trash site

[omfgzilla]

This thread's gettin my tail so hard.............MOAR!!!

Sponsored

 

[darkvader0]

I look forward to your input on how to tune for an upgraded turbo. That is where it loses me. I haven't even figured out yet how it derives a target boost pressure from the relative load.

 

[razorlab]

I look forward to your input on how to tune for an upgraded turbo. That is where it loses me. I haven't even figured out yet how it derives a target boost pressure from the relative load.

Compressor table, PID tables, wastegate base tables, boost setpoint tables.

 

[razorlab]

Is this thread actually helping anyone? It takes a good amount of work to create these posts so if it's not actually helping anyone, I'll use that time for something else.

 

[Branflakes123]

Selfishly, keep going, I enjoy consuming this.

Practically, I don't use it.

 

[i3igpete]

eagerly awaiting the next lesson. i don't know what I don't know, so i don't know what to ask until you clue us in.

but as someone that also lived the arc2/mapecu/AEMEMS life, the multiple knobs to achieve the same cylinder fill is paralyzing. dont know where to even start if I were to upgrade the turbo, beyond starting with ewg tables and extrapolating that last column in the reference torque table.

I think someone said in a different thread that the dme can be programmed to translate pedal angle to throttle angle in a 1-to-1 way. am i misremembering?

 

[S2gizzle]

Is this thread actually helping anyone? It takes a good amount of work to create these posts so if it's not actually helping anyone, I'll use that time for something else.

I am also eagerly awaiting the updates. Super helpful thread

 

[ColonelAdama]

Is this thread actually helping anyone? It takes a good amount of work to create these posts so if it's not actually helping anyone, I'll use that time for something else.

I am saving every post for when i go unlock down the line ?

 

[razorlab]

I think someone said in a different thread that the dme can be programmed to translate pedal angle to throttle angle in a 1-to-1 way. am i misremembering?

Yea, that's an easy one. It's a pedal torque request.

So, for example. If you press the throttle pedal 55% at 80kmh (50mph), it will give you 70% throttle. This will also be a percentage of torque request.

One thing to note here. Some tuners set the Torque Request/Limit tables very high out of the way and let the Load Request/Limit tables do all the work. However, this has a really bad consequence that now the throttle request mapped to torque request in the table above will now be trying to target VERY HIGH torque and the throttle gets super twitchy, the throttle snaps open at low pedal requests and then slams shut as the Load Request/Limit kicks in. Driveabliity suffers. I would even say it makes the car suck to drive.

So don't be lazy like that.

 

[i3igpete]

that actually brings up a good point, are there multiple generally-accepted tuning strategies for bmw DMEs? If so, what are the pros/cons when you compare them?

 

[razorlab]

that actually brings up a good point, are there multiple generally-accepted tuning strategies for bmw DMEs? If so, what are the pros/cons when you compare them?

Most of the bigger tuners just crank up the boost control PID and let the DME figure it out. This makes the boost turn into a lightning bolt. They do this to crank out tunes faster. It's lazy tuning. Sometimes it's fine but it also can make the car feel really inconsistent. You can even hear the WGA pulsing with this kind of tune. I'm sure the turbo wheels love being sped up/slowed down hundreds of times during a WOT pull.

What boost should look like in a good tune:

Vs this trash:

 

[itspeat]

I like my boost control like I like my relationships.
A rollercoaster of ups and down and unhealthy for me over the long term.

 

[razorlab]

Let's start talking about boost control.

You might have seen or heard some people stating that the B58 doesn't use a base WGDC table like other cars do, or like older Japanese cars do...

This is 100% incorrect. It uses a base WGDC to set, you guessed it, the base wastegate duty cycle, before any of the dynamic boost control starts.

In our DME it's called the "Wastegate Position Feed-Forward" table and it looks like this:

It is best practice to get this table set to the WGDC needed to hit as close to your boost target as possible. Depending on the setup, you might want to buffer this a little bit and and understate the WGDC numbers a little bit and let the PID system add in the extra. This can help with boost spikes and such. We will get into the PID part of the system later.

The table has Turbine Distribution Factor on the X-axis, MAF Pre Turbine on the Y-axis, and Wastegate Duty as the values in the main section. How this table does differ from other cars is that there is no RPM axis, instead it does it by a combo of the two things on the X and Y axis. Turbine Distribution Factor is basically how much or how little amount of exhaust gases are passing through the wastegate (the flapper thing on the turbo). If all gases are passing through the wastegate, you get little to no boost. If the flapper is closed, you will get all the boostz. MAF Pre Turbine is, you guessed it, how much air flow you are shoving into the engine. Usually, more flow equals higher MAF numbers.

You will notice that with the OEM values, 99.998 wastegate duty cycle corresponds to a 1.00 Turbine Distribution Factor. All the duty used to basically close the door and hit maximum boost that the turbo is capable of. So the OEM programming is almost a nice 1:1. This works nicely on the oem setup. When upgrading the turbo and/or adding more free flowing other parts like an exhaust and such, you will want to tune this table a bit to get everything working well again, which is what I explained in the first paragraph above.

In order to tune this properly you need to datalog these items:

Boost (Pre Throttle)
WGDC (Base)
Turbine Distribution Factor (Turbine/Wastegate)
MAF Pre Turbine (Actual)

The following screenshot is with the dynamic PID boost control system turned off, so what is in this log is the DME running 100% off the Wastegate Position Feed-Forward table. Depending on how you like to tune, this might be a more helpful way to set the correct base WGDC for the boost desired before turning on the PID system and wrestling with it. If you are new to tuning, this will be a MUCH easier way to tune this table. You could even run straight off this table without dynamic PID control if you like, just know that you might/will run more/less boost depending on the environment, gear,etc. PID helps add/subtract to hit a certain boost level no matter the environment, gear, and driving style.

Right away you will notice that you will have to rescale the Y-axis to accommodate higher airflow numbers that a OEM turbo running higher boost than stock will hit, or an aftermarket turbo, that will have higher airflow at even the same boost numbers as the smaller OEM turbo.

I run a Pure700 turbo and I rescaled my Y-axis like this:

This covers the range of airflow this turbo sees at the boost levels I have used.

As far as the X-axis goes, you will see that the car usually runs around .650 to .850 range at WOT, again, depending on the turbo and what boost levels you are running.

So, rescale the Y-axis as needed and set the WGDC values to get close to what your boost to be, as a base. So if you are aiming for 24psi and 88% WGDC gets you that in 4th gear at 1600 MAF, at a .850 Turbine Distribution Factor, then consider that cel good. Rinse and repeat to get a tuned base boost curve. Good practice is to ramp WGDC in as Turbine Distribution Factor goes up. You can also enhance the response of the car by inflating these numbers past the 1:1 that the factory programming has.

In the end, the general shape should look something like this:

Next up... The "Compressor Map With Required Compressor/Turbine Output (Kw)" which shifts the actual Turbine Distribution Factor being used, based on a certain Kw for the turbine to spin to hit the desired boost. Fun stuff!

 

[sw9991]

I'm tentatively taking my first step into self tuning, and I really appreciate the time and effort put into these posts, as well as the hours of testing and research that has enabled you to have the findings to share. It's a massive help!

I'm running an F22 M240i, and unfortunately that means I haven't got the nice 275 wide rears and 10 inch rims that you guys have (skinny 245's on 8.5 inch rims for me). One thing I'm looking at from self tuning is the ability to restrict the torque in the first 3 gears. There appears to be the following table to adjust. Are there any others that you are aware of?

 

[razorlab]

I'm tentatively taking my first step into self tuning, and I really appreciate the time and effort put into these posts, as well as the hours of testing and research that has enabled you to have the findings to share. It's a massive help!

I'm running an F22 M240i, and unfortunately that means I haven't got the nice 275 wide rears and 10 inch rims that you guys have (skinny 245's on 8.5 inch rims for me). One thing I'm looking at from self tuning is the ability to restrict the torque in the first 3 gears. There appears to be the following table to adjust. Are there any others that you are aware of?

Yes that is correct. On Gen2 DME, you also need to set some flags to make that table active, not sure on Gen 1 if you need to or not.

You can also set torque % per gear in the BM3 options.

Sponsored