that partial throttle closure was a BMW 'safety' strategy to reduce boost when it exceeds a threshold, but that discussion was at pressures around 17 psi. I'm seeing the same thing around 13.5-14 psi, so I doubt it's a hard limit.
I find it very hard to believe that anyone would close the throttle even a bit when the driver is commanding WOT. I'm just an unfrozen caveman engineer, and I don't know much, but what I do know is that the pressure drop across a partically closed throttle body results in efficiency losses that are not well aligned with my intentions when I push my right foot to the floor.
Boost overshoot and the resulting throttle closure could probably be avoided if wastegates could start opening a little earlier as boost approaches the commanded value.
Based on log data posted on Cobb's site, I was expecting to see maximum steady state boost of around 17 psi and fairly rich AFR, but steady state boost did not exceed 13.5 psi or so and AFR's didn't look as rich. I've asked if the current calibration a little more conservative than the initial release, but have not heard back on this topic.
No logs today, but I can post a record of what I've observed when I can make some time.
Overall, I'm happy with my experience and look forward to the release of tuning software that will let me customize my throttle map.
03-18-2011, 07:30 PM #2Timeout
- Join Date
- Feb 2010
- Rep Points
- 318 Post(s)
- Rep Power
yea, throttle closure is common under certain circumstances, the how and why is better left to the tuners here.
but an FYI, your not going to get 17psi from COBB, at least not yet. That is way past the efficeny range of the OEM turbo's, a lot of the piggyback guys do it because they want more and have meth to supplement with the cooling plus octane
actually, im not even sure if the flash guys have been able to get 17 psi, maybe GIAC, but vaguely recall that being a long battle.
best of luck to you, and you seem well informed of whats going on with the car/human connection
03-18-2011, 07:32 PM #3Member
- Join Date
- Apr 2010
- Rep Points
- 6 Post(s)
- Rep Power
do they allow for a custome tune with the access port if you go to a shop that purchased the rights from cobb??
03-18-2011, 08:01 PM #4Stage 2 or 2.5 E9X M3 S65 V8 supercharger kit for sale: http://www.boostaddict.com/showthrea...r-kit-for-sale
03-18-2011, 08:09 PM #5Stage 2 or 2.5 E9X M3 S65 V8 supercharger kit for sale: http://www.boostaddict.com/showthrea...r-kit-for-sale
03-18-2011, 11:47 PM #6
07-10-2011, 10:43 PM #7
Throttle mapping made it onto the radar!!!
Installed the V1 (Linear) map and the V2 (more aggressive onset) and went out to play.
Version 1 Throttle Map - Thank you Cobb for turning my right pedal from an on/off switch to a throttle! The car feels less powerful initially until you realize you need to push the pedal much further to get the same torque. I've read reviews of the Dinan flash where owners have lamented the lack of low end torque, but accepted that 'loss' in exchange for the midrange reward. I will bet that if they pushed the pedal to the floor they'd find the torque is available from the same rpms, but requires more right foot action. This will be fantastic on the track!
Initial pedal travel results in linear TP indicated on the AP, but around 60-80% pedal travel causes a very quick transition on indicated TP and a corresponding surge of torque. This coincides with changing engine rpm and load, so there may be more at play than just relationship between pedal and commanded TP. Also of interest is that 'TP act.' (does this mean 'actual'? 'Actuator'? is this parameter the best indication of what the throttle butterfly is doing?) reads a maximum of 81% (which I understand is the max it ever achieves) at something less that pedal on the floor, but if you actually do floor it, more torque is available. Functionally, this is exactly what I want. It's just curious that torque doesn't seem to track TP at the top end.
Version 2 throttle map is more lively than the pure linear map, but does a toned down version of what the OEM thottle map does. Throttle adjustments still result in overshoot which is less severe than OE, but still a bit annoying.
Linear for me, for now.
07-10-2011, 10:48 PM #8
07-11-2011, 09:48 AM #9
The 'linear' map feels very soft near the top of the pedal travel. I'd prefer a slightly more aggressive throttle onset, but not the overshoot associated with version 2.
I'd also prefer not to have the sudden increase in TP in the 60-80% range, but HEY, it's much much better.
09-08-2011, 11:53 PM #10
I can't rep you for this post (still too new) but I found the exact same response for the 535 on stage 1 oct 93 v.203. I re-flashed the alternative throttle map 1 and just got back from a short, non-spirited drive and it's much more behaved. I'll have to pay more attention as i'm driving around tomorrow but just wanted to bump your post back up as it is indeed right on the money! Thanks for taking the time to write it up!
09-09-2011, 02:38 AM #11Member
- Join Date
- Nov 2010
- Rep Points
- 52 Post(s)
- Rep Power
FYI: the throttle closure during WOT is simply the ECU's inability to control the wastegates properly, so it overshoots load target. Cobb should come out with different WGDC maps for loose, tight, medium wastegate stiffness.