EDIT: I took some of the original screenshots using datalog 04-08-14, but realized it was not a very good log, so I switched to datalog 04-30-14. Datalog 04-20-14 is not a single gear pull, its about a 40 roll in 2nd into 3rd, but it had more data that I was looking to show. Some of the original screenshots may not match, but the point is the same, please disregard datalog 04-08-14.
Also, for those good in excel, I have attached a copy of my template I use for Mazdaspeed datalogs using conditional formatting. It really makes things easy to look at with just a quick glance through the log.
A few things I have highlighted:
A sliding Boost scale
A highlight scale on AFR
Highlight where the car switches from Closed to open loop (the mustangs are full time closed loop, so this doesn't matter)
Since the Mazda is direct injection, I have fuel pressure outside of the optimal range highlighted as well.
I have a highlight on Injector duty cycle approaching 100%
Lastly I have a highlight scale on MAF flow.
I really encourage everyone to poke around with some conditional formatting, As all of the tools in this thread will hopefully make looking through your logs a BREEZE. Now to the good stuff though.
Alright, so I know I preach the importance of datalogging your vehicle to interpret and understand changes, but I realize a lot of people probably don't know where to start, so I figured I'd give yall a rough breakdown and hopefully give you the means of better understanding your mustang.
This is basic. Basic, basic, basic, really just breaking down the log in Excel format and kind of getting an understanding of what you're looking for/at. If this is met with a really positive response, and people are engaged and wanting more, i'll take the time to go through the SCT software as well.
So for starters, I have attached one of my old datalogs and i'll try to walk through it with yall. The one were interested in is called Datalog 04-30-14.
So First thing is first, its a LOT of data and can be overwhelming at first, so the first thing I always do is freeze the top row. This is not meant to be an excel class, so sorry you'll have to be familiar with that on your own
. Once you have the top row frozen, your columns will basically always have a header as you scroll up and down so you can see what each column is.
So what are we looking for in a datalog? The answer is anything and everything. Almost every single aspect of your engine can be recorded and interpreted through a log, so lets just start with some basics.
1. Where does the log actually start? So my experience is with turbo 4 cylinders, and knowing where the WOT section of the log started was really easy; its where the boost went positive lol, but how do we tell in the mustang? Well, I'm sure there are several ways, but I think the easiest would be looking at the accelerator position.
Here it looks like 82% is Wide Open Throttle on the mustang (or at least 5.0) platform. Its not uncommon for cars to see less than 100% in this case, you'll notice the mazda log I attached is only about 76%. So we know I'm Wot but... now what? Well, the next thing I like to do is break the log down into 500 rpm segments. Most tuning software is not broken down into every single RPM, they're broken into 500 rpm segments typically and the computer will interpolate between those breakpoints.
Looks like I went WOT right about 2000 rpms in this log, and breaking the log down into these breakpoints makes it easier to analyze. 500 rpms at a time is easier than a single 7000 rpm mass of data all at once.
2. Alrighty, now that the log is broken into bite size chunks, the first thing I always check for is Knock. Here is an excerpt I have shamelessly stolen from JAJ5point0 on another forum to explain what it is we are looking for:
The knock sensor PID is measuring crank degrees in the direction of rotation that are added to or subtracted from the normal (usually borderline knock table) spark timing.
When the KS PID reads positive numbers, it means there are more crank degrees at ignition, which means a spark that happens later than normal, and most people call that ******. A negative number means fewer crank degrees at ignition; sure enough, that's ADVANCE.
When you first encounter it, it seems odd, but once you understand it, it makes sense.
You should normally see negative numbers as the KS looks for incipient knock - that's how the gets the most power out whatever octane you're running. It hunts relentlessly for the edge of knocking. Positive values mean your spark tables are providing too much advance.
In laymen's terms, this means "negative knock," is the computer adding more timing, and "positive knock" means the computer is pulling timing due to actual knock.
Sho nuff, right here we can see the computer adding .05* and 1* of timing respectively. Notice anything interesting? It changes right at a 500 rpm breakpoint
, and if you follow that log you'll see a few more changes occur at 500 rpm breakpoints towards redline. This is why we break things out.
3. So now that we know we're good on knock, lets check our fueling using a combination of AFR / Lambse / Short term fuel trims and long term fuel trims. Its a lot to take in, so lets try to break them into useful pairings again.
For those that don't know, Short term fuel trims are instantaneous corrections your computer is making to your base MAF (mass air flow sensor) curve to correct for any anomalies. Since they're instantaneous, I ALWAYS have my computer and datalogger hooked up any time I do anything major in the engine bay and monitor STFTs when I start the car. This is an excellent way to see if you have any vacuum leaks, or if the car is otherwise unhappy. Excessively positive STFTs (typically around 30-40%) are a dead indication that you have a leak somewhere, and can prevent you from driving around under dangerous conditions.
LTFTs on the other hand, are stored averages of the STFTs for given breakpoints. We don't have access to these breakpoints since we don't know how the tune was written, so for now we're just concerned with the values. In my personal application I view it like this:
+/- 2% is good
+/- 5% is acceptable
+/- 8% needs adjusting
So just a plain and simple sweep of the log looking at LTFTs will give us an indication of how our fueling looks since its had a time to acquire some averages. Like I said, STFTs change entirely too much to be accurate at ANYTHING other than idle where the car Is stationary with no throttle application, this is why we're looking at the long terms now.
Here we can see my LTFTs look to be about .99 and 1 respectively (at this specific point, you cannot assume the whole log has these same values!). This means my computer is multiplying the fueling in one bank by .99 (removing 1%) and not making any changes to the other. This is what I meant earlier when I gave the +/- percentages I like to see.
.98 to 1.02 is good
.95 to 1.05 is acceptable
.92 to 1.08 is really getting unacceptable and needs to be adjusted. The MAF curve will change as the weather changes however, you will never get a perfect 0. Just something worth keeping an eye on.
We can also check out our measured Lambda in this shot, which is measured as a ratio to stoich. The Stoichiometric relationship between gasoline and air in our case is 14.7:1, so that's "1" as far as our relationship is concerned. So we can see my Lambda is right about .82, or 82% of 14.7 which means my AFR is right about a 12.0 AFR.
Really as far as 95% of us our concerned, that's enough to check and insure that our car is running well. Are we knocking? And how does our fueling look? Another thing we can do to measure airflow and power potential is measure gains in MAF flow and MAF frequency. Now, this is basically like a drag strip or dyno, you can really only measure against yourself.
You cant look at someone with similar mods and say "wow, my max Lb/min of flow with the same mods is like 2x yours, your car isn't running right!" There are way too many factors to accurately say that. What is their altitude? What is the weather there like? Is it cold, hot, humid... Too many factors, so lets try to break it down.
Much like a dyno, we are looking for changes against ourselves. The MAF frequency is the measured airflow coming into the engine. There are several Frequency breakpoints all along the MAF curve (again with interpolation) with a specific fueling target attached to each one. If my MAF curve is even slightly different than yours, my MAF flow value will be different than yours at any given point, and it can look like one of us is flowing substantially more than the other when that is simply not the case.
So again, what we are looking for is increase in MAF flow. This is very difficult to measure due to changes in weather, and all manner of other things, but its something I believe worth looking at.
Lets again maybe just break it down into 1000 rpm increments. Lets say the most MAF flow you saw before your CAI was 40 lb/min in the 5000 rpm range. After the addition of the magical CAI you now see 50 lb/min max in the 5000 rpm range! What does this mean? In a nutshell, we are flowing more air, and making more power! Now, its not quite that simple lol, but looking for changes in flow can be an indication that a mod is doing something, and we'll cross that bridge when we come to it. For now, I hope everyone enjoys poking through their datalogs. These SCTs log a TON of data in every pull.