Pro Racer - Tuning For Fuel System Modifications

[JoeBoomz]

I had discussed this a bit previously in my LiveLink thread here, but I need to start a dedicated thread to give this subject the detail it deserves.

Here I will also discuss changing XNorm and YNorm (table axis) values in Advantage III, which is an important step for building a tune specific to one vehicle. Stock tunes carry values that are built to work across a large number of vehicles but you can greatly improve driveability by focusing table ranges on only what your car will actually encounter (a very basic example is if a tune contains data for running at 7000 RPM but you've set your rev limiter to 6000 - you do not need the 7000K data).


The Marauder ECU has a voltage table with the voltages required by the fuel pump(s) to maintain a desired fuel flow rate. From this base table, the ECU will adapt ("learn") the voltage values that are actually required to reach specific flow rates.

With factory hardware and a factory tune, the programmed base table is very close to the real values that are needed, and very little learning is needed. The result is that there is no negative impact on driveability due to the learning process if the car memory is lost (i.e., you disconnect a battery terminal).

The problem is, if you change your fuel hardware at all, the voltage table in your tune will become more inaccurate. Every time the car is reprogrammed or the memory is lost, the ECU has an incorrect table to start with and the learning process can result in some very poor driveability (stuttering, shuddering, etc.) and in the worst case, can lead to lean fuel conditions and detonation.

So, as you change your fuel hardware, the base fuel pump voltage table in the tune should be adjusted to compensate, in order to minimize the required learning process. Changes to fuel hardware can be any of replacing the fuel filter, improving the fuel flow path, replacing a failed FPDM, changing or upgrading a fuel pump, or adding a BAP. All of these can affect the voltages required by the fuel pump at different flow levels!


In SCT's Advantage III, here is the stock Marauder fuel pump voltage table:


The table is a function of Target Rail Pressure (vertical or y-axis) and Fuel Flow (horizontal or x-axis). The values in the table are actual voltage values passed to the fuel pump.

SCT helps with some fuel pump value files - for example if you replace a stock Marauder fuel pump with a GT40 pump, you can load a value file for the GT40 pump to bring your voltage table more in line with what the pump will need.

However there is no value file for upgrades like the BAP - so we must do some data logging and build a new table ourselves. This means that we drive the car for some time to allow the ECU to learn a new table (do it gently if you've made drastic changes, to avoid pinging or detonation). Once the car drives normally and the shuddering, loss of power, and wrong rail pressures are all gone, the learning is complete and we can log the data to build a new fuel table.

Data logging is done using SCT's LiveLink (I describe it here), and when we've got some good data with all load values and RPM ranges, we can use the logged data to build a new fuel table.

The values to log to do this are FPUMP_FLOW_RATE, FPUMP_ACT_VOLTS, and fuel_rail_pres_psi.

We start by looking directly at our data log CSV file first in Microsoft Excel. We will use Excel's graphing capabilities to eyeball where we need to focus on for axis values.

So here, I select the entire FPUMP_FLOW_RATE column by clicking on "B" at the top...


Then I hit F11 to make me a graph of the data:

This is a nice visual representation of our fuel pump flow rate throughout our drive. But it doesn't help us much in this case.

So I go back to the sheet again, and sort the data in ascending order using Excel's SORT button (looks like an A-Z with a down arrow).

Here's the graph with the data sorted:


NOW we can see the flow rates where we spend most of our time driving. Most of our driving resulted in flow rates under 0.5 with the majority under 0.250. And our maximum value was under 3.00 (higher HP cars will see higher fuel flow rates).


Let's take another look at the voltage table in Advantage III:

The fuel flow values on the horizontal (x) axis are somewhat evenly distributed and they go to 4. If you look at the graph above, we see that it would be better to focus this table more on the lower values, and less at the higher values, stopping at 3.

Advantage III allows us to change the table axis configurations through XNORM and YNORM tables. Here's the stock XNORM table corresponding to the x-values in the fuel voltage table:


...we change it as follows:

Examine the top part of the table and you will see we increase the axis value in increments of 1. But the bottom of the table has a much finer resolution, with increments of 0.0625.


Again, let's take another look at the voltage table in Advantage III:

Check out the vertical (Y) axis - it references rail pressures ranging from 40-75 PSI. I am curious why they do this, as the rail pressure is tuned to a specific targets (i.e., 40PSI, 50PSI in some temperature conditions).


So let's look at our actual fuel rail pressure, and do the same thing as we did for the flow rate.
Here's the graph (sorting the data is not required for this one):

The datalog shows we spend most of our time around 40PSI, with some spikes to 50 and even 60. Why bother having such a broad PSI range in the table if the car is not going to encounter those values?


Continued in Post 2...

[JoeBoomz]

...Continued from Post 1

So here are the YNORM values for the Fuel Pump Voltage table:


...which we change to:

IMPORTANT to Note: I have found greatly enhanced driveability by changing this table to focus on a smaller rail pressure range (the values actually encountered by the vehicle).


Now that we have the axis values set to where our car actually operates, it's time to populate the table with values. I use my Data Analyzer Spreadsheet to do this (posted here).

Before you open the spreadsheet, set up the rawdata.csv file with the columns in this order:

...then open up Tuner V3.xlsx. For this task I work in the SmallTable tab (the tabs are shown at the bottom left of the sheet).

Notice your horizontal (X) and vertical (Y) axis are set up with rail pressure and flow rate, just like the Fuel Pump Voltage table in Advantage.
Start by erasing the default axis values:


...and get the new axis values from the Advantage table (notice they have been changed by the XNORM and YNORM tables we edited above):


...and plug these into the Tuner V3.xlsx spreadsheet and edit the compare-at thresholds to the minimum change amount in the axis (shown here as 0.0625 for flow rate and 1 for rail pressure, adjacent the yellow squares at the end of each axis)


Voila, the table populates with values from the datalog that meet the axis values and criteria. We can copy-paste this back into the Advantage Fuel Pump Voltage Table:


Notice we have some empty (white) squares in the table in Tuner V3.xlsx. This is because our compare-at thresholds are TOO LOW and need to be increased. Specifically in this case it is the FLOW RATE - need to bump it up to 0.250, then 0.3, then 0.4, then 0.5, and as the missing squares fill in, paste them into the corresponding spot in Advantage III.


Here's my completed table:



We are done! And going forward, the fuel is bang on every time the car is reset or the tune is updated (until we change the fuel system again). There is no shuddering, or loss of power, or detonation when driving the car right after a reset because of an incorrect Fuel Pump Voltage Table.
Continued in Post 3...

[JoeBoomz]

There is another method to doing this than using my spreadsheet. LiveLink Gen-II can build a good histogram that you can copy/paste into Advantage as well. I posted details here in post #4.