BMW N54 Boost Control – PID and Tuning Limits
Posted on 26 January 2015
Given recent developments in flash based tuning we wanted to offer our insight into the inner workings of the N54 DME when it comes to boost control tuning or more specifically, PID based boost control.
Boost control in the N54 OEM DME as well as most modern electronic boost controllers out there is based on what is called a PID algorithm. PID algorithms aren’t exclusive to boost control and are applied to various systems where correction to a signal can be applied by feeding the actual value of the signal through the algorithm and adjusting control for it automatically by comparing setpoint/requested to actual and adjusting the controlling device which in the case of boost control is pulse to a boost control solenoid.
Concept of a PID algorithm is quite simple. It is based around knowing two values:
1) Setpoint – in the case of boost control this is requested boost pressure that the DME is targeting for a given Load
2) Actual Value fed into the PID algorithm – in the case of N54 boost control this is the signal coming from the OEM TMAP sensor feeding back a 0-5V value representing absolute pressure
DME converts the analog 0-5V signal coming off of the TMAP sensor into engineering units based on the Map Conversion table you can find in your maps.
OEM TMAP sensor is a 2.5bar sensor. That means it can read 2.5bar of absolute pressure. This means your actual boost will vary based on your atmospheric/barometric pressure which needs to be subtracted from 2.5bar to understand how much readable range you’ve got out of your MAP sensor. This readable range above your barometric is what is referred typically as the word we all love and we’re here for, “BOOST” (i.e. positive pressure or pressure above barometric).
The OEM N54 DME works on the principle of Load based tuning and PID based boost control. Load is a calculation based on various parameters primarily around “boost” (i.e. pressure at the MAP sensor – barometric presure) and charge air temp which in turn calculate into Mass Air Flow (MAF). N54 doesn’t use a MAF sensor to meter airflow on the intake side. It calculates/infers it based on other values mentioned.
In order to make boost from the turbos the DME regulates boost control solenoids via a PWM (pulse-width modulated) signal pulsing them and allowing vacuum to hit the wastegate actuator. The frequency of that pulse/signal is what is typically referred to as wastegate duty cycle (WGDC). This is where PID based boost control comes in and helps the DME regulate the amount of WGDC applied to the solenoids so that Target Load is hit as required.
Very few of us speak about our tunes in terms of target load and it makes sense as most are used to talking about “Boost”. While the DME targets a given load and given that Load is defined based on primarily boost+IAT+compressor maps and airflow modelling around stock turbos/intakes/intake pipes it ends up calculating a Requested Boost level. In your Cobb datalogs you can monitor this value by looking at the Req. Boost Abs. channel.
You can refer to the Req. Boost Abs. as the DME’s Setpoint for the PID based algorithm. When using Cobb’s ATR software there is a hard limit at 1.28bar in terms of maximum Requestable boost. If for any reason your car boosts past this level the DME’s safety mechanism will kick in to prevent motor damage due to what it perceives as an uncontrollable boost condition for its settings. For a tuner using ATR to get around this issue and provide a smooth tune they need to effectively remove this first DME safety by calibrating tables related to boost control and throttle management. In other words, if at any point in time the turbos were to spike boost and cause it to overshoot target to say 35psi instead of where they’ve set/expect your tune to be, given how they’ve set the DME tables up to get that extra 2-5psi out of your car, it ends up being quite unsafe and won’t have the DME intervene as it normally should/would. With appropriate PID control and DME failsafes in place the signal to the solenoids is taken way, DME closes throttle causing DVs/BOV to open and vent boost and the DME goes into limp mode.
If you’ve been around the N54 scene long enough or even owned an N54 for while, especially a tuned one, you know that there are countless N54s out there that have had an experience with either overboost (30FE) or underboost (30FF) induced limp mode issues that otherwise could’ve ended up in a tune that knocked damaging the motor potentially ending in a really unnecessary and expensive repair.
Some N54 ROMs, such as the I8A0S and IJE0S, have had support added by Cobb (i.e. race code) so that the full range of the OEM TMAP sensor can be tuned with PID based boost control and the DME safety net entirely in place. It is only available with AccessTuner Pro (ATP) software and only by request to Cobb from a licensed protuner shop.
Even with the above race code in place going past the limits of the OEM TMAP sensor with flash only tuning carries an extreme level of risk for a motor. N54 is a stout motor that we all know can take quite a beating but that hasn’t prevented engine failures in the past due to various reasons but also due to poor and unsafe tuning.
If you’re running a flash-only tune on your N54 today, with or without the N20 3.5bar capable TMAP sensor, the hard limit is in place for the readable boost range of 2.5bar absolute (~21.5-22psi at sea level) inside which the DME’s boost safety mechanism works. This is also the range in which PID based boost control has authority to adjust your boost. If your tune is pushing past 22psi rest assured that even though some tuners love throwing the word PID in their descriptions that PID simply doesn’t exist and is virtually eliminated. Why? As mentioned before, PID based boost control revolves around knowing the actual boost. If you’re seeing boost flatline in your logs at ~21psi it is at that point that PID based boost control and safety has basically gone out the window for your car/DME and all you’re left with basically is the wastegate duty cycle which is mapped based on what is now a really skewed MAF calculation.
Can you make power in this way? You sure can, but so could the old versions of popular piggybacks before CANBus days. There will always be multiple ways of making power on any given motor. We’ve also shown that pushing past the TMAP sensor limit is quite doable setting a former N54 HP/TQ record on a beta Stage 3 VTT twin setup. However, the difference at least with us is that we explicitly stated that such tuning is not recommended and WILL NOT be provided for any customer car given boost control safety mechanisms aren’t in place. We are not ones to hand responsibility of appropriate tuning and throw risk into customer hands or make them stare at a boost gauge in hopes that they’ll react in time to save their motor from uncontrollable boost spikes. It is also why we recommend going with an external form of boost control when tuning any car past its OEM TMAP sensor as there simply isn’t any other proper way of doing it and certainly no way of using PID based boost control.
Dyno glory numbers and tinkering are always fun and certainly help marketing buzz but its one thing to make power and another to do it with appropriate safety and control in mind. Bottom line is, if you see anyone claiming PID based boost control while going past the TMAP limit you can refer them to this thread and let them know they’re missing a key piece of information to claim effective PID based control. You can also ask them what happens if your wastegate actuator sticks (not uncommon especially with wear on the turbos) or a boost control solenoid fails in the open position or when the car is already tuned to the edge at altitude and you drive down to sea level with the same map and heavily overboost. Not pretty, potentially very costly, irresponsible and just plain unnecessary given how far the N54 has come.