[Biginboca]

One of the benefits of Independent Throttle Body is there are no pumping losses as there is no vacuum ahead of the cylinder intake ports (or throttle body to be exact). This increases throttle response because on NA motor the engine is able to draw air in quickly and without much resistance, which is the primary benefit of ITB

Valvetronic also has this same benefit because the throttle body is always wide open there really is not much restriction also

Our (N52) throttle response is mainly limited (controlled) by how fast the valvetronic motor can turn the worm gear which increases valve lift. Supposedly the ITB s65 has faster throttle response than the N52, but to me they feel pretty similar. But I’ve not driven the cars with the intent of testing that just general feel of the drive.

Does anyone have a video of the valvetronic motor in action? How fast does it open the valvetronic to full lift? Early valvetronic specs were published at .3 seconds which frankly seems kind of slow if you imagined letting off gas, shifting (.3 seconds maybe) and then applying throttle and waiting (another .3 seconds) again for the engine to wake up. I suspect and hope our valvetronic is faster than .3 seconds to go from full close to full open.

I’ve seen the BPC video where he turned it with a drill and it was very slow, like I hope it’s much faster than that with the valvetronic motor working it. Start around 5:00:

Has anyone seen a way to program the speed at which that motor activates the worm gear? Can the voltage to the motor be stepped up slightly to increase its speed of actuation? Is there a faster motor?

Today I was imagining going old school and attaching a cable from the accelerator to the valvetronic mechanism tooth gear so it just worked instantly lol

[hassmaschine]

From what I have seen datalogging the eccentric shaft - you are barking up the wrong tree. Valvetronic changes position surprisingly fast.

A cable throttle? Good god... Good luck.

Btw, yes there is a parameter for vvt change speed - but it makes no difference that i can tell.

[Biginboca]

Originally Posted by hassmaschine

From what I have seen datalogging the eccentric shaft - you are barking up the wrong tree. Valvetronic changes position surprisingly fast.

A cable throttle? Good god... Good luck.

Btw, yes there is a parameter for vvt change speed - but it makes no difference that i can tell.

That vvt change speed is really what defines our throttle response, correct? So altering that would be how a car would in essence change throttle response from sport to comfort modes? I can’t think of any drawback to having that set to balls out lol

Will the new pro tool access that parameter?

[hassmaschine]

No - IMO things like AE and the physical characteristics of the engine (velocity, weight) are far more important.

What you are talking about requires a standalone computer, which is more likely than not inferior to stock in so many ways. Which begs the question - why bother with an N52 if you want to bypass valvetronic?

The posistioning of valvetronic is not directly affected by pedal input either - which is just your first problem.

[6ixSpd]

Originally Posted by Biginboca

Supposedly the ITB s65 has faster throttle response than the N52, but to me they feel pretty similar. But I've not driven the cars with the intent of testing that just general feel of the drive.

Not sure about the other technical discussion here, but I can - without a shred of a doubt - assure you that the N52's throttle response is not anywhere near as instantaneous as the S65's. Not even close.

[Biginboca]

Originally Posted by hassmaschine

The posistioning of valvetronic is not directly affected by pedal input either - which is just your first problem.

Ok, thanks. I’d like to understand the valvetronic better... do you know what data the ECU uses beside throttle position to set the valvetronic position?

I assumed since the engine runs fine without a manifold and TB that throttle is regulated by the amount the valves open. So “more pedal” means more valve opening and increase in rpms. (And then the timing of the opening is continuously varied by Vanos throughout the rev range to optimize power output as rpms increase.)

Also you said the valvetronic opens surprisingly quickly. Do your logs show a time it takes to fully open? I’m really curious how fast it is.

[hassmaschine]

That's a tough question. There are whole models that are used as inputs to the valvetronic actuator/position, which each have their own massive set of variables/inputs. Obviously, pedal position is one of them - it just doesn't affect it directly (except at WOT).

You can only log at about 60hz, so you probably can't see it happening. I was expecting to be able to watch the eccentric shaft position move as I went from ~10% to 100% throttle, but it just opened all the way. Not to say there is no delay, but whatever there is is imperceptible. Other factors (the shape/size of the manifold, inertial, accel enrichment) are going to be much bigger factors. Consider the S65 has 'instant' throttle response, but also uses the same throttle pedal and has electronic throttle actuators.

There are a few scalars that could possibly increase the response time, but I think you risk damaging the VVT system if you change them - because it actually calculates a time constant for how fast the VVT motor will move, but if it moves too fast it could hit the travel limits of the eccentric shaft. Presumably, the engineers left *some* safety margin that could be used up, but I have no idea how much, and logging is too slow to see any change anyway.

[rjahl]

Originally Posted by hassmaschine

That's a tough question. There are whole models that are used as inputs to the valvetronic actuator/position, which each have their own massive set of variables/inputs. Obviously, pedal position is one of them - it just doesn't affect it directly (except at WOT).

You can only log at about 60hz, so you probably can't see it happening. I was expecting to be able to watch the eccentric shaft position move as I went from ~10% to 100% throttle, but it just opened all the way. Not to say there is no delay, but whatever there is is imperceptible. Other factors (the shape/size of the manifold, inertial, accel enrichment) are going to be much bigger factors. Consider the S65 has 'instant' throttle response, but also uses the same throttle pedal and has electronic throttle actuators.

There are a few scalars that could possibly increase the response time, but I think you risk damaging the VVT system if you change them - because it actually calculates a time constant for how fast the VVT motor will move, but if it moves too fast it could hit the travel limits of the eccentric shaft. Presumably, the engineers left *some* safety margin that could be used up, but I have no idea how much, and logging is too slow to see any change anyway.

I think a bigger factor is the load management and how that affect the ignition advance when a load is initially applied. Stab the throttle a 3,000 RPM and the timing will drop to 12 degrees. Stab the throttle at 2,000 RPM and runs run past 3,000 RPM the rimming will be near the target of 18 degrees or so. This will change pending adaptions, gear and temp.

[rjahl]

I dug through some old logs and found data that shows a VVL eccentric cam close, 169 degrees to 54 degrees, in 250 milliseconds.

Quickest opening log I found was 110 to 171 degrees in 483 microseconds. In this period of time the engine RPM ranged from 3,300 RPM to 3,800.

This is a factory three stage N52 tune. Max VVL opening is about 172 degrees = 9.7MM lift and idle is about 40 degrees.

This is not saying that the VVL can't open any faster, I just did not intentionally log the opening speed. I was easy on the Tires for most of my logs, rolling starts and such.

[gT-BMW]

this is juicy stuff

[Biginboca]

Originally Posted by rjahl

I dug through some old logs and found data that shows a VVL eccentric cam close, 169 degrees to 54 degrees, in 250 milliseconds.

Quickest opening log I found was 110 to 171 degrees in 483 microseconds. In this period of time the engine RPM ranged from 3,300 RPM to 3,800.

This is a factory three stage N52 tune. Max VVL opening is about 172 degrees = 9.7MM lift and idle is about 40 degrees.

This is not saying that the VVL can't open any faster, I just did not intentionally log the opening speed. I was easy on the Tires for most of my logs, rolling starts and such.

Wow, 483 MICROseconds to go to full opening? So that’s .0483 seconds to go from a little more than half open to full open, right? That’s pretty dang fast if you ask me!

I actually have a hard time believing that after seeing and fiddling with the mechanism and it’s worm gear. Thinking out loud here... That motor would have to make probably at least 10 rotations in that .05 seconds time to get that much movement. I guess that is possible... I wish we had a video of it in action real time I’d like to see that.

That 110 to 171 degrees is as measured and confirmed by the eccentric cam sensor right? Not a requested value of the ecu but the actual measured time and measured degree of movement that resulted?

[rjahl]

Originally Posted by Biginboca

Wow, 483 MICROseconds to go to full opening? So that's .0483 seconds to go from a little more than half open to full open, right? That's pretty dang fast if you ask me!

I actually have a hard time believing that after seeing and fiddling with the mechanism and it's worm gear. Thinking out loud here... That motor would have to make probably at least 10 rotations in that .05 seconds time to get that much movement. I guess that is possible... I wish we had a video of it in action real time I'd like to see that.

That 110 to 171 degrees is as measured and confirmed by the eccentric cam sensor right? Not a requested value of the ecu but the actual measured time and measured degree of movement that resulted?

Yes, those were actual values. Wert and not target Sol values.

[gT-BMW]

an ITB with cable throttle and tps with an EMS that can read the values will increase throttle response for sure. combine the ITB with a vacuum box can increase HP in all rpms. i think the m3s have this setup. but there FBW systems and not cable. must of been alot of R&D on bmw side.

There are forums like k20.org and honda-tech.com that forums users created these systems for the k-series with great success. combine it with cnc head porting these motors rev to 10k+ rpm. now anybody can buy it. no reason why n52 can benefit from this but i dont know about 10k+ rpm on n52

[PhaseP]

Originally Posted by gT-BMW

an ITB with cable throttle and tps with an EMS that can read the values will increase throttle response for sure. combine the ITB with a vacuum box can increase HP in all rpms. i think the m3s have this setup. but there FBW systems and not cable. must of been alot of R&D on bmw side.

There are forums like k20.org and honda-tech.com that forums users created these systems for the k-series with great success. combine it with cnc head porting these motors rev to 10k+ rpm. now anybody can buy it. no reason why n52 can benefit from this but i dont know about 10k+ rpm on n52

Something interesting from http://www.autozine.org/technical_sc..._Ignition.html

"Now you might think about Honda's VTEC. Can Valvetronic use its variable lift to increase power like VTEC ? Sadly, Valvetronic is actually less efficient at high rev than conventional engines, let alone VTEC. As you can see, the camshaft drives the long intermediate rocker arms, in turn drive finger followers, this generates quite a lot of friction. Therefore the efficiency and refinement of Valvetronic engines drop rapidly at over 6,000 rpm. No wonder in the foreseeing future BMW will not equip its M-power engines with Valvetronic."

[hassmaschine]

Yeah, thats pretty much nonsense. Every moving part of valvetronic uses roller bearings. And the S55, an M engine, uses the same system as the N52. We already know the N52 can make good power until at least 7500rpm.

Vtech isnt analogous to valvetronic anyway - its variable timing, which Vanos does better anyway.

[PhaseP]

Originally Posted by Biginboca

Wow, 483 MICROseconds to go to full opening? So that’s .0483 seconds to go from a little more than half open to full open, right? That’s pretty dang fast if you ask me!

I actually have a hard time believing that after seeing and fiddling with the mechanism and it’s worm gear. Thinking out loud here... That motor would have to make probably at least 10 rotations in that .05 seconds time to get that much movement. I guess that is possible... I wish we had a video of it in action real time I’d like to see that.

That 110 to 171 degrees is as measured and confirmed by the eccentric cam sensor right? Not a requested value of the ecu but the actual measured time and measured degree of movement that resulted?

300 msec from close to full open:

http://www.iosrjournals.org/iosr-jmc...volume-2/5.pdf

Which makes sense, I counted like 17 revolutions in the BPC video from closed to full open. 300 msecs per 17 revs makes 3400 RPM for the valvetronic motor speed. Sounds reasonable.

[hassmaschine]

Originally Posted by PhaseP

300 msec from close to full open:

http://www.iosrjournals.org/iosr-jmc...volume-2/5.pdf

Which makes sense, I counted like 17 revolutions in the BPC video from closed to full open. 300 msecs per 15 revs makes 3400 RPM for the valvetronic motor speed. Sounds reasonable.

Cool, thanks for finding that! - I haven't seen that doc before.

I should point out the quoted specs are for the first generation valvetronic - not the second generation system used on the N52.

[PhaseP]

Originally Posted by hassmaschine

Cool, thanks for finding that! - I haven't seen that doc before.

That is Valvetronic-I (used on 316ti), reading more and finding this one

https://www.e90post.com/forums/attac...9HWQotTBL1qgOG

Valvetronic-II is N52. N52's valvetronic maybe faster. They mention Valvetronic-I didn't have rollers in this one above, so that comment I copied for high rpm friction may apply more to first version.

Valvetronic-III is what is used on N55.

N52 is using a stepper motor, those motors cannot be very fast, not expert but that is what I know.

N55 is using a brushless servo DC motor with internal sensors. That maybe faster.

[Biginboca]

The hard stops look cast into the head and pretty rubust. I don’t really see the risk of accelerating the operating speed if possible...

[rjahl]

For those that are interested,

I've attached one of the very few logs of my N52 where I actually logged the foot pedal "Stat_fahrerwunchsch_pedal_wert" values along with everything else.

I found it easier to leave everything in German, so it might be a struggle for some to read.

I've attached the Raw Datalog in a zip file.

This is just some random log. Can't really remember why I took this one. But you can see that due to the Driver's wish mapping, the DME has already gone full load, well before the throttle is depressed to the floor board. The VVT opens a little slower, presumable to keep the intake port air velocities at lower RPM.

Screen capture is only part of the data that can be found in the CSV file.

Good study file for anyone who has not logged their engine.

[nsjames]

Originally Posted by PhaseP

N52 is using a stepper motor, those motors cannot be very fast, not expert but that is what I know.

N55 is using a brushless servo DC motor with internal sensors. That maybe faster.

that's mostly dependent on the stepper driver. There are certainly NEMA motors capable of 3K RPM with the appropriate driver. I'd imagine that the VT motor is a custom bit, and with a custom driver, so it can really be anything that BMW wanted. I'd say it's not out of the realm of possibility that it's a 2-3K RPM motor. The biggest problem of driving steppers quickly is a lack of torque. I'm not sure what the requirements are to move the valvetronic mechanism.

A sensored brushless motor will be superior in torque and speed, but probably not be as accurate. I imagine that it's not such a big deal because you still have an external sensor to tell you shaft position, and you're not relying entirely on the accuracy of the motor.

[Phyrexia]

Found this while doing some research after lunch. Haven't ever seen it before. I wonder if VT II is less prone to frictional losses than VT I. Seems to be the case.
http://115.112.66.136/Employee_Infor...MTZ/aug_08.pdf