[blindrooster]

Originally Posted by TheAxiom

At least message Bob and have him look it over, he is a AA dealer (and a tuner) and he might be able to give you some direction.

I've been working with Nic @ AA and he's been extremely helpful.

[Taskmaster]

Originally Posted by hassmaschine

I've said it before, but that's what the 328i fuel curve looks like. The stock 330i curve is about 13.5:1 until the last cell, where it drops to ~12:8. the stock 325i 14.7:1 until almost 5k before it drops to the 13's and tapers off to 12.35. The 328i curve is similar to the 325i.



Theoretically the DME just follows the AF targets. in reality, the cat overheat protection can add up to 30% more fuel. That's why it always drops as far as 10:1 at the end of the graph. My fuel curve is nearly stock except that last cell I leaned out (I hadn't learned about cat protection at the time, so it still is too rich).



Every time I see one of these wacky fuel curves, I wonder what the tune looks like. There's no reason to target 14.7:1 at full load except for emissions reasons. I can't imagine AA didn't even touch the fuel curve, but I've never seen it personally and it doesn't make sense otherwise.

BTW, I had been looking for this map for a while. Took me a while to realize what the german word for cat protection is.. it's not obvious!

I know, I've seen you post it before.

Which is why I was asking you, I don't see a reason why some tunes don't even touch the stock fueling curve, even with the 3 stage manifold. My OE tune was one of those instances, and I think Evolve's tune were a another situation. Same with the CAT Overheat...

[Taskmaster]

Originally Posted by blindrooster

I've been working with Nic @ AA and he's been extremely helpful.

Well, let us know what you find. It sounds (and looks) like there may be something else going on here.

[hassmaschine]

Originally Posted by TheAxiom

I know, I've seen you post it before.

Which is why I was asking you, I don't see a reason why some tunes don't even touch the stock fueling curve, even with the 3 stage manifold. My OE tune was one of those instances, and I think Evolve's tune were a another situation. Same with the CAT Overheat...

maybe they don't know where it is or why it's happening? I've learned quite a bit from the disassembly - I doubt AA is doing that.

there are 8,000 parameters on MSV70 and about 2,000 more on MSV80. it's not easy to sort out what all of them do or to even find them across different software versions.. plus half of them are in german and translations aren't obvious. I wouldn't be even a little bit shocked if none of them know the location of the map I posted (which is why I didn't include the name, sorry guys).

[Taskmaster]

Originally Posted by hassmaschine

maybe they don't know where it is or why it's happening? I've learned quite a bit from the disassembly - I doubt AA is doing that.

there are 8,000 parameters on MSV70 and about 2,000 more on MSV80. it's not easy to sort out what all of them do or to even find them across different software versions.. plus half of them are in german and translations aren't obvious. I wouldn't be even a little bit shocked if none of them know the location of the map I posted (which is why I didn't include the name, sorry guys).

Hass Pls. I needs the horse powers

[rjahl]

Originally Posted by hassmaschine

maybe they don't know where it is or why it's happening? I've learned quite a bit from the disassembly - I doubt AA is doing that.

there are 8,000 parameters on MSV70 and about 2,000 more on MSV80. it's not easy to sort out what all of them do or to even find them across different software versions.. plus half of them are in german and translations aren't obvious. I wouldn't be even a little bit shocked if none of them know the location of the map I posted (which is why I didn't include the name, sorry guys).

Have you noticed any lag between when you place the car at full throttle and when the engine is really working at full load? IE. I've seen a good lag between full demand load and full valve lift with Full load lambda. I don't have my logs with me, but my memory is telling me this lag can be 500 milliseconds or more. Could that be part of the low fueling in the early RPMs on the dyno?

[stevesingo]

Is it the case that the distance the Lambda sensors are placed from the exhaust poert is related to the DME monitoring cylinder to cylinder variations in Lambda and correcting individually?

In such a case the DME will have a model based on mass air intake, fuel, ignition angle, exhaust cam timing etc, where the DME will calculate a window the Lambda sensor is reading to the appropriate cylinder. This window will be a time (ms) from spark event or TDC.

If this is the case, then moving the Lambda sensor further away from the port will change the sensing window and therefore lead to a miscalculation of the fuel correction.

Nothing a good tuner couldn't overcome.

[hassmaschine]

I haven't noticed anything. There really shouldn't be any lag. There is also wall-film based fuel enrichment to deal with transients so you really shouldn't have a lag between extra air and matching fuel. Although I have wondered how long it takes valvetronic to go from idle to full lift - that could be what you are seeing?

Originally Posted by stevesingo

Is it the case that the distance the Lambda sensors are placed from the exhaust poert is related to the DME monitoring cylinder to cylinder variations in Lambda and correcting individually?

In such a case the DME will have a model based on mass air intake, fuel, ignition angle, exhaust cam timing etc, where the DME will calculate a window the Lambda sensor is reading to the appropriate cylinder. This window will be a time (ms) from spark event or TDC.

If this is the case, then moving the Lambda sensor further away from the port will change the sensing window and therefore lead to a miscalculation of the fuel correction.

Nothing a good tuner couldn't overcome.

that's exactly what I'm saying. I don't know if it looks at individual cylinders or not (it's certainly fast enough to). I should track down the map that controls it but I haven't the slightest memory of what it was called..

[rjahl]

Originally Posted by hassmaschine

I haven't noticed anything. There really shouldn't be any lag. There is also wall-film based fuel enrichment to deal with transients so you really shouldn't have a lag between extra air and matching fuel. Although I have wondered how long it takes valvetronic to go from idle to full lift - that could be what you are seeing?

Hass,

I hope you can see this clip, I had problems formatting this to fit.

At time stamp 5038 milliseconds I went full throttle STAT_BERECHNETE_LAST_WERT @2391 RPMs but I did not see full valve lift STAT_VT_EXCENTER_WERT until 5413 milliseconds. Lamba set point also does not match the tune until 5631 time stamp. I'm pretty sure I have other logs that show similar results.

[hassmaschine]

I think that makes sense - there's a VVL adjustment speed setting that is 16.7mm/s. 180 degrees is about 10mm, so if you divide 180 / 1.67 you get about 112 degrees/second. Yours moved about 80 degrees in .8 seconds. I have tried playing with that setting but I think it's more a limitation of the Valvetronic motor than just software.

There's no real way to tell what the lambda target is doing in transients to full load, and I expect it probably follows the actual VVL setting anyway. A "full load" state is basically anything above 92-96% (depends on RPM) which is easily reached when you put your pedal to the floor. It doesn't mean that the VVL is maxed out though.

[justpete]

Originally Posted by hassmaschine

I think that makes sense - there's a VVL adjustment speed setting that is 16.7mm/s. 180 degrees is about 10mm, so if you divide 180 / 1.67 you get about 112 degrees/second. Yours moved about 80 degrees in .8 seconds. I have tried playing with that setting but I think it's more a limitation of the Valvetronic motor than just software.

There's no real way to tell what the lambda target is doing in transients to full load, and I expect it probably follows the actual VVL setting anyway. A "full load" state is basically anything above 92-96% (depends on RPM) which is easily reached when you put your pedal to the floor. It doesn't mean that the VVL is maxed out though.

Beginning to sound like a race tuned N52 should lose both Valvetronic and Vanos...

[hassmaschine]

There's nothing to gain by ditching vanos - and a whole lot of torque curve to lose. You might gain a little throttle response without valvetronic but I think it's a wash since the throttle plate can be 100% open, and now you're adding that restriction - the transient time doesn't seem to affect it much IMO. the throttle response on my 330i is pretty sharp. Not Porsche GT3 RS sharp but good enough for a $12k sedan..

[rjahl]

Originally Posted by hassmaschine

There's nothing to gain by ditching vanos - and a whole lot of torque curve to lose. You might gain a little throttle response without valvetronic but I think it's a wash since the throttle plate can be 100% open, and now you're adding that restriction - the transient time doesn't seem to affect it much IMO. the throttle response on my 330i is pretty sharp. Not Porsche GT3 RS sharp but good enough for a $12k sedan..

I think I read somewhere that the MSV 80 variant N52k gets in upgraded Valvetronic motor. Could have been marketing BS

[rjahl]

Originally Posted by hassmaschine

I think that makes sense - there's a VVL adjustment speed setting that is 16.7mm/s. 180 degrees is about 10mm, so if you divide 180 / 1.67 you get about 112 degrees/second. Yours moved about 80 degrees in .8 seconds. I have tried playing with that setting but I think it's more a limitation of the Valvetronic motor than just software.

There's no real way to tell what the lambda target is doing in transients to full load, and I expect it probably follows the actual VVL setting anyway. A "full load" state is basically anything above 92-96% (depends on RPM) which is easily reached when you put your pedal to the floor. It doesn't mean that the VVL is maxed out though.

You knowledge of this DME is amazing. I need to look at other logs and see if I can find something that runs a larger load spread.

[rjahl]

I looked at a few more logs and the 112 degrees per second rule seems to pan out. Lambda sees to lag a little behind the VVT but not much.

Back to the topic, Being completely ignorant of how a car is run on a dyno, could any of this lag be what we are seeing in the lower RPM Lambda anomalies?

If a the VVT cam / essentric cam is at 40-60 degrees when the driver nails the throttle, it will be almost a full second or more before full VVT can be reached, then another lag before full load lamda is operational. How many seconds does the dyno allow between 2,500 RPM and 7,000 RPM? Are they all the same?

[hassmaschine]

Originally Posted by rjahl

I think I read somewhere that the MSV 80 variant N52k gets in upgraded Valvetronic motor. Could have been marketing BS

I dont think that is true. Valvetronic 3 did get a faster motor but that started on the N55.

[E46driver]

Originally Posted by Kgolf31

My car is only an autocross car now. All it sees is high punishment.

I'm over revving the engine (7200 rpm but still) and still don't have issues. Have had AA headers for almost 2 years now. I was inital release buyer

What other mods do you have besides the headers and is the motor an n52 or n51?

[Dan Bilzerian]

Originally Posted by hassmaschine

I dont think that is true. Valvetronic 3 did get a faster motor but that started on the N55.

Would that motor work in the other n variant engines?

[Kgolf31]

Originally Posted by E46driver

What other mods do you have besides the headers and is the motor an n52 or n51?

Engine wise?

aFe Stage 2 SI Intake, Headers, Custom Burns Merge and 3" Custom Exhaust, AA Tune.

N51 Engine

Everything Else? Lots of stupid stuff.

[Taskmaster]

Originally Posted by Dan Bilzerian

Would that motor work in the other n variant engines?

Yeah Hass, give us answers!

[Taskmaster]

Originally Posted by Kgolf31

Engine wise?

aFe Stage 2 SI Intake, Headers, Custom Burns Merge and 3" Custom Exhaust, AA Tune.

N51 Engine

Everything Else? Lots of stupid stuff.

I'm falling down that hole myself, some people like Pete, are already there.

[hassmaschine]

Originally Posted by Dan Bilzerian

Would that motor work in the other n variant engines?

No, different hardware to drive the motor.

I'm wondering though if increasing the mm/s speed changes the opening time at all. I tried 20mm/s and couldn't tell any difference.