Why newer engines dont bog like older engines do

[B C]

I learned something that I think @m42b32 discovered on his M54 swap, which is: the ECU on a car with an electric throttle body does not open the throttle 100% all the time. It is programmed to open to a fraction of WOT depending on the RPM. I'm not sure if the strategy is to avoid lean spikes/NOx emissions (very likely), or better control of enrichment to prevent bogging and keep particulates down (probably not). I observed this while diagnosing a bad throttle body/TPS on a car (not a BMW) I was working on while plotting the throttle opening vs resistance. It only opened 50% max when the pedal was applied 100% from idle speed, and continues to open up a fraction of WOT until you are higher in the rev range where the volumetric efficiency will allow more complete cylinder filling. On Nissan products I believe people refer to this as "Torque Limiting" and sometimes this throttle restriction is removed when a performance tune is flashed.  I do think this concept is known about but misunderstood since I think some people believe the reason is to protect the drivetrain. 

If anyone else has more information on this, chime in. I like this kind of stuff 

[m42b32]

From my understanding, the ms43 ecu (earlier M54 cars use this) has two functions that adjust the throttle body response to pedal inputs: 

 

The first is Driver Wish Input Correction, which can limit the rate of increase of the throttle body plate based on a number of different inputs. (This is all from ms4x.net)

ip_pvs_cor_max_rpl_[gear] is used by the ecu to decide if the drivers requested pvs input increase should be limited. The values in the table is the lower limit and the X-axis is the upper limit. If the drivers requested pvs input is between these values then the ecu will start limiting the pvs input increase.

If the following conditions are met then the ecu will not try to start limiting the pvs input increase:

• The driver requested pvs input is decreasing.
• The driver requested pvs input change gradient is larger than c_pvs_av_grd_max_rpl(59,99° pvs).
• The clutch is pressed.
• The driver requested pvs input is higher than c_pvs_cor_max_rpl(42,5° PVS)

When the ecu starts limiting the pvs input increase the pvs input will be increased by the value taken from ip_pvs_cor_rpl_lgrd_[gear] until the following conditions are met:

• The limitation duration specified in ip_t_pvs_cor_rpl_[gear] has expired.
• The driver requested pvs input change gradient is larger than c_pvs_av_grd_max_rpl(59,99° pvs).
• The limited pvs input is larger than the driver requested pvs input.

If any of those conditions are met then the ecu will use the driver requested pvs input and will not start limiting the pvs input again until the time specified in c_t_dly_pvs_cor_rpl(0,2s) has elapsed.

 

The second is Throttle Request Correction, which sounds extremely similar but is used in lower RPM ranges and is dependent on current engine load. Again from the wiki:

ip_tps_req_ltc_min_[gear] is used by the ecu to decide if the throttle setpoint change should be limited. If the requested throttle setpoint is lower than the value in the table the throttle setpoint change will be limited.

If the following conditions are met then the ecu will not try to start limiting the throttle setpoint change:

• The clutch is pressed.
• The requested throttle setpoint is lower than c_tps_req_ltc_min(0.248° TPS)

When the ecu starts limiting the pvs input the throttle setpoint will be increased by the value taken from ip_tps_req_ltc_lgrd_[gear] until the following conditions are met:

• The limitation duration specified in ip_t_tps_req_ltc_max_[gear] has expired.
• The requested throttle setpoint is larger or equal to ip_tps_req_ltc_min_[gear].
• The clutch is pressed.

If any of those conditions are met then the ecu will use the requested throttle setpoint and will not start limiting the throttle setpoint again until the time specified in c_t_dly_tps_req_ltc(0,85s) has elapsed.

 

I believe the throttle request correction is what was making my car very hard to shift smoothly as it is likely tuned for the original clutch, gearing, etc. which I am not using. I was able to disable both of these by basically zeroing out a table and parameter as defined in the wiki. This made a huge difference in drivability as it should now function similar to a throttle cable. I have not dug into if there is any WOT limiting but I wouldn't be surprised if there is some form of it for emissions purposes, etc. 

 

[straight6pwr]

my guess is that its designed for Karens. If a Karen's BMW jerks or stutters or bogs or do anything unordinary, they are going straight to the dealership to demand to know whats wrong. They dont know how to properly apply throttle, so this nanny was applied. Comfort Throttle Mod

(more likely fuel economy or emissions) 

[jc43089]

Interesting stuff.  On the V70R I had I had an issue where clutch engagement suddenly became jerky.  What I found was that the clutch did not have a simple switch, it was a linear potentiometer so that the ECU could determine clutch pedal position.  The potentiometer failed and the ECU must change its response rate during clutch engagement to assist smoothness and was no longer doing that.  Those had a Bosch Motronic ME7 ECU.  The way I figured the issue out was that shortly after it became jerky the cruise would not set either so that got me looking at ECU values for the brake switch and clutch sensor.

Not exactly the same thing but another example of the electronic throttle body response being different during different conditions.

[B C]

Just when you think you have things figured out and then a door opens up to uncharted territories

[HipMF]

Sounds like a CV carb. Presenting the fueling system with a step-change in manifold pressure makes it's job a lot harder. Most of the sensors that are used to measure what's going on in the engine have some sort of time-constant associated with them, so the computer only knows what happened maybe 20 or 30 milliseconds ago. Meanwhile, it's trying to guess about what the correct parameters are for the next combustion event when the current one hasn't even completed. Slowing down the rate-of-change helps it to guess better, which should result in better combustion.

There are probably other issues as well; the air that's rushing in to the intake manifold to bring it up to ambient pressure has mass and therefore inertia, so there's likely to be a little bit of reverberation in manifold pressure until everything equalizes.That's the most likely explanation I can come up with for the CV carb. Air could even potentially flow backwards through the carb, which would cause it to get fueled 3 times before it got into the engine...

[m42b32]

1 hour ago, HipMF said:

Sounds like a CV carb. Presenting the fueling system with a step-change in manifold pressure makes it's job a lot harder. Most of the sensors that are used to measure what's going on in the engine have some sort of time-constant associated with them, so the computer only knows what happened maybe 20 or 30 milliseconds ago. Meanwhile, it's trying to guess about what the correct parameters are for the next combustion event when the current one hasn't even completed. Slowing down the rate-of-change helps it to guess better, which should result in better combustion.

There are probably other issues as well; the air that's rushing in to the intake manifold to bring it up to ambient pressure has mass and therefore inertia, so there's likely to be a little bit of reverberation in manifold pressure until everything equalizes.That's the most likely explanation I can come up with for the CV carb. Air could even potentially flow backwards through the carb, which would cause it to get fueled 3 times before it got into the engine...

Very interesting. I hadn't heard of that before but it makes a lot of sense