Everything was going so well and now this...
Thread Starter
AudiWorld Super User
Joined: Apr 2006
Posts: 2,962
Likes: 0
Everything was going so well and now this...
<center><img src="http://i132.photobucket.com/albums/q33/x1rider/logjune272007.jpg"></center><p>no CEL or DTC since my O2 spacer install. Today I tried a few WOT runs and noticed that my boost seemed to be falling quite quickly. I did two 3rd gear WOT runs, plotted the specified vs actual boost and came up with this graph. It seems like my actual boost follows the specified all the way up but then falls way below the curve. My first guess is defective N75 opening the wastegate too quickly. Any thoughts on this? Any input would be greatly appreciated, thanks.
Thread Starter
AudiWorld Super User
Joined: Apr 2006
Posts: 2,962
Likes: 0
sorry I forgot...
GIAC FXK03 with Genesis injectors
Awe-Tuning FMIC
APR TIP (I doubt it's collapsing but I did bang the N75 around quite a bit installing it, maybe it just gave up and died)
TT HFC
Awe-Tuning Cat-Back exhaust
EVOMS intake
Stratmosphere HXc DV
That's about it.
Maybe I'll do a boost leak test tomorrow. Last time I had a boost leak though, I couldn't boost more than 10 psi.
Awe-Tuning FMIC
APR TIP (I doubt it's collapsing but I did bang the N75 around quite a bit installing it, maybe it just gave up and died)
TT HFC
Awe-Tuning Cat-Back exhaust
EVOMS intake
Stratmosphere HXc DV
That's about it.
Maybe I'll do a boost leak test tomorrow. Last time I had a boost leak though, I couldn't boost more than 10 psi.
AudiWorld Super User
Joined: Oct 2005
Posts: 6,127
Likes: 3
The departure from specified, may be the ECU's active response to some other...
parameter that is outside optimum/expected normal values, as determined by the ECU, based on sensor inputs.
Try logging other mb values concurrently with mb 115, like O2 sensors, lambda, ign timing knock control retard, IAT, etc, to see if there is some other parameter that is deviating from normal along with the reduction in actual boost pressure. The ECU response could either be based on accurate sensor data or faulty sensor data. Or, it could be an ECU adaptation defect involving the ECU tuning re mappings.
BTW, if the N75 was defective, (opening early or stuck partially open,) then it is likely the achieved boost would not reach the specified values at any point. From the graph, I believe the boost control system (N75 and wastegate/actuator mechanism,) is OK, and is responding to the ECU accordingly.
Try logging other mb values concurrently with mb 115, like O2 sensors, lambda, ign timing knock control retard, IAT, etc, to see if there is some other parameter that is deviating from normal along with the reduction in actual boost pressure. The ECU response could either be based on accurate sensor data or faulty sensor data. Or, it could be an ECU adaptation defect involving the ECU tuning re mappings.
BTW, if the N75 was defective, (opening early or stuck partially open,) then it is likely the achieved boost would not reach the specified values at any point. From the graph, I believe the boost control system (N75 and wastegate/actuator mechanism,) is OK, and is responding to the ECU accordingly.
Thread Starter
AudiWorld Super User
Joined: Apr 2006
Posts: 2,962
Likes: 0
Thanks for the input, here are 2 more graphs...
Lambda (actual & specified).
<img src="http://i132.photobucket.com/albums/q33/x1rider/lambdagraph.jpg">
I looks like the curves are very similar except for a few deviation spikes (maybe when I was shifting). If you need the actual values of the logs, I can email them to you.
CF (cyl1 to cyl4).
<img src="http://i132.photobucket.com/albums/q33/x1rider/CFgraph.jpg">
The maximum CF is 6 (which I regularly get). I have heard rumors that the 91 gas here is not really 91, we are getting screwed on that. One of my friends has a Harley that kept on pinging when riding in town and as soon as he went down south and filled up there, the pinging magically stopped.
I'm looking to get some xylene at the local paint store and I'll try raising the octane level or our "91" and see what it does.
Do you think it's worth it to order a N75 and give it a shot anyway? My car has 100k miles, it wouldn't exactly be a luxury!!!
I'll do a boost leak test this afternoon and see what I come up with. Also I'll look through the logs and see if the boost drop corresponds with the higher CF's.
Thanks Diagnosticator1!
<img src="http://i132.photobucket.com/albums/q33/x1rider/lambdagraph.jpg">
I looks like the curves are very similar except for a few deviation spikes (maybe when I was shifting). If you need the actual values of the logs, I can email them to you.
CF (cyl1 to cyl4).
<img src="http://i132.photobucket.com/albums/q33/x1rider/CFgraph.jpg">
The maximum CF is 6 (which I regularly get). I have heard rumors that the 91 gas here is not really 91, we are getting screwed on that. One of my friends has a Harley that kept on pinging when riding in town and as soon as he went down south and filled up there, the pinging magically stopped.
I'm looking to get some xylene at the local paint store and I'll try raising the octane level or our "91" and see what it does.
Do you think it's worth it to order a N75 and give it a shot anyway? My car has 100k miles, it wouldn't exactly be a luxury!!!
I'll do a boost leak test this afternoon and see what I come up with. Also I'll look through the logs and see if the boost drop corresponds with the higher CF's.
Thanks Diagnosticator1!
AudiWorld Super User
Joined: Oct 2005
Posts: 6,127
Likes: 3
The key thing, is graphing concurrently to see what may be deviating at the same time..
and with a similar "slope" or rate of change.
It won't hurt to replace the N75, other than the cost of a replacement.
Also, interpreting the data would be easier if you use a graphing app with finer resolution.
The graphed spikes/points for the lambda, known as "data outliers", are common, and not usually significant in relation to the data "on curve", and based on the context of the data collected, is often omitted from the graph.
It won't hurt to replace the N75, other than the cost of a replacement.
Also, interpreting the data would be easier if you use a graphing app with finer resolution.
The graphed spikes/points for the lambda, known as "data outliers", are common, and not usually significant in relation to the data "on curve", and based on the context of the data collected, is often omitted from the graph.
Trending Topics
Thread Starter
AudiWorld Super User
Joined: Apr 2006
Posts: 2,962
Likes: 0
graph with everything...
<center><img src="http://i132.photobucket.com/albums/q33/x1rider/loggraph.jpg"></center><p>sorry, I'm just learning how to graph data, I'm not sure how to do it with finer resolution yet and I still have to figure out how to put RPM on the x-axis.
As you can see, the CF's and boost rise/fall at the same time, other than that it doesn't say much. The O2 sensor looks like it's OK because the lambda specified/actual are fairly close right?
I remember a few weeks ago, I did a few runs and was peaking at 19-20 psi and holding a solid 17-18 psi after the peak.
Right now I'm thinking either boost leak that seals ok until 18-19psi and then starts letting air out or a defective N75 dumping boost too soon. Do you think I'm on the right track?
Maybe I should log wastegate duty cycle?
Thanks.
As you can see, the CF's and boost rise/fall at the same time, other than that it doesn't say much. The O2 sensor looks like it's OK because the lambda specified/actual are fairly close right?
I remember a few weeks ago, I did a few runs and was peaking at 19-20 psi and holding a solid 17-18 psi after the peak.
Right now I'm thinking either boost leak that seals ok until 18-19psi and then starts letting air out or a defective N75 dumping boost too soon. Do you think I'm on the right track?
Maybe I should log wastegate duty cycle?
Thanks.
AudiWorld Stories
Bringing Audi to Life for Audi Fans
Top 10 Most Expensive Audis Ever Sold on Bring-A-Trailer
Brett Foote
10 Audi Features & Options We Miss the Most!
Joe Kucinski
Audi Recreates Crazy-Looking Speed Record Breaker From 1935
Verdad Gallardo
Coachbuilder Recreates the 1995 Audi TTS Concept
Verdad Gallardo
Every Audi V10 Car Ranked!
Joe Kucinski
9 Audi Designs That Aged Like Fine Wine
Verdad Gallardo
2026 Audi Q5 Buyer's Guide (Every Model Explained)
Brett Foote
Ranking Every Current Audi 'S' Model From Worst to First!
Joe Kucinski
10 Best Audis That are FINALLY LEGAL to Import to the U.S.!
Joe Kucinski
AudiWorld Super User
Joined: Oct 2005
Posts: 6,127
Likes: 3
Yeah, a graphing app that can "smooth" the data, would be helpful. ..
also, extending the scale range of the "X" axis, will help legibility. For example, specifying the divisions on the X axis as 100 rpm intervals instead of 1000 rpm intervals, will spread out the graph, aiding legibility. The apparent slope of the graphed data at any point is determined by the scale applied to the Y axis. Specifying smaller intervals on the vertical Y axis will help reveal more detail also. More advanced capability graphing apps can auto scale the graph axis data intervals for best legibility, according to the value ranges in data set being analyzed.
The coincident variation of CF and the
boost achieved, is an important clue to the situation occuring. The fact that the boost achieved, is being lowered, and not the specified boost, suggests that the variable factor involved, is not a standard value used in the torque demand calcs, but is instead, a result of a protective action by the ECU, or possibly caused by variation of a boost relevant parameter not directly monitored or controlled by the ECU.
Logging the N75 duty cycle along with mb 115, would be very useful, to determine what may be effecting the boost levels observed.
The coincident variation of CF and the
boost achieved, is an important clue to the situation occuring. The fact that the boost achieved, is being lowered, and not the specified boost, suggests that the variable factor involved, is not a standard value used in the torque demand calcs, but is instead, a result of a protective action by the ECU, or possibly caused by variation of a boost relevant parameter not directly monitored or controlled by the ECU.
Logging the N75 duty cycle along with mb 115, would be very useful, to determine what may be effecting the boost levels observed.
AudiWorld Super User
Joined: Oct 2005
Posts: 6,127
Likes: 3
It is important that the data points be closely matched time stamps,
simply overlaying graphs without the time stamps being coordinated, can distort the information shown, that may or may not be occuring together real time.
Data collected while logging concurrently two measuring blocks, will have the required time stamp coordination for real time comparisons.
Data collected while logging concurrently two measuring blocks, will have the required time stamp coordination for real time comparisons.