BMW N54 – Data Logging with COBB AP – Part 1
Posted on 18 July 2012
Data logging is an invaluable tool when it comes to maintaining reliability and checking on your modified car. It allows you to know what your car is trying to do (requested values) and what your car is actually doing (actual values). Knowing the concepts is one thing – being able to understand it and utilize it is another.
The most common question I get in regard to datalogging is people asking me what is useful to record when doing a pull. The good news is that Cobb setup the default logging parameters really well. It has just about everything you would want to see. It seems the confusion then becomes what all the values in that .CSV file actually mean. This is what I am going to cover in today’s blog entry.
The first thing I do when opening a datalog is find the exact data that represents the WOT run I am looking for. There are a few ways to do this. One way is to look at pedal position to determine at what point you put your foot all the way down. You can simply delete all the rows above that data point. Then go to the end of the run which is often represented by the pedal position no longer being maxed. I usually like to start by looking at a single gear pull. If your log is of a multiple gear pull it may help to isolate just one of the gears and remove all of the other data (rows in excel).
Now that we have isolated this pull it is time to see if the car is doing a few things.
-Is the car requesting the values that I would like it to request?
-Is the car hitting the values that I am requesting?
-If the answer is yes to both – is the car happy running these values?
Lets start with the first question. Depending on the modifications you have done you will want to start by determining if the boost curve is appropriate. The car uses “Boost Req Abs” as the name for how much boost the car is requesting. This value is representative of atmospheric pressure (14.7psi) + boost pressure. Basically, if the value shows 31.58 then the car requesting is requesting 16.88psi (31.58-14.7). This boost curve is directly correlated to the load table that is set in your map. The higher the requested load the more boost you are asking the car to make. Our turbos are rather tiny and can handle 18-19psi (FBO vehicles) in the lower to middle RPMs but quickly need to taper as the RPMs rise to remain within their efficiency island. Once you are aware of these requested values you have the answer to the first question. These are what the car is trying to request – is it what you want the car to be requesting? That is up to you.
Now it is time to take a look at the actual values to see if the car is running what you are requesting of it. First we take a look at the “Boost Mean Abs” column which represents how much boost the car is making. This is the result of the “Boost Req Abs” and is the actual value the car is creating. In the lower to middle RPMs these values should mimic closely what your “Boost Req Abs” is calling for. As your RPMs increase they may begin to fall a bit below the requested values. This is perfectly okay as Cobb setup the maps to run within acceptable ranges of WGDC so that the turbos are not out of their efficiency range. Your actual load typically will not reach the requested load as there are many limiting factors the car is using when determining how much boost to run. Essentially the car will run the least amount of boost to satisfy the load request for the environmental conditions. I can command 190 load but if It is 84* and humid the car may only hit 178 load and satisfy the many correction tables and calculations in determining output. It is normal for the actual values to not hit the requested values. You want to see that it follows the same increasing and decreasing slope of the requested values as the RPMs rise.
Now that we have covered the first two questions – we move into the fun part. Is my car happy? To determine this we look at a few things. The first thing to consider is environmental factors. Take a look at the “Charge Air Temp” to see if the conditions are reasonable. Heat is an engines worst enemy and can lead to increased knock occurrences and other power reducing events. As such when your boosted values are higher than 100* throughout the pull it is not ideal – but it is Summer so it is rare to see anything less (water and meth will help). When the temps get this hot it makes for “noisier” datalogs. The second thing I like to look at is the “TPS Act” which represents what the throttle blade is actually doing. The throttle plate will typically show a value of 80-81.xx when it is fully open. During a WOT run if you see it close it may be because of slight overboosts. Small changes in the throttle position are nothing to worry about. In fact, small percentage changes will not even be noticed on a dyno plot. If you have massive closures you may want to investigate into why before continuing. If nothing shows up (or minor blips) then you can continue on to the next area I like to check. Make sure the car is running the AFR you would like. Although we have a Direct Injection motor there are negligible gains from leaning it out too much. In high load, high RPM settings I like to see AFR curves in the high 11:1 range. Now we can move on to timing! “Timing” is a great way to increase power while maintaining steady boost levels. By monitoring knock events we can see how “happy” the engine is running the timing curve set in your map. The n54 is VERY sensitive and seeing -3.88 and other occurrences of timing being pulled is okay. What you are looking for is multiple adjacent cylinders pulling large amounts of timing at the same time. If you are getting a few random negative timing values on a few cylinders here and there it is most likely just noise. Take a few logs and see if they are repeatable in the exact same spot or if it just comes and goes randomly. While playing on a dyno and testing out multiple variations of maps we learned that these random timing corrections are not noticeable in the tq curves. Only when the timing corrections are simultaneously occurring on multiple adjacent cylinders is output actually affected. Therefore we are looking for these larger events. In their absence – it usually means the car is pretty “happy”.
This is a good start to understanding how logging works. Next week I will do an entry on a good way to setup logs in excel for easy viewing/reading. I am a visual person and like to add columns using formulas to see calculated values as well as inserting graphs for the visual representation of what the car is doing.