×

Sale ends todayGet 30% off any course (excluding packages)

Ends in --- --- ---

Practical Reflash Tuning: Step 4: MAF/Injector Scaling

Watch This Course

$229 USD

-OR-
Or 8 easy payments of only $28.63 Instant access. Easy checkout. No fees. Learn more
Course Access for Life
60 day money back guarantee

Step 4: MAF/Injector Scaling

30.01

00:00 - In the next step of the tuning process, we're going to address our mass airflow sensor scaling as well as our fuel injector scaling if that's applicable.
00:08 Now in this particular example, as we've already discussed, we're running the FA20 DIT engine with the stock injectors and there is no facility currently to add aftermarket direct injectors.
00:19 So the injector scaling is something we're not going to be touching, we're going to be focusing solely on our mass airflow sensor scaling.
00:26 Now we also have discussed that our off the shelf map from Cobb should theoretically give us a pretty accurate mass airflow sensor scaling straight off the shelf.
00:37 However we know that that's not always going to be the case and we've purposefully built in a little bit of error here so that we can show you the process of adjusting this.
00:46 This is going to be particularly applicable to any of you out there who aren't running the Cobb intake system and are working from one of their off the shelf maps as a starting point.
00:57 It's worth mentioning here that the technique that we are going to use for rescaling the mass airflow sensor is exactly the same as what we would use for adjusting our injector scaling as well.
01:07 So if you are running an EJ series engine and you're using this guide to help you with the tuning of an EJ engine, then if you have changed the injectors, you can use this information to help you there.
01:19 This is a good time to point out however that if you are intending to swap to an aftermarket intake system, as well as a larger set of injectors, then it's always advisable to do one of these changes at a time, scale in the mass airflow sensor, then swap your injectors and then any further errors need to be dealt with by your injector characterisation.
01:41 If you do both together and you've got short term and long term fuel trim errors, then it's impossible to know whether those errors are a result of inaccurate injector scaling or inaccurate mass airflow sensor scaling.
01:54 Let's jump back into our Accesstuner software and we finished up the last step of this module where we had saved our base calibration ready to flash into our ECU.
02:04 So we're going to do that now, what we'll do is we'll start by keying on our ECU, once the key is in the run position, we can go to our ECU dropdown menu and what we want to do is click on flash map.
02:17 Again there is a shortcut key that you can follow to do this.
02:21 What we want to do here is simply follow the on screen prompts until the flash process has completed.
02:29 So let's go ahead and do that now.
02:36 Once we've followed all of those on screen prompts and the flash process is complete.
02:39 we're at a point where we can start our engine and make sure that it's running properly, so let's do that now.
02:49 Now our engine's started up and it's idling comfortably, this is a good time to bring up our gauges using the control B shortcut.
02:57 And then we can view all of the parameters and see exactly what's going on.
03:01 Now it will take a little while before the short term and long term fuel trims start working so at the moment, we can see that those are both sitting at zero.
03:11 Likewise our measured air/fuel ratio is currently sitting at zero.
03:15 So what we're going to do before we make any changes here and start gathering any data is we're going to allow the engine to come up to normal operating temperature, at the moment we can see our coolant temperature is sitting at 56 degrees celcius.
03:29 So that's a little bit cool, we want to make sure before any changes or any data's gathered, we want to make sure that there is no heat soak going on and that our engine is operating normally so let's just allow our engine to come up to operating temperature.
03:43 Alright we've allowed the engine to come up to operating temperature now and we can get a better idea of what's going on with our trim.
03:49 So we can see here, we've got a little bit of activity from this trim, showing us that there is indeed some error with our mass airflow sensor scaling.
03:58 And how this occurs is the number on the left, our A/F correction #1, this is our short term fuel trim.
04:04 So this is an immediate response to any error in the air/fuel ratio between target and measured.
04:10 If we end up with a short term fuel trim active for a certain period of time, that will end up getting transferred across to our A/F learning #1.
04:19 So this is our long term fuel trim.
04:21 So essentially what happens is that the short term fuel trim is an immediate response to an air/fuel ratio error.
04:27 After a period of time that error gets baked into the long term so that the short term fuel trim ends up coming back to close to zero.
04:34 So essentially the total error is the difference or the addition of these two numbers so while our trims are moving around a little bit, we can see that our error is somewhere in the region of about minus two to minus three percent.
04:47 So it's not too bad but a little bit of work to be done there.
04:51 We can also see that our measured air/fuel ratio or lambda is siting at 1.0 and our target of course is 1.0 so everything's doing what it should be doing there.
05:01 We've got our coolant temperature sitting at 81°C.
05:05 So our first task here is to gather some data to help us with rescaling that mass airflow sensor.
05:11 And the information that we need in order to do that, we've already chosen from our logging parameters but we essentially need our short term and our long term fuel trims and we need our mass airflow sensor voltage.
05:21 If you wanted to get really fussy you could also log into those parameters there, the measured airfuel ratio and our target and take into account any error there.
05:30 Turns out that the closed loop control strategy in the Subaru WRX is relatively quick though so you're going to get a pretty good result just using the combination of the short term and long term fuel trims.
05:42 So what I'm going to do now is get our engine up and running here on our Dynajet dyno, we are using a load controlled dyno with an eddy current brake.
05:51 And what we're going to do is just set this to hold a relatively low RPM and we want to get our engine up and running and we're going to hold as little throttle as we can.
06:00 And we're going to gradually increase our throttle opening, essentially 'til we're all the way at wide open throttle and we want to do this smoothly and we want to do this slowly.
06:11 The reason we need to do this smoothly and slowly is to make sure that we don't end up bringing in any transient enrichment that could affect our results.
06:19 Now traditionally I would normally do this in two ways, we'd do this under closed loop conditions where we are targeting 14.7:1 stoichiometric air/fuel ratio or lambda one and then we're going to repeat that process under open loop conditions.
06:33 Again here because we are running continuously under closed loop conditions, that's not strictly necessary and as long as we're smooth and slow with our throttle input, the ECU will be able to really accurately and quickly track those changes.
06:47 So what we're going to do now is get ourselves into third gear which is the gear we're going to be using for our tuning and we're going to start off with as little load and as little throttle as we can and then we're going to start our datalogger.
07:00 So let's just get ourselves into third gear first.
07:07 Alright so we're in third gear here, we're barely touching the throttle, we're not on the eddy current brake at the moment, the rollers are just spinning.
07:15 And what we're going to do is start our datalogging.
07:18 Now we can do this by going to the ECU menu and clicking on datalogging or alternatively use the control D hot key.
07:25 So right now just while I am rolling along with essentially no throttle, we can see we've got about a -6% trim in our long term fuel trim.
07:35 Adding onto that we've got about a minus one to two percent short term fuel trim.
07:39 So again definitely showing there's some work to do.
07:42 So as I'm talking here I'm just gradually opening the throttle a little bit.
07:45 We've got our set point on our eddy current brake at about 2200 RPM.
07:50 So we'll just come up on that and then once the dyno has clamped and is holding the RPM steady, we're going to slowly but surely open our throttle.
08:01 So let's go ahead and do that now, just making sure all of our throttle inputs are nice and smooth.
08:13 Alright so at this point we are on the brake on the dyno, the dyno is applying load now to control our engine RPM.
08:19 You can see we're actually sitting at just under 2000 RPM and I try and do this initial test here at relatively low RPM just so that if we do have any inconsistencies in our air/fuel ratio, we're not likely to do any damage to the engine.
08:34 However again, the closed loop control here is so good that it's unlikely to end up with any error from our target.
08:42 So now that we're on the dyno control set point, we're just going to again, smoothly open that throttle and once we get to wide open throttle I'm just going to simply use ctrl D again to close that datalog file.
08:54 We want to do this before we back off the throttle so that our results aren't influenced by anything when the engine is slowing down.
09:01 So let's go ahead and do that now.
09:14 With our first test complete there and a datalog file saved, we can now have a look at our data.
09:21 Now in this instance with the Cobb Accesstuner software, we don't really have an integrated built in complete data viewing software package. And I find the easiest solution here is to use MegaLogViewer HD which is able to read any .CSV file.
09:37 It's a very powerful datalog viewer and allows us to manipulate the data to suit what we're trying to find.
09:43 So let's have a look at our laptop software here, we've got that file opened up and I'll just show you what we're looking at here.
09:50 So in our top set of graphs in white here at the top we've got our engine RPM.
09:54 So we can see that our engine RPM is relatively consistent because of the break on the dyno there sitting, you can see at about 1940 RPM.
10:04 We've also got in that same graph data there, we've also got our throttle position which is in green, so you can see particularly right at the end of the run there, gradually we've opened that throttle up.
10:18 We've also got our manifold pressure in yellow here and our requested torque, not really relevant for our current task.
10:25 Then our next set of graphs here we've got our A/F correction #1 which is our short term fuel trim and in green we've got our long term fuel trim so we can see what they're doing through this file.
10:37 So for example at this particular point here, we've got our short term fuel trim at -0.78, our long term fuel trim at about -6.25.
10:47 So essentially a total trim there of around about -7%.
10:51 Now this is where the problem comes in because using this data to make changes can be quite tricky.
10:58 Trying to figure out exactly what the sum of those trims is doing at any particular point.
11:03 So what I'm using here in the next set of graphs is our MAF error.
11:07 So this is a calculated field that we can make up.
11:11 We can do this, if we go to our calculated fields tab here and we can go down to our custom fields and we can see we've got a few added in there already.
11:21 If you're starting from scratch here, we'd just simply click on add custom field.
11:24 However we've already added that field in here and if we look at our formula, we can see that this is A/F correction #1, our short term fuel trim, plus our A/F learning #1 which is our long term fuel trim.
11:35 Pretty simple formula there.
11:37 Now if you were to add that in yourself, just so you can see how it's done.
11:40 You can give the calculation a name here, the custom user channel a name here.
11:45 And then to add the formula you simply need to follow the instructions which is to press ctrl + space, that will give you a full list of all of the available channels.
11:55 Double click on the one you're interested in, in this case A/F correction #1 and that will add that parameter or channel in.
12:02 In this case, because we're going to add them together, we want to use the plus symbol and then if we press control space again, we can bring in our second channel there.
12:12 So I'm not going to do this because we've obviously already got it added in there.
12:15 So we can see what that channel is doing and also in white here we've got our mass airflow sensor voltage so we know where abouts in our MAF scaling we need to make any changes.
12:27 Now lastly at the bottom here we've got our air/fuel ratio from our wideband sensor that is in green and we've got our target in yellow so we can see how closely those are tracking right through this log file.
12:40 Again, just showing how well that closed loop mode is working.
12:44 Now in the perfect world, when it comes to gathering data like this and optimising our mass airflow sensor scaling, my personal preference is to use a histogram.
12:53 And we can do this using the histogram table generator tab which we'll find over here on the right hand side of MegaLogViewer HD.
13:00 However because there are 60 points in the mass airflow sensor scaling table, it's going to be a very time consuming process to resize this particular table and then add the break points which we can see here, this is currently set up for RPM.
13:16 We'd have to add the voltage break points manually for each of those points.
13:20 So rather than do that, while yes it is the superior method, I know that not a lot of tuners are going to go to that trouble.
13:27 So what we're going to do is use our log viewer instead and we're going to use a little bit of common sense and apply that to our mass airflow sensor scaling.
13:35 So what we want to do is basically look for trends in this data and apply those trends to our MAF scaling.
13:42 So we can see here right at the very start of our log file, we were down, it's a bit hard to see here but we were about 1.46 volts out of our mass airflow sensor.
13:52 And we can see that we're around about -7% total trim.
13:56 So you can see that up until around about 1.7, 1.8 volts, we are at about that same trim.
14:06 So what I'm looking for here, rather than trying to be really specific for this first iteration, we're just going to go through and make some bulk changes.
14:13 What we'll do is we'll head back across to our Accesstuner Pro and before we make any changes here, we do want to just pay some attention to what our closed loop fuel trims are doing here at idle.
14:23 So while you can see they are moving around, right now we're sitting at around about -3%.
14:28 We also need to take note of what our mass airflow sensor voltage is.
14:32 We can see that's sitting at around about 0.90 volts.
14:35 Now it can be a little bit tricky at a glance to see where abouts we're operating in any of these tables or where to make these changes.
14:41 So what we can do is go to our view dropdown menu and if you click on live tracking, or ctrl t, that will bring up a little blue box that moves around as we move through the table so making it really easy to decide where exactly to make your changes.
14:57 Now in this case we are going to make a change for a start up to just beyond our idle position so let's take a position here of 1.04 volts.
15:07 We're going to highlight all of the numbers to the left of that including the area we're at at idle and what we're going to do is press the M key which stands for multiply, that's going to bring up a box that allows us to make a change, we're going to take 3% away from all of these cells.
15:22 Now to do that we're going to enter a value of 0.97, this is a floating point value.
15:28 So numbers higher than one we'll add or multiply by a positive number.
15:33 Numbers smaller than one, we'll multiply by a negative number.
15:36 So 0.97 will take away 3% and what we'll do is just wait for that change to take place and straight away we can see now our short term fuel trim has gone positive.
15:47 We ca see our long term fuel trim responding to that and within a few seconds we can see that our short term and long term fuel trims now essentially sitting right on zero where we want them to be.
15:57 When we are making changes to our mass airflow sensor scaling, we want to make sure that we make relatively smooth changes.
16:04 So this is why I'm not going to necessarily go through and make exact changes to every single cell as shown by our logging.
16:10 What we're really aiming for here is our short term and long term fuel trims combine to be no more than about plus or minus 5% maximum.
16:18 Although ideally we should be in the range of plus or minus about 2-3%.
16:23 So some variation is tolerable here, we're never going to get perfectly repeatable results but what we do want to do is make sure that our mass airflow sensor curve is smooth.
16:33 So we've made that change now up to 1.04 volts and we looked at our logging up to about 1.8 volts so what I'm going to do is now highlight the numbers out to 1.84 volts.
16:44 I'm going to make a smaller change than our logging actually called for.
16:49 For our first iteration of changes what we're going to do is multiply that by 0.95.
16:55 So this is going to take 5% away from this.
16:58 So again this is an iterative process, we're not likely to get this perfect in one shot.
17:03 Let's head back across to our logging now.
17:05 So we've made our change up to approximately 1.8 volts and what we want to do now is have a look at the trends above 1.8 volts.
17:11 So we're still at around about -5% and we look at what happens here, we can see that within reason, other than this little bump here, we are still staying at around about -5% right up to the point where we're at two volts.
17:27 So again I'm not trying to split hairs here and get everything perfect, what I'm going to do is select up to this point here, 2.1 volts and we're going to take out a further 5% so let's head back to our Accesstuner and do that.
17:44 With that change made, let's head back to our logging and we'll have a look at the rest of our log.
17:49 So here we can see our voltage from our MAF sensor is starting to climb as we increase our throttle position and again we're trying to just look at a general trend.
17:57 So what we can actually see is that our trims in this position aren't too bad, we're sitting probably between about -2% and around about 0%.
18:08 So for our next iteration of logging, what I'm actually going to do is leave that section completely alone.
18:14 What we're going to do now is transition across to do some wide open throttle ramp runs and we'll gather some data that way.
18:22 And what we're going to do here is look at a higher range of our mass airflow sensor voltage.
18:28 Ideally what we're trying to do is span between the entire range of our mass airflow sensor scaling using our steady state test that we've just looked at and our ramp run test.
18:38 So let's set up our dyno and perform a ramp run now and we'll gather some data.
19:05 So we've got our run complete there, on the dyno we can see that we've produced 255 horsepower at the wheels.
19:11 Although we're not interested in power at the moment, we're just using the dyno to allow us to apply load in order to gather that data.
19:17 What we really want to do is have a look at that data that we've just gathered in our MegaLogViewer HD software so let's load that up now.
19:26 Alright we've got our log file open here and remembering that our white trace here in our top run, on our top graph is our engine RPM.
19:34 So we can see the area for our ramp run there is the entire log file itself.
19:40 We've started the log file just before we started the ramp run on the dyno and we've finished it as we've backed off the throttle.
19:46 And the area of interest for us of course is our short term and our long term fuel trims and we can see that they are relatively good through this log file.
19:57 We've got a maximum of around about 3-4% positive and a minimum of around about -3% so we're actually already pretty good and that was just one iteration of changes there.
20:09 If we have a look at the MAF error channel that we've created, we can see that straight away the calibration is pretty close.
20:16 We've got a maximum error right at the start here of about 5%, through the mid range of our run, about 3% at this point, around about 2.5% and then about 3% again.
20:28 What we can see is in the top area of the run, the high RPM area, we do start to move a little bit rich and right at the top of the run we're removing around about 2% of fuel.
20:41 So we need to now apply these changes to our mass airflow sensor calibration.
20:47 Now the important thing here is making sure that we're overlapping between what we've done in steady state and what we've done in our ramp run.
20:53 And at the very start of this run here our mass airflow sensor voltage is sitting at around about 2.5 volts.
21:00 So this means that we've got a gap between about two volts and about 2.5 volts that we do need to fill in.
21:05 We can do this by performing some steady state tuning under slightly higher RPM conditions or alternatively, we can choose to do a ramp run at part throttle just to fill in those areas and see what's happening between 2-2.1 volts where we do have information and in this case 2.5 volts where we can see that we are actually a little bit lean and the short term and long term fuel trims are adding fuel.
21:27 What we're going to do though is address the rest of this first and we'll come back and have a look at that.
21:32 So at this point we're looking at around about 2.98 volts from our MAF sensor.
21:38 You can see that we've essentially got zero total trim so we're not going to need to do anything there.
21:44 We can see that once we get up to about 3-3.1 volts, we start to go a bit positive in our trim.
21:50 Again I'm not trying to make specific changes here.
21:52 From 3.1 volts up to around about 4.1 volts, so 3.1 - 4.1 volts, we are around about 3% lean, we're adding 3% with our trims.
22:04 So let's apply that change.
22:08 So I wanted to make that change between 3.1 volts, so we'll highlight that point there and 4.1 volts so we'll scroll out until we're at 4.1 volts and we need to add some more fuel so in order to do that we use the multiply function and multiply by 1.03, that will make that change.
22:27 Let's have a look at our log file again, and as we go past 4.1 volts, we can see that our trim does start to drop back down at around about 4.18, let's call it 4.2 volts, we are adding 1% but that's close enough, I'm not too worried about that.
22:45 What we can see though is once we go past about 4.2, 4.3 volts, we start to go a little bit negative in our trim.
22:51 So from 4.3 volts and above we're going to pull out 2% so lets go ahead and do that now.
22:59 I'll scroll across to 4.3 volts and what I'm going to do is extend that error out all the way to the end of our MAF scaling here at five volts.
23:08 So we're going to click the M key again, we want to remove 3% so let's make that 0.97.
23:16 So we can see that in this instance we haven't made any changes between 4.16 and 4.24 volts, we're going to leave that for the moment.
23:23 And if we scroll back across to the left we know we've got this area here between 2.16 and 3.04 volts that we've got no data so we can't make an educated decision on what to do there.
23:37 What we're going to do for our next test though is we'll perform another ramp run and we'll test to see what the effect of that change to our MAF scaling is above 3.1 volts, we'll see if we've cleaned up that scaling a little bit better.
23:50 So let's go ahead and perform another test now.
24:12 So with our second ramp run complete there we can see on our dyno that we've produced essentially the same power as our first run, completely expected given we're really not making any changes to our actual fuelling, ignition or boost.
24:23 What we're interested in of course is our closed loop trim so let's have a look at our log file in MegaLogViewer HD again and we'll see how that change has affected things.
24:34 So again we're going to go straight to our mass airflow sensor error channel here.
24:37 We can see that in red again in our third set of graphs.
24:41 And what I've done purposefully is started this particular ramp run with partial throttle just to try and fill in that gap between 2.1 and 3.1 volts that we didn't have before and you can see that's worked quite well.
24:54 We're starting at actually about 1.8 - 2 volts, the beginning of that ramp run.
24:58 So we can see that we've got this big positive trim here, around about 4.8% at that particular point I've just chosen.
25:05 We can see by the time we get up to 2.6 volts, that's fallen back to essentially zero so we've got some work to do here.
25:13 We can see for the most part if I just randomly click through this and we look at our trims, we can see that we are much closer to the mark there.
25:20 I'm not really seeing anything much more than about plus or minus 1% for the majority of this.
25:26 We do however have this area here where it looks like we've now got a positive trim coming in so we're going to need to address that.
25:35 Right up the top of the run though where you will remember we were a little bit rich and the short term and long term fuel trims were pulling fuel out, we can see that that is now sitting at 0.7% so it's actually well within the area I'm happy with so let's address this lean area here where we're adding fuel for a start.
25:52 So if we try and find sort of an average position here, from around about, let's say 4.1 volts through to around about 4.3 volts, we're adding about 3% so let's go and make that change now.
26:05 So I've found the area that we need to make that change, 4.16 through to 4.32 volts.
26:10 We've highlighted that area, we'll use the multiply function again and we'll add 1.03 to add 3% there.
26:19 Now let's go back to our logging and we'll have a look at what was happening down below.
26:22 Now if we zoom in a little bit on the area where we had this positive trim here and we have a look at that, essentially from around about 2.1 volts through to around about 2.3, 2.4 volts, I'm just going to split the difference there and we're going to add 3% there so let's head back to our Accesstuner and make that change.
26:49 OK so with our changes now committed, let's go through and gather some more data.
26:54 What we'll do is start by having a look at our operation under steady state conditions and see if that's improved and then we'll perform another ramp run and we'll have a look at our data.
27:25 Now let's load up the data from our ramp run and we'll have a look at that.
27:28 And as we go through our ramp run if we just click through this log file, we can see that our trims are actually looking pretty good.
27:36 I've got a slight positive here as we get up to around about 6000 RPM at this point, we're positive 1.18% but definitely not going to be too concerned about that.
27:48 As I've said, if we can get within about plus or minus 2-3% we're doing pretty well and that's what the closed loop trim system is there for.
27:56 We've got an area here down at 2300 RPM, 3.06 volts where we are a little bit negative in our trims there and again right at the very start of the run, no we're actually pretty good there at about 0.7% positive trim, 2.5 volts.
28:13 So at this point you've seen how to use the dyno under steady state conditions as well as ramp run conditions to help us scale that mass airflow sensor.
28:21 You've seen how we can gather data and datalog the information we need in order to decide what changes need to be made and how to do this across the entire voltage range of our mass airflow sensor scaling curve.
28:34 Now this is an iterative process, you're not going to get it right in one set of changes.
28:38 Normally it'll take between two and four sets of changes before you've got your mass airflow sensor scaling correct.
28:45 It's also important to just reiterate here that you don't need to beat yourself up trying to get an absolute perfect result where your combined trims are sitting at zero the entire way through your datalog.
28:57 That's not realistic and usually not even possible.
29:00 Even if you could get to a perfect zero, it's not really going to help your tune in the end.
29:06 As long as you're within a range of plus or minus 2-3% then we can be confident that our mass airflow scaling is accurate and that's going to mean that as we move through the rest of the steps of this process, it's going to make our fuelling and our ignition tuning much easier and much quicker.
29:21 Now we've jumped ahead here, I've done a few more rounds of modifications, let's jump into my laptop software and we'll see what our combined trims are looking like once we've got our mass airflow sensor curve dialled in.
29:33 So here's our ramp run here, we've started our ramp run from about 2000 RPM, running through to 6500 RPM and we can see that our trims are sitting between around about 1.6 to minus two percent.

We usually reply within 12hrs (often sooner)

Need Help?

Need help choosing a course?

Experiencing website difficulties?

Or need to contact us for any other reason?