Summary

Once you have your air fuel ratio dialled in on the dyno, it’s still important to confirm that your air fuel ratio is correct out on the road or the race track. In this webinar we’ll look at how we can use MoTeC’s i2 data analysis package to help confirm and optimise the air fuel ratio using logged data.

Transcript

It's Andre from High Performance Academy. Welcome to this webinar, where we're going to be looking at how we can use data analysis in order to help optimize and analyze our fuel mixtures. Now, predominantly here we are going to be looking at using MoTeC's i2 data analysis package, although we're also going to be having a quick look at the MegaLogViewer HD a little bit further through the webinar. I'm just going to give you one example of how we can use that particular package to help us with our tuning. Of course, as usual, a lot of the concepts that we're going to be dealing with here will be applicable to other brands.

It's really the concepts that we're going to be looking at that really are the most important. As always, we will have an opportunity for questions and answers at the end of the webinar. So if there's anything that I discussed during the webinar that you'd like me to discuss in more detail, or anything related to the topic at all, please ask those in the chat, and I'll deal with them at the end of the webinar. So I think we probably need to start by discussing why we want to analyze our air-fuel ratio or lambda data in the first place. Now, if we've done our job properly on the dyno, then surely our air-fuel ratio should be okay once we get off the dyno, head out into the real world, be that the racetrack or simply driving on the road.

The reality is that that's often not the case. And as I discuss in our practical standalone tuning course, even with a quality dyno for our dyno tuning process, and if that's fitted into a well-equipped, well-designed dyno cell, we still can see discrepancies between our air-fuel ratio while we're on the dyno computer, what we see out in the real world. It's really difficult for us to replicate those real-world conditions perfectly in a dyno cell, particularly if we haven't got a budget of several million dollars like OEM manufacturers may have. It's very difficult to replicate the air flow and temperatures that we will see when the car is out on the racetrack, particularly in a high gear and at high speed. The conditions that the engine is exposed to can vary quite significantly from what we have on the dyno.

So the situation is that obviously since we aren't driving or racing our cars on the dyno, what we are most interested in is ensuring that our air-fuel ratio is correct on the road or racetrack where the car is going to be used. Now I am solely focusing here on our air-fuel ratio. Of course boost pressure and ignition timing are just as important. We need to really be monitoring all of those aspects to ensure everything is sitting within the realms of what we're expecting to see. But again, for our webinar today, we are focusing predominantly on our air-fuel ratio.

That's what we're going to be looking at. So the easy way of analyzing this data is by looking at what the air-fuel ratio is doing out in the real world. Now, we can fit a wide-band air-fuel ratio meter, we could maybe suction cup that to the inside of our windscreen, and we can view that while we're driving around. However, that creates a problem insomuch as we've got a large, three-dimensional table for our fuel log volume efficiency table, and depending on what our current manifold pressure and throttle position, sorry, and engine speed might be, obviously this puts up in a different cell in our efficiency table, so it's really difficult to visually correlate the data that we're getting from our wide-band meter and decide where exactly the air-fuel ratio may have been rich or may have been lean, and in that case where we need to make our changes when we've got the car stopped. So the obvious solution to this is that we log the values from our air-fuel ratio meter, our wide-band controller, into the ECU.

That way we can correlate that data directly in the logging software and see exactly what was happening at a wide range of data points, both manifold pressure as well as engine speed. So once we've got this though, once we've got a log data file with our air-fuel ratio information, the next problem we have is it creates a lot of data. There's a lot of points to look at. And it can be quite difficult with a large log file to decide exactly what to focus on, what's important, what's not, and then of course what changes do we need to make based on the data that we've got in that log file. It really can be quite difficult when we're focusing solely on air-fuel ratio to pinpoint a problem and decide what to do about it.

We also are inevitably going to see some fluctuation in our air-fuel ratio data, which is really important to understand. Particularly novice tuners, when you get started, you like to think that lambda or air-fuel ratio is one fixed number. If we have 0.80 lambda at 4,000 rpm and 200 kpa of manifold pressure, then that's what we should have, no matter what gear we're in, no matter whether we're going uphill or downhill, what the air temperature is. The reality is a little different, of course. And our lambda data is always going to move around a little bit.

So it's not uncommon, for example, to do four or five pulls through the rev range through a fixed gear, let's say fourth gear, and find that our lambda value varies perhaps 0.01 to 0.02 across those runs. So that's another aspect we need to take into account. Basing a decision on a single data point when we're looking at a log file is often futile and can also be quite misleading. So we need to know how we can take account of that. So let's get to the start of it.

What are we going to look for first? And it's always a good idea to have a bit of a step by step process or guide to go through when you're looking through a data log file. Of course, this particular webinar really backs onto some of other data analysis webinars, so you can also view those to get an overview of what other aspects you should be looking for. So if I was to look at the key aspect that I'm going to be most interested in when I'm viewing solely the air-fuel ratio data, what I first of all want to be very sure of is that the air-fuel ratio is safely rich when I'm under the most dangerous conditions, which would be high load and also moderate to high rpm. That's the first place I want to focus my energy. And the reason I want to do this is if we are potentially lean under those conditions, nothing else really matters because we may be in a position where we're going to end up damaging the engine.

So that's the first place to look, make sure that our engine is going to be safe and it's actually going to hold together. Once we know that our air-fuel ratio is at least safe, we can start being a little bit more specific, and we can look at how close we are to our desired or our target air-fuel ratio. Now once we've done that, once we know that the engine is safe, we know where our air-fuel ratio is compared to our target, then of course we can go ahead and make some changes to our efficiency of fuel table based on that data in order to make some corrections. Now the problem with doing this though, as I've kind of already touched on, is it's difficult or can be difficult to find trends when we're looking at a large sample of data. So we need to understand some techniques that we can go through in order to do that.

Let's start by jumping across to our i2 software. And for this example I am using MoTeC's i2 Pro. This is some lap data from the same racecar that we looked at previously for some brake data. This is a lap around Manfeild Racetrack in the North Island of New Zealand, so the engine here is a naturally aspirated five liter V8. Now what I've done here is I've created this particular worksheet, which is known as a rainbow track map.

I've named it lambda. And this is a really easy way of getting a really quick overview of where our air-fuel ratio is. Now if we just jump back for a moment and we have a look at our, let's go through to our mixture, here we go, let's have a look at a larger chunk of data here. So this is a range of laps that we're now just looking at from that same car. At the top of the screen here, we've got our engine rpm, then we've got our throttle position, we've got our manifold pressure, which in this case, being naturally aspirated, isn't too much of a concern for us, and then we've got two lambda inputs coming from MoTeC LTCs.

So we're looking at this data. This is difficult to really make much sense of. There's large spikes in the lambda from gear change ignition cut, and really it's difficult to really highlight anything particularly useful. So remembering one of the first things I just said was we want to make sure that the engine is safe under high load sustained, wide open throttle operation. So just zoomed in now, so we're just solely looking at one lap of data.

Obviously as I do this, everything starts to look a little bit more sensible. But even so, it's still quite difficult to really pull too much from this data in the way of trends. In terms of looking at that wide open throttle condition, what I'm going to always do is straight away look at the areas of sustained wide open throttle. So we can see if we look at our throttle, we've got quite a few long straights on this particular track. And this is the easiest place to start when we're analyzing our data.

We can simply zoom into an area where we're under sustained, wide open throttle. You can see, this is a sequential gearbox, so we have flat shifting. The driver is not lifting off the throttle for the gear shifts. And we can look at our data. So straight away, we've got both LTC1 left bank and LTC2 right bank, and we can see that data right through this time graph.

So we can get a good indication that we're probably pretty safe there. I bet she put in some bounds here on the i2 time graph. We can just click on a single spot, and we can see for that particular point 6400 rpm. It's only at 0.88 lambda on the left bank, 0.89 lambda on the right bank. And I'll just show you as well, if you want to add the red and blue lines, which we can see here, just to give us a quick visual guide, we press F5 to bring up our time graph properties and we double click on the group five, which contains our lambda data.

We can see down the bottom we have the option of our scale lines. So by default these are switched off. However, what I've done is I've enabled a minimum line at 0.88, and I've enabled a maximum line at 0.92. You can also choose some colors for these. And it's just a quick graphical guide to know, are we within those bounds? Yes, we are, we know we're safe.

Okay, we can carry on. At the same time as well, it's also worth mentioning, quite often depending on the magnitude of the values that you're getting from the entire data set for your lambda, you may find that the scaling is initially pretty much useless. I'll give you a quick example of that. It may be that we end up with something that looks a little bit like this. So at the moment, our lambda data is scaled between zero and five, so really difficult to get anything particularly useful out of there, makes everything look really nice and flat, but of course that's not our concern here.

So if that's the situation you've got, you can just double click and you can change your mode from auto scaling to manual scaling. And then you can set a sensible minimum and maximum. So in this case I'm just going to go between 0.7 and lambda one. So you can see that gives us a lot more detail and exactly what's going on. So I'll just zoom out a little bit further because I wanted to also talk about the lean spikes that we're seeing here, which may be a little bit scary to some of you.

So we can see here, in particular, there's quite a large lean spike on this gear shift. Now, as I've said, that might sound a little bit scary. We'll also got another one here as we shift into sixth gear. Those lean spikes are not real. These are a result of the gear change ignition cut.

So what's happening here is that the ECU, when the driver's pulling back on the gear lever, initiating a shift, the ECU is cutting the ignition to allow the gearbox dogs to disengage and the next gear to be pulled through without lifting off the throttle. So what this does is it creates a situation where cylinders full of unburnt fuel and air are passing through the exhaust system. The lambda sensors detect the oxygen passing through the exhaust system, and that's what gives that lean spike. So we really want to ignore that lean spike. This is not a situation that we need to worry about.

I mentioned before it's not real. Well, it kind of is real, there is unburnt oxygen, unburnt fuel and oxygen passing those lambda sensors. But in terms of what we need to be worried about, it is not a lean condition. Let's zoom back out so we're looking in the whole of that lap two data again. Now we'll just jump across because I started by viewing this rainbow track map, and this, as I've said, is a really good way of just getting a quick graphical glimpse of your data and deciding if the engine is going to be safe.

What we've done here, if we right click and go to properties or press F5, what I've done is I've set up a color channel for this rainbow map. At the moment I'm just solely looking at lambda one, LTC1. And we've set a manual scaling there between 0.8 and lambda one. So what the ECU will do here is break down the data for that entire lap into bands based on what we select here. I've actually selected also that we'll specify a bandwidth of 0.025 lambda.

So how that works, if we look over here at the bottom, it breaks all of our data down into different color bands based on what the lambda value was at that particular point on the track. What I've done is I've selected manual coloring here. Now under wide open throttle, this is all I'm trying to analyze at the moment, I really want to be in the range of 0.85, 0.88 through to 0.90. So again, I'm not trying to be very accurate with this data, but it's a quick visual guide as to whether or not the engine is going to be safe. So you can see there I've selected the color band, the range between 0.875 and 0.90 lambda to be green.

As we move leaner, we start going yellow, then orange, then orange, and then red. As we go richer, obviously a safe condition, we start going to lighter blues and then darker blues. So once we've got this set up, we can see straight away the result of that. So what, again, just like when we're looking at our time graphs, we're going to be concentrating on the straight sections of the track. These are the areas where the driver is on open throttle, so these are the areas that we're most interested in what our lambda or your air-fuel ratio is.

Any time we're under sustained, wide open throttle operation, we want to be making sure that our air-fuel ratio is safe. And what we're looking for here is, ideally, we should be in the green range, which for the most part you can see we are. If I start seeing those colors under wide open throttle move to red, then obviously that's a concern. And we can start looking in a little bit more detail. Do need to be careful when you are viewing this, obviously, we get some strange things happening towards the end of the stripes.

The track direction is clockwise on this particular track. And these areas here, where we've got these darker colors and also some lean spots indicated by red, that's where the driver is under braking, so we can't really take too much notice of those, so this is just a really quick visual way of seeing if we're in the ballpark. We're not trying to be too specific. And in the i2 software, what we can also do is use the N and P keys so we can press N for next and we can just cycle through all of the laps. P will take us back a lap.

And that just lets us get a really good overview of what's going on as we go from lap to lap. So once we've that overview and we know that our air-fuel ratio is safe, we can start looking at how we can make some specific changes, how we can get an idea of what specific changes we need to make. Now there's a few ways we can go about doing this. One of the ways that we can get a better idea of what's happening with our air-fuel ratio data is to use the mixture maps in the i2 tuning, the i2 analysis software. So on the mixture map screen now, and we'll just quickly go through how to use the mixture map.

Now the reason we're going to use the mixture map, this is probably a little bit more relevant when we're looking at a large log file from a road car, is that we're not just going to be interested in tuning a wide open throttle, particularly if we've got a turbo charged engine as well, we may be at wide open throttle, we may be running various boost points. We're also really interested in obviously, for a road car, what's happening when the car is in the cruise areas. So how we're going to use this feature will be a little bit dependent on exactly what we're going to be tuning and what we're trying to find. So let's have a look at the setup of this first. We right click on this and click properties again.

We can go to the setup of our mixture map. You can see we've got the two channels here that will be our color channels, our inputs there, lambda one and lambda two. We've also got a load channel, so in this case because it's a naturally aspirated engine and we're also using throttle position or alpha N as our load input, we're using throttle position as our color channel, sorry, our load axis, which becomes how the data is separated into colors. Again, this is able to be set up however we want. We've selected a manual scaling here and our throttle position points will be between zero and 100.

And we've also got 11 bands between that zero and 100. The colors are separated here, so we go from closed throttle, where we're blue, through to red, where we're at wide open throttle. Now if we go back to our channels as well, we'll see down the bottom, in the middle for a start, we've got our X axis input, which is our engine rpm. Again, we can choose some sensible minimum and maximum values for that data. In this case, we're going for 2,000 and 8,000 rpm, although because this is out on a racetrack, obviously we're running with a sequential gearbox over a relatively narrow rev range.

Moving down, we've also got some settings here for our lambda. We can set a bit of a delay. There's sort of, depending on the engine rpm, the position of the lambda sensor, and the exhaust system, there's going to be some amount of transit time between the air-fuel ratio actually leaving the exhaust ports and actually being metered by the lambda sensor. So we've got the ability to set a delay in there. We can also exclude lambda values that are richer than a specific value.

In this case, you can see that set to 0.5. And on the right hand side, we also have the ability to employ some filtering for gear shifts. So what's going to happen here is that data is going to be ignored when the rpm either rises or drops by more than two and a half thousand rpm per second. So it's just really is indicative that the engine is being, it's going through a gear shift rather than normal acceleration. So with that all set up, we can go back and have a look at our data.

Now again, because this is a racecar, really we're most interested in what's happening under wide open throttle conditions here. What we can do with our mixture map though is select the color ranges that we're interested in viewing. And we can do that simply by clicking on the relevant areas that we're interested in, so let's just have a look now. What we'll have a look at is all of the data points that we ended up with from 70% throttle and above. And that's what we've got on our screen right now.

So what the MoTeC software does is also gives us a trend line, which we can sort of see running through that data. This gives us a bit of an indication of all of those data points, what we've actually got going on when it's all averaged out. And this really comes back to what I was saying that if we're looking at a large amount of data, there's always going to be some amount of discrepancy from one event to another. So we're never going to get exactly the same air-fuel ratio data every time we go through, let's say 6,000 rpm at wide open throttle in fourth gear. We may see a very small variation.

And that trend line helps sort of get rid of that. We can also see some outliers on the mixture map. We've got them on both channels here. Those outliers again, looking at the rpm where that's happened, probably are most likely to be related to a gear shift ignition cut. Again, we need to understand the data and how to read it.

If we're seeing something like this particularly happen right up near the rev limiter or if we are using a sequential gearbox where the gear change cut, then it's important to understand that that's what causing that and we don't need to try and take that into account. So just looking at this data here, first of all what we can see is that we do have, if we look purely at our left bank for a start, we've got a reasonably tight spread vertically of our lambda, which is great. That makes we've got a reasonably consistent lambda value through all of our data that we've analyzed. We've got a slightly wider spread on this right bank here in comparison. And this is one of the aspects we do need to look at, regardless whether we're using the mixture map, which we're looking at here, or we've come back and we're looking at our mixture using the time graph.

It's important to look and see if we've got any large variations in our lambda. So a good example of this would be if we're looking at data from a racecar, car goes out of the pits and for the first two laps the air-fuel ratio lands exactly on our target, and then after four or five laps we start seeing the lambda data start to move richer and richer. So at the start of the, start of the laps we may have been right on our target and by the end of 10 laps, for example, we may find that we're six or eight percent richer than our target. Now if we're looking at that sort of variation, it's really important to pick that up. If we're seeing a variation across multiple laps in our lambda, then we want to be looking a little bit deeper at what may be causing that.

And what we really want to be doing there is looking for any aspects in the data that could affect the air density or the state of tune of the engine. Good examples there might be intake air temperature. Is the intake air temperature either rising or perhaps falling through the course of 10 laps? If that's the case and everything else is staying relatively fixed, then it may be that our intake air temperature compensation needs a little bit of work. Likewise, engine coolant temperature, what's going on there? Although engine coolant temperature seems to have a lesser effect on our air-fuel ratio, it still can have an effect. Of course fuel pressure is another good one to look at.

Is our fuel pressure staying fixed? So all of these things we want to have a wider look rather than just focusing solely on our air-fuel ratio data. What other aspects could influence the data that we're seeing and give us the kind of trends that we're seeing there? And when we're looking at the mixture map, if we just pop back to that, if we're seeing, obviously at the moment we can see that I'm only focusing on one single lap. But this particular graph is zoomed linked. So we can look at all of our data across all of our laps. And if we're seeing a trend where the air-fuel ratio is starting to move richer or leaner as lap after lap goes by, we're going to see a wider spread across our lambda and compared to what we're seeing on just one lap.

We've got a little bit of that going on there, but of course I just also included our out laps where the engine may also be suffering from some heat soak as well. So it's really important to not keep your blinders on. Look at all of the data, understand how all of the data relates, how all of that sort of can impact on the way the engine's running. The last one there is also our barometric air pressure. Now obviously that's hopefully not something that's going to be an influence from one lap to the next.

We're obviously going to be at a relatively fixed barometric pressure when we're at the same place. But particularly if you're used to racing at sea level and perhaps you go to an event where you're at a significant elevation and the barometric air pressure there can influence your air-fuel ratio data. Okay, so we've got a couple of options now. We've looked at a couple of techniques of analyzing the data. And my personal preference here is to use this mixture map, just because it combines all of our data points, much easier to make, to get a more realistic idea of the average air-fuel ratio or lambda that the engine's running at at a fixed rpm and throttle position or manifold pressure if we're using the mixture map rather than going back and looking at our time graph and just trying to analyze one particular point on one part of the track and make an educated decision based on that single data point as to what the engine's doing lap after lap.

So the mixture map is a really, really powerful option there. Now I'll just mention as well, I really, I touched on it briefly, but I'll go back to it. If we press F5 to bring up our mixture map properties and go to our load channel, as I mentioned, currently I am using throttle position here because the engine is tuned on alpha N. That's really important to make sure that the load axis we are using matches whatever the load axis is in our efficiency or fuel table. If it doesn't, it's gonna be really difficult to make useful decisions on where to make our changes.

So if you were running a conventional engine using manifold pressure as the load axis, simply click here and we can, in this case the channel is, no, it's, manifold pressure. And we can go through the same process and set up some break points that will suit a manifold pressure based system. Okay, so we've got our data. We know we're rich or we know we're lean, and we know whereabouts exactly that's happening. How can we use this data in order to make our changes? Well, there are a few options there.

Now there are some specific options for MoTeC that we'll look at, bu we'll also look at some broader terms as well. So for example, let's look at our data here from lambda one. And let's say for example that our target at 6,000 rpm was 0.90 lambda. Well we can see here, so if you look in the middle of our trend line, I'm sorry that's, my line is red here as well, we can see that we're actually running probably closer to 0.88 lambda. So this means we're a little bit richer than our target.

Just try and draw a line through there so we can see. Yeah, so we're a little bit richer than our target. So if we can think back to our EFI tuning fundamentals course, one of the key equations that we learned there is how to make a correction based on our measured air-fuel ratio or lambda. So what we used there is the air-fuel ratio that we've measured, so our measured air-fuel ratio divided by our desired. So measured over desired, in this case 0.88 divided by 0.90, it'll give us a correction factor to apply at that point in our efficiency table.

Essentially in this case, what we want to do is remove around about 2% fuel. Multiply the value in our efficiency table at 6,000 rpm, and in this case wide open throttle, by 0.977, 0.98 just to round it off, and that should give us the air-fuel ratio lambda value that we want. So makes it nice and easy. Now if we are using MoTeC software, then MoTeC will also incorporate the lambda was function. There's a couple of ways you can use this.

Let's just jump over to the M8 100 or 100 series ECU manager software. I've just got a sample start map configured here. And let's say we were at 6,000 rpm. Obviously the numbers in this table are currently just garbage. We've got an efficiency of 50% in there.

But that doesn't really matter. So to use lambda was, we want to press the L key on our keyboard. And it brings up this box that says Lambda Was. Now, what we wanna do here is just enter the lambda that we averaged in our mixture map, so in this case 0.88. Now you'll see that just below this, it says LA Table.

What this is doing is taking the value from our lambda target table, and at that particular point 100% efficiency in 6,000 rpm. Our lambda target table is currently at 0.9. But what that's going to do when we press OK is it's going to make exactly that same change. You see that it has taken the value from 50% down to 48.8%, so it's made that same mathematical change for us. Just takes the requirement to use a calculator out, makes it really nice and easy.

So we can just go through using the data from our mixture map, and we can just go through and we can start say here, at 5,000 rpm, look at our values, and we can make adjustments right the way through the range based on what that data from our log file is showing us. Now if we're using MoTeC's i2 software, the technique is essentially much the same. We can press L, it's going to bring up our lambda was. In this case, we can enter both our fuel mixture aim as well as our measured air-fuel ratio, measured lambda, and it's going to make the, again, the required change for us in our efficiency table. So a couple of nice ways there if you're using MoTeC software to kind of automate or speed up that process of making tune changes to your efficiency table based off the log data file.

Now the other way we can also go about analyzing our data is to create a math channel. If we jump across to, this is just a file I've got open in i2, and what I've got in this particular time graph, we can see we've got our engine up here, we've got our throttle position, we've got manifold pressure, and this is coming from a MoTeC M1, where we are logging lambda one as well as our fuel mixture aim. So here we can actually see how our measured air-fuel ratio lambda is deviating from our target. So what I've done is I've created a math channel, which we can see down the bottom, which I've called Mixture Error. To do that we can click on the little maths icon in the toolbar.

And we can add an expression, in this case it's a really simple one. All I'm doing is looking at lambda one minus our fuel mixture aim. And that will generate this particular channel, which we're now displaying. So what we do based on this channel here is we can see, let's have a look at an area where we do have a small error. This is pretty close, so there's not a lot of work to do.

You can see at this particular point, let's try and find something with a little bit more error. Okay, so in this point here, our fuel mixture aim is 0.82 and our measured lambda is 0.83. So we can then, using our fuel mixture error, what this is showing us is essentially we've got a 1% error there. And the way I've got this configured, this is what we need to add to our efficiency of fuel table. So essentially at that particular point in our efficiency table, if I added 1% to that number in our efficiency table, that should correct that error.

So this is just another way that we can use i2's analysis software in order to generate a channel to help us speed up and fine tune our end result. Okay, we're going to move into questions and answers really shortly, so if you have anything you'd want me to discuss further, please ask those in the chat and I will go, do my best to answer them there. Now there's one function that I'm going to show you here. It's outside of MoTeC's i2 data analysis software. It's another software package that I have used from time to time, which is MegaLogViewer.

And this is from the same team that bring us the MegaSquirt range of ECUs. It's actually quite a powerful log analysis viewer, not just for the MegaView, sorry, for the MegaSquirt ECUs, but it's great log analysis software package for basically any .csv or common separated variables data file. Now the reason I'm going to show you this is one of the functions I really like in the MegaLogViewer is the ability to create a histogram based on the same break points as our fuel efficiency table and then fill that histogram in with data to help us speed up our tuning. This is really a lesson that I learned from tuning reflashing factory ECUs using the HP Tuner software. They had some very, very powerful histograms in the VCM suite scanning software, which makes it very, very fast and easy to optimize both a mass air flow sensor calibration as well as a speed density map using these histograms.

So I really replicated how that works using MegaLogViewer. So the first problem with this is if we are dealing with a MoTeC log file, we can't open the MoTeC log file directly in MegaLogViewer. What we need to do first of all is export the file as a .csv file. It's really easy to do. If we go to our file menu and we go down to export data, and you'll see under the export options we have output file format, and you can see at the moment I've just selected the .csv file.

We can now output that file, and it will be saved as a .csv file, which we can open straight into MegaLogViewer. We'll just jump to MegaLogViewer. I've got a file open already, which is actually that same lap two data from Manfeild that we've already been looking at. At the moment I've got the, essentially it's the time graph view, where we can see that we've got graphs, we've got engine rpm on our top graph, we've got throttle position on the second graph, and on the bottom graph I'm logging both lambda one and lambda two. So at this point really no useful difference from what we're seeing in MoTeC's i2 software.

However, the key point of difference, which I really like with MegaLogViewer, is if we go across to the right hand side, we have this histogram slash table generator function. Now we'll click on that. We'll click on that, and what we've got here is the ability to set up our axes. So in this case, what I've done is I've replicated the same axes that I've used for the efficiency table, again remembering we're on alpha N here. TPS is our load axis, so I've selected our Y axis as throttle position.

Our X axis is engine rpm. And in this situation, what I've done is I've just simply filled our actual data points with the values from lambda one. Now if you've got a data point, a parameter logged such as fuel mixture error, then you can use this directly into the histogram. But what it's going to do is basically populate this histogram here with the average values that we have logged while the engine has been running. Important points to note here is that we can adjust our table dimensions as well, both the number of rows and columns, and then we can change the individual break points.

So this makes it really easy. What we can do is go through this and we can adjust our break points on these tables, so both the rpm and throttle position match exactly what we are using in our actual fuel efficiency table. The important point there is if the break points match, it's really easy to take data points from this histogram and then go and apply them directly into your efficiency table. So here we can see we're logging our air-fuel ratio data. This currently is also colored, as we can see.

And the color is based on the hit weight, was what MegaLogViewer calling this. Essentially what it's doing is telling us how many samples we've got in each of these cells. In this case, the green color means that we've got more data points, so we can trust those values more closely. Really broadly, what we can do here, we can obviously move through the time graph that we've got on the bottom, and we'll see exactly, by the cursor, whereabouts in the table we're accessing at each particular point. The important point with this is if we're analyzing a large amount of data, the histogram's a really good way of filling in the entire table, getting a pretty good picture of what our air-fuel ratio is doing compared to what we want it to do, right through the entire engine operating envelope, or in other words our entire efficiency table.

And we can then look at any areas where the air-fuel ratio is dramatically different from what we actually want. We can go in and focus on those areas. Point to watch out for here though is that we can end up with some garbage points. There's no filtering being applied at the moment on this table, so we can end up with some garbage points like where we're on overrun fuel cut, for example, or with those points that we've already discussed, where we've got lean spikes caused by gear change ignition cuts. We do need to be a little bit careful and understand how to interpret and read this data.

We can then go and copy these changes that we need to make straight into our efficiency table, and we'll use that measured over desired calculation or correction in order to make our changes. Alright, we'll move into some questions and answers now, so again, if you do have anything else that's cropped up, please ask those and I will do my best to answer them. Our first question is, have you played with gated math channels on i2 Pro for analyzing lambda or do you find the functionality in i2 Standard sufficient for most cases? I'll be honest, no, I haven't used the gated math channels. I'll try and say that again. I've been using the, the gear shift that we've talked about, the gear change filtering, but I'm not gating the data on anything else.

And again, I think probably the key point here is to understand what you're looking at, understand how the data can be influenced by the likes of gear change ignition cut, by the likes of overrun fuel cut, and how that can influence the data that you're looking at. Really again, when I am personally analyzing data from a racecar, what I'm most interested in is making sure that the engine is going to be healthy under the worst possible conditions, which is those wide open throttle, sustained high rpm, high load operation. And we had a look at two ways we can analyze that data. So for me, I haven't personally found too much need to use gated math channels. But by all means, that functionality is there if you want to get a lot more advanced with it.

I don't believe that it will stop you doing a perfectly good job of analyzing data in i2 Standard, though. Next question is, how have I found the closed loop lambda control for track applications, specifically there in the MoTeC M1? I have found it absolutely excellent. And in our own Toyota 86 development car, I rely on the closed loop lambda everywhere right through the entire efficiency table. What I have done with that functionality is I have limited its range under higher load conditions, so typically I'll allow a little bit more variation down at idle and maybe light throttle cruise, where we can't see the influence of temperature changes et cetera. I'll make a little bit more of a difference to our actual error, our actual lambda, under wide open throttle conditions I limit, tend to limit the closed loop control to perhaps plus or minus 5%.

And if I'm outside of plus or minus 5%, probably means that I haven't done my job properly in the first place with the fuel table, and really that needs some more work. But it is exceptionally fast. And it really does do a great job of just making sure that any slight fluctuations day to day or venue to venue aren't going to have any impact on the air-fuel ratio that the engine is operating at. And you can also use the log data based on the closed loop lambda trims in order to help you with your tuning. Now that is a good point, though, that I've just touched on there.

If you are running closed loop lambda, again, just simply really need to have an understanding of the impact of all of these parameters on our tuning. It's quite often going to be the case where our actual air-fuel ratio ends up really nicely matching our target. Of course that's the purpose of closed loop air-fuel ratio control in the first place. So on the face of it, if our ratio matches our targets, our job's done, we don't need to look any deeper. If I am using closed loop fuel control, what I will also do is analyze the trim, the closed loop fuel trim channels and see what those are doing.

So if my air-fuel ratio matches my target but my trim's sitting at let's say positive 6% almost everywhere, then what that means is that our efficiency table is lean, we need to add some fuel, add some numbers to our efficiency table in order to get those trims close to zero. Long answer to a short question. But hopefully that was helpful. Next question comes from James, who's asked, can you explain how a target air-fuel ratio map interacts and influences the base fuel map during a dyno run or a ramp run? Yes, I can, but this is a complex answer, or complex question, probably more complex than it seems to be on face value, simply because everything here depends on whether we're talking about an injection time based fuel model or a volumetric efficiency based fuel model, two very different aspects that you need to understand. For a conventional, basic, millisecond based fuel injection model, then the closed loop, sorry, the lambda target fuel mixture aim map can have no interaction, no change on the fuel being delivered to the engine.

In that aspect it would act as a target for closed loop fuel control, act as a target for our own tuning. But physically, if you tune the engine to 0.90 lambda under wide open throttle at 6,000 rpm, and then you go into your fuel mixture aim table and you change your target to 0.8, nothing's going to happen. With a VE based fuel model, on the other hand, everything works very, very differently. And I really probably don't have the ability to get too deep into this, but the volumetric efficiency based fuel model like the MoTeC M1 uses is really integral with that fuel mixture aim map. It's critical to make sure that one of your first jobs is to set your fuel mixture aim to your actual desired air-fuel ratio targets, assuming you've got all of the other configuration aspects such as your injectors, your fuel characteristics, your engine capacity, et cetera set correctly, you then go about adjusting the volumetric efficiency table til your fuel mixture matches your target.

Now once you've done that, the efficiency table is deemed to be tuned correctly, and now if you wanted to make change to your fuel mixture aim, you can actually do this in the fuel mixture aim table. So an example, change your fuel mixture aim from 0.9 to 0.8, and the actual measured air-fuel ratio should match that new target. So again, long answer to a short question, but it is quite a complex question. Hopefully that's cleared that up for you. MCR's asked, can Link input GPS so I can log track runs? That's a good question.

I have not tried to do that, so I'm probably not the best person to answer that. My understanding is at the moment, no, that is not possible. But what I will do is confirm that with the guys at Link, and I'll answer that in the forum after this webinar has aired, just so I can give you accurate answers there. Barry has asked, on one of the example logs shown, one of the lambda channels seems significantly more choppy than the other. Are there any reasons for that? Yes, there are reasons for that.

I think if I remember back to what we were going through with that car, it had a pinhole crack in the header on the right manifold, and this was influencing the air-fuel ratio data that we were seeing. I will also mention though, when you are logging lambda data from two banks of a V configuration engine like the data we presented there, it's the opportunity to check your bank to bank balance. So what we're doing there is we can find that one bank of cylinders will be perhaps richer or leaner than the other bank. That's quite often the case. We will find that through one rpm range one bank is richer and then it may come back to being the same as the other bank.

And in order to fix that, this is a situation where we can use individual cylinder trimming, and we can adjust one bank or the other in order to even out those lambdas so that both banks are running broadly the same. Remember when we're looking at air-fuel ratio or lambda data, what we're actually getting is the average of all of the cylinders that the sensor is sampling. So in that case we're looking at lambda one is sampling all of the four cylinders on the left bank, lambda two is sampling all of the cylinders on the right bank. It's possible as well when we see a lean or a rich area that there may be affected predominately by one or maybe two cylinders. But of course without individual cylinder lambda, it's impossible to tell.

My own preference there is if I've got one bank of cylinders that's running leaner than the others, what I tend to do is add fuel to the lean bank. It's just my own personal preference. I prefer to add fuel to the lean bank in order to even out any discrepancy rather than removing fuel from the rich bank. Alright, that looks like it's taken us to the end of our questions. So hopefully the information in that webinar is going to help you do a better job of analyzing your engine's performance in terms of the fuel mixture on the track or the road using some data analysis software.

As usual, if you do have any further questions that crop up, please ask those in the forum and I'll answer them in there. Thanks all for joining us, and I look forward to seeing you all next week.