Managing data to track progress

So after some messing about with the new air intake I decided it was time to think about tracking changes over time. The Rotofab intake took about 1hr to fit. Note the nice large cone.

Sounds much nicer with a bit of intake roar now. Some may note the crusty MAF hiding out there. I was hoping this model had the newer LS7 blade maf. But its still the old 2005 85mm LS2 type maf. Of course it must eventually die. But I'm going to use the MAF to give me a bit of info first.

Rotofab installed

A piece of straight 4 inch silicon hose has been ordered to sit where the MAF is now. I want to get all the airflow and torque tables right first before removing the MAF though so I will be failing/moving between Speed Density and MAF airflow models to identify any changes in calculated torque values etc. The summary is in Speed Density the tables are not right. This is because the OEM Speed Density values and tables are not right. I guess its easier to get one column in one table right rather than tables like this one right! There are 450 cells in this table and there are 4 tables like this!

E38 VE Polynomial factors

Also make any modification to Speed Density airflow and the tables all need redoing! Thats because they are static and map engine efficiency based on RPM, Intake pressure, temperature etc. This is also why lazy types just tune MAF airflow models. Thats if they arnt even lazier and just tune the PE table like one tune I saw this week!

All the UK tuner tunes I have seen have been guilty of this at least historically anyway. I suspect some of this is driven by the fact that people think they can get a good tune in 1hr for £250! Ask GM how long it takes them to build a tune for a new application. Im certain its months.

Anyway there is a very good forum post on EFILive regarding the engine torque tables which are crucial in managing communications with the TCM.
See here https://forum.efilive.com/showthread.php?18710-Improving-shifts-via-the-ECM

E38 Torque Model Tables

The table above is the one mentioned in the article and allows the engine to calculate torque output based on a number of factors similar in concept to how Speed Density airflows are calculated.

The plan is to visit Mr Wright at Surrey Rolling Road after most mods to get a good view of things. But rather than relying on a trip there every week to guage progress its possible to use EFILive logging to get information which is just as good depending on how accurate airflow models are. (Hence the discussion above about keeping the MAF in place until the Speed Density airflows match the data from the MAF).

So what data can we use? Well basically we can use measured and calculated grams of air per cylinder for a torque proxy and grams of air per second for a bhp proxy.

We can also use calculated torque values from the MAF and/or Speed Density (The topic of the torque table above) This is the main reason to keep the MAF for now. So I can get good torque calculations. Once Speed Density is giving me similar/same values I will retire the MAF.

As a plus EFILive Scan Tool allows me to export the data to CSV and therefore Excel or Access. This should allow me to do conditional evaluations of the data and tell me what spark timing at what RPM and load point gives the best calculated torque value. I was going to do the same thing but with g/cyl to proxy torque, but Jessie Bubb said he uses calculated torque while we were chatting at PRI this year. (He mentioned he even has some sort of Calc PID, Ill try and pry it off him to see how its supposed to work)

So anyway, what does one of these CSV files look like in chart form? Ok this is from the first Dyno day at SRR.

Log generated Dyno Graph (fwhp)

Does it look even remotely like the Dyno chat from SRR in the previous post? Yep. The data set at the bottom is not synced nicely. So cleaning that up would help. Also just querying the data is going to be more helpful and accurate. If I can figure out Microsoft Access I can upload each log as a CSV with different timing values in it and query out the highest calculated torque at each load point vs spark timing. Worse case picking out points and looking through each CSV import in Excel should allow a manual way of doing this.

So that basically is how I will be monitoring and tracking changes of modifications and ensuring spark timing is optimal. If these values increase, power should also increase by the same amount. So I can look at percentage gains to get the relative merits of each change. First up will be to log the Rotofab intake and see what that achieved. Hopefully that will be the next topic in this Blog.