Practical Reflash Tuning: Aftermarket Injector Scaling - Subaru
Watch This Course
$229 USD
-OR-
Or
8 easy payments of only $28.63
Instant access. Easy checkout. No fees.
Learn more
Aftermarket Injector Scaling - Subaru
40.05
| 00:00 | - In the body of the course, we've already covered the process required to rescale the injectors on a simple ECU where we don't have full characterisation data, typically where we've got a flow scaler value as well as a latency or dead time or offset table. |
| 00:16 | So we're going to actually go through that process in this example here using our v11 Subaru STi so you can see first hand how it's done. |
| 00:27 | Before we get too far into this, let's just talk a little bit about what we're going to be doing and some of the traps that we can fall into. |
| 00:34 | Here we're going to be upgrading from the stock STi injectors to a set of Injector Dynamics ID1000s, although the same process that we're using will work irrespective of what injectors you are using. |
| 00:46 | The 1000s would probably be a reasonably common upgrade on this sort of engine. |
| 00:50 | Now one of the problems we can come across is that we're assuming that the stock calibration is in fact stock as well as the mechanical setup of the engine. |
| 01:01 | That's not always going to be the case so a little bit of due diligence there before we get started can save a lot of headaches further down the track. |
| 01:09 | One of the real big traps we can fall into though is trying to rescale the injectors at the same time as we're trying to rescale a different mass airflow meter. |
| 01:20 | Even if we've got the stock mass airflow meter fitted, even changes such as a new intake pipe or airbox can actually alter the accuracy of the MAF scaling. |
| 01:30 | So this is a really big no no, when we're getting started rescaling our calibration for larger injectors, it is critical that we're either starting with a stock mass airflow sensor and stock intake piping, plus the stock MAF calibration or alternatively that we've already rescaled our mass airflow sensor calibration to suit the different intake pipe etc before we get onto our injectors. |
| 01:55 | It's not possible to rescale the injectors and the mass airflow sensor calibration at the same time. |
| 02:02 | And the reason for this is that both the injector data as well as the mass airflow sensor calibration will affect the amount of fuel we're delivering. |
| 02:10 | So what I mean here is that if we've got a different intake system on the car and we've fitted larger injectors and we're in a situation where we're at 3000 RPM and our air/fuel ratio closed loop trims are highly negative, where do we fix that? We could fix that either in the mass airflow scaling or in the injectors, chances are that both are incorrect, there's no way of us knowing where specifically to make those changes so that's really critical, we must make sure we are either dealing with a stock intake system and stock MAF calibration, or alternatively we've already made the required changes to our MAF calibration to suit our new intake. |
| 02:49 | And typically what I'm looking for here in order to know that everything is on point is I'm looking to see that my closed loop trims, my combination of short term and long term fuel trims are within +/- about 5%. |
| 03:02 | Ideally the closer we've got these to zero the better but we also don't want to beat ourselves up too much. |
| 03:09 | The trims are always going to move around a little bit, that's why they're there in the first place so trying to get that dialled in to absolutely being spot on 0 is really just going to waste a huge amount of time. |
| 03:21 | If we're within that range, +/- 5% we're pretty much good to go so that's one of the first things we want to do is check that our calibration is correct. |
| 03:29 | It's also a good idea just to visually inspect the car. |
| 03:32 | Pop the engine bay open and actually have a look at the intake system, is it still stock or has it been modified and those will usually give you some pretty clear clues that something is not right. |
| 03:44 | Let's jump into the software that we're going to be using and just navigate around and see what parameters we're actually going to need to be changing. |
| 03:51 | So in the software we can see we're using EcuFlash, there's a variety of different options available for the reading, writing and flashing so all of them are going to give us access to the same parameters, doesn't really matter what you're personally using. |
| 04:04 | Now you can see there are a lot of parameters available in the menu structure on the left hand side and we're going to just focus on what we need for our injector scaling. |
| 04:12 | Most of which we can see comes under the fuelling-injectors menu structure. |
| 04:18 | Now first of all we have our injector latency and then we've got our injector flow scaling. |
| 04:22 | Now the first part we can see here is our injector flow scaling in the stock calibration is set to 562 cc/minute. |
| 04:30 | Not the point here is that that's actually not exactly what the stock injectors flow. |
| 04:35 | Depending on whose data you want to believe, stock injectors in these engines, generally around about the 550 cc mark, close but not quite right. |
| 04:45 | And that's important because we'll see how that plays into things when we actually start making changes to the injector scaling. |
| 04:53 | Next we've got the injector latency table over here. |
| 04:56 | You can see that it is set up as a 3D table there with our manifold pressure referenced on the vertical axis which, as we can see, is not being used, the data is the same irrespective of that axis. |
| 05:07 | So we can treat it as a 2D table and we've got battery voltage on the horizontal axis. |
| 05:13 | Now again it's tempting here when we buy a set of aftermarket injectors from the likes of Injector Dynamics, just take their data straight from their website for injector offset or latency and drop these numbers into this table. |
| 05:25 | But that is not going to work, it's not going to be accurate so we're going to need to develop the numbers for this table as we go and you'll see how that's done. |
| 05:36 | Next, although it's not specifically relevant or related to our injector scaling but just because it does come into play here is our mass airflow sensor scaling so we'll come down to our mass airflow sensor scaling here and what we want to do is just make sure that the table hasn't been changed, now that's not always going to be easy to see but one thing that is worth just checking out here is just visually making sure that this table is nice and smooth. |
| 06:02 | Any time we've got a mass airflow sensor scaling that's got big bumps with dips in it, that's an indication that something is not right there so we want to make sure that that does have a normal, smooth, exponential shape to it. |
| 06:15 | Alright we can close that back down now. |
| 06:18 | What we're going to do is just check and see where our trims are with our stock calibration, again just making sure that everything is on point before we actually start modifying our calibration to suit injectors, so obviously at this point we haven't changed our injectors, we still have the stock 550 cc STi injectors so let's get our engine up and running and we'll see what we need to look at. |
| 06:43 | Alright we've got our engine up and running here and obviously for our demonstration we're using our Mainline dyno but this process can easily be achieved out on the road, it's not very difficult, we don't even need to be driving the car fast. |
| 06:55 | What we're going to be doing is logging the fuelling under two conditions. |
| 06:59 | First of all idle and then under light load between about 3200 and 3500 RPM. |
| 07:06 | Now before we get into that, let's just try and understand what we're trying to achieve here. |
| 07:11 | And first of all, if we are adjusting our injector scaling number, obviously if that number's incorrect, either too large or too small, that's going to affect our overall fuelling and we're going to either be rich or lean everywhere and we're going to be using our closed loop short term and long term fuel trims to give us that indication. |
| 07:29 | However, we can get our scaling correct and find that our fuelling's correct at maybe 3500 RPM but it may still be showing too rich or too lean at idle. |
| 07:39 | And if we've got that situation where it's correct at one point and too rich or too lean at another point, obviously we can't fix that with our scaling. |
| 07:47 | That is a red flag that our latency or offset table is incorrect. |
| 07:51 | So let's just understand what the latency actually means. |
| 07:55 | Our fuel injectors are obviously a mechanical device and when the ECU sends a pulse width to the injectors to open them, they can't open instantaneously, there's a bit of a delay for the injector to physically move off its seat and to open and start flowing fuel. |
| 08:11 | We actually have a really small delay when the injector closes as well, it doesn't close instantaneously so you could think of latency as essentially being the difference in time between the signal of the ECU asking the injector to be open and how long the injector actually is open and supplying fuel. |
| 08:30 | Now why that's important is that it creates some errors in our fuelling. |
| 08:34 | Essentially the ECU is asking for a given mass of fuel to be delivered but because the latency's wrong, that fuel isn't actually getting into the engine. |
| 08:42 | The other aspect and what we're going to be using to our advantage is let's say our injector latency is 1 millisecond and let's say we're idling with an injector pulse width of 2 milliseconds. |
| 08:53 | Well the percentage of the overall pulse width being delivered to the injector, in order to get 2 milliseconds of injector open time, is clearly 3 milliseconds and that deadtime of 1 millisecond makes up a large percentage of that overall 3 milliseconds or obviously a third of that, 33%. |
| 09:12 | However if we're up in the RPM and up in the load a little bit further and maybe now our injector pulse width is maybe 6 or 7 milliseconds, now the dead time is a much smaller component of that so this is what we're using to our advantage, essentially what I'm saying here is the deadtime doesn't impact our fuelling quite as much at 3500 RPM as it does at idle so that's how we're going to be using those two pieces of data to help us with our scaling. |
| 09:40 | If it's sounding a little bit complicated, don't worry, it'll all fall into place as we move through our demonstration. |
| 09:46 | For our logging which is obviously an essential part here, we're going to be using RomRaider and just like our flashing software, there's a variety of options here. |
| 09:55 | They all essentially give us access to the same parameters. |
| 09:58 | Let's jump into the software and see what we are looking at here. |
| 10:02 | So for a start we want to be logging our air/fuel correction 1 and our air/fuel learning 1. |
| 10:09 | And air/fuel correction 1 is essentially an instantaneous response, this is our short term fuel trim that's reacting because there is an instantaneous error. |
| 10:19 | And while I'm talking we can see that that's sitting at 0.78 so very slightly positive, it's adding a little bit of fuel. |
| 10:25 | Our air/fuel learning 1 is essentially our long term fuel trim and this essentially starts building up or down over time. |
| 10:34 | If we've got a consistent short term fuel trim, over time that'll basically transfer across to our long term trim and what that means is that when the ECU transfers through that same operating region again, the long term fuel trim will essentially correct any error that's there, meaning that our short term fuel trims over time should sit really close to 0. |
| 10:54 | So let's have a look and see what our conditions are at the moment. |
| 10:58 | We've already said we're just under 1% positive trim on our short term and our long term is just under 4% so we're a little bit positive there, just under our 5% threshold so we're in the ballpark here already, which is pretty much what we want from a stock calibration. |
| 11:16 | Now other aspects that are important to understand here as well, we are logging our air/fuel ratio sensor 1 so this is our front wideband sensor which sits in the downpipe off the back of the turbo. |
| 11:27 | This is a true wideband air/fuel ratio sensor in this particular engine so we can use this to see what our actual air/fuel ratio is and we're also comparing this to our closed loop fuelling target which is what the ECU is requesting. |
| 11:41 | We can see that that is basically unsurprisingly sitting at about the same, about 14.5, 14.6. |
| 11:47 | So we're at stoich here, sitting at idle. |
| 11:50 | Playing into this as well is another parameter that I'm logging which is our ooen loop and closed loop fuelling status. |
| 11:57 | So that's CL/OL fuelling. |
| 11:59 | And we can see that that is sitting at 8. |
| 12:03 | This is meaning that the closed loop is functioning at the moment. |
| 12:08 | There's a variety of different status flags here, but we want to make sure that when we are logging data that we are sitting in the closed loop mode, it is operational which is as I've mentioned there, a status flag of 8. |
| 12:22 | Now the other area that's very easy to make mistakes when we're gathering this data is gathering the data essentially under conditions that aren't realistic. |
| 12:31 | And particularly if you're doing this on the road, what you're going to be doing is starting and stopping the car a lot and making changes, flashing these into the ECU. |
| 12:40 | What that results in is a lot of heat soak in the engine bay and this for a period of time after we restart the car and get back up and running can result in some inaccuracies and we want to make sure that we're not operating under those conditions so it's always a good idea before we start taking notice of our logged data, taking notice of our trims, make sure that we drive the car for a period of time and just allow that heat soak to evaporate. |
| 13:05 | So at the moment we know that our overall trim is just under our 5%, it's a little more positive than I would like but we are under those exact conditions I just talked about, we've only just started the car so I wouldn't really take too much notice. |
| 13:18 | Let's have a look at our other point and here I'm going to be using 3200 RPM light throttle so this would replicate cruising. |
| 13:27 | A little intricacy here with the v11 STi or the Subarus in general is that they are known to have a fuel system oscillation around about 2800 to 3000 RPM and normally I would do this test at about 3000. |
| 13:45 | I'm just pushing it up a little bit just to get outside of that oscillation range so we don't have that fuel system weirdness that the Subaru is known for affecting our results. |
| 13:56 | So let's get ourselves running in fourth gear now, 3200 and we'll have a look at our data. |
| 14:00 | We're up and running now at 3200 RPM. |
| 14:03 | I'm just at light throttle and what we can see here, if we look at our short term and long term fuel trims, we've got a short term that's slightly positive at around about 1.5-2%, it's moving around as I'm talking. |
| 14:16 | We've also got a long term trim of -3. |
| 14:21 | And while I've been talking here we've seen that that short term fuel trim is moving around quite a bit. |
| 14:25 | So it's a really good point here that when we are doing this, we want to make sure that we allow the engine to run for long enough that we actually get stable results there and while I'm talking we can see that those have stabilised to a point and if we look at the combination of the short term and long term fuel trim, we are really really close to 0 so obviously that's exactly what we want and it is important to mention there that these are cumulative so we want to add the short term and long term fuel trims together so probably about +1% give or take, it's moving around and it's very very close to 0 so I'm really happy with that. |
| 15:03 | Another couple of things that I'll just point out here, we are logging our coolant temperature. |
| 15:07 | We can see that that's sitting at 84° just like our head soak aspect we don't want to be making these changes when the engine is warming up and it's in cold start, that's not going to give us realistic results and we don't want to be doing this logging when it's overly hot either so normal operating temperature. |
| 15:25 | And we also are logging our mass flow sensor voltage and our mass airflow so we can see those values here. |
| 15:32 | And that's just going to give us an indication if there was something weird going on with our calibration in stock form, that'll give us a reference to where abouts it's actually accessing on the mass airflow sensor scaling. |
| 15:43 | So again, I've been talking for a while here, we just monitor our short term and long term trim again, yeah we're looking at a combined total of about 0 to maybe 1%, as I've talked it's actually gone slightly negative but we're within that plus or minus 5% so I'm happy with that. |
| 16:00 | What we'll do is we'll just stop the vehicle now on the dyno, that little period of running there would have helped get rid of any heat soak as well so we'll come back to idle, we'll give the engine a few moments to stabilise and we'll have a look again at our short term and long term trims under idle conditions. |
| 16:17 | Alright we've been idling for a few moments now and we can see that those trims are again working out to be really close to zero You can see that the short term trim kind of ends up transferring across to the long term trim but at the moment we're about -1, 1.5% so it's a really good indication of what we should be looking for and it's also really backs up what I was saying before that we want to make sure that we aren't operating in a heat soak condition because it's going to give us inaccurate data. |
| 16:48 | So at this point we're comfortable that our stock calibration looks like it is on point, we've got our short term and long term fuel trims, the combined trims are within that +/- 5%, in fact in our case very very close to zero which is looking good. |
| 17:04 | We know that we've got a stock intake system on the car so we can shut down now and we'll actually get our larger set of injectors installed and then we can come back and look at how we can address our scaling. |
| 17:14 | Alright we've gone ahead and installed our Injector Dynamics ID1000 injectors and while we're not really focusing on the mechanical job of changing injectors, there's a couple of aspects I just want to touch on here. |
| 17:27 | Obviously any time we're dealing with fuel systems, there is the risk of a fuel leak and the potential for fire so it goes without saying that a lot of care is required here. |
| 17:38 | Couple of things that I will mention, it's always a good idea to basically drain down or reduce the fuel system pressure before you go ahead and remove the existing injectors. |
| 17:48 | Easy enough to do this by removing the fuel pump relay so that the pump will stop running and basically just idle the car until it stops so that's just going to reduce that pressure that's going to be in the fuel rail so you're not going to get fuel spray out when you remove the rail and the injectors. |
| 18:05 | Also a good idea as well just to keep some rags on hand, you're inevitably still going to get some fuel leak out, even though you have reduced that pressure so we don't want that leaking everywhere, we want to make sure that we clean that up straight away. |
| 18:19 | The other aspect here as well is once we've got the new injectors fitted, when we first pressurise the system, we want to be very mindful of watching carefully and making sure that there are no fuelling so don't want to just try and start the engine straight away, always a good idea to just cycle the ignition a couple of times, it'll build pressure in the system, have someone else even watching the injectors carefully and just make sure that there's no visible leaks. |
| 18:44 | Last point I'll make here, always a good idea to keep a fire extinguisher on hand while we're doing any work like this. |
| 18:52 | Alright with our injectors fitted though, everything tested, we are ready to get the engine up and running but we can't start it just yet, we've got a little bit of work to do in our calibration to get ourselves ready to go so we're just going to go through that process now. |
| 19:06 | Let's dive into our EcuFlash software and the first area that we're going to want to focus on is our injector latency and our flow scaling. |
| 19:16 | So let's just open both of those up here for the moment, we'll put these side by side, again these are our stock figures. |
| 19:23 | We know that we've gone to a set of Injector Dynamics ID1000 and this flow scaling here for our stock injectors, 562.45. |
| 19:31 | Can be tempting just to take the data from the Injector Dynamics injectors and drop that in, it's not necessarily going to work properly though so let's just have a quick look at the scaling data for the Injector Dynamics ID1000s and what we've got here is the latency or offset depending on what you want to call it, as well as the flow. |
| 19:54 | So what we want to do for a start is just understand what our fuel pressure is because that will affect both our latency values and our injector flow. |
| 20:03 | In a typical Subaru fuel system, we're expecting our fuel pressure to be 3 bar or 43.5 psi so essentially we are looking at the data solely from this row here. |
| 20:16 | Now we can see here that our slope or flow is 1015 cc per minute and we can see our offset data as well, I'll just focus on one particular parameter here, we can see that we've got 990 microseconds here, 0.99 milliseconds and that's the latency that we'd use in an aftermarket ECU but again if we just take this data and we drop that all into our table here, we'll probably be close, it'll probably get up and running OK but it's definitely not going to be quite right and this is where it gets a little bit tricky but as we go through this you'll see how to deal with this. |
| 20:55 | So what we're going to do for a start is simply drop in our scaling data and we're going to leave our latency exactly as it is. |
| 21:03 | With this number here, what we can do is just enter our flow data that we looked at from the ID1000 flow sheet there and we know that that was 1015, press enter and that will change that number. |
| 21:17 | OK so just making that one change there to our scaling, this should be enough to allow us to flash this into the ECU and get us up and running. |
| 21:25 | There are a few other tables though in the Subaru ECU that are based around a injector pulse width value so of course when we essentially almost double the size of our injectors from 560 to 1000 cc, we're going to need to address those tables as well. |
| 21:42 | So let's go through those one at a time. |
| 21:45 | The first table we're going to have a look at is found under our fuelling, closed loop to open loop transition menu. |
| 21:53 | A lot of parameters here, the one we are interested in though is this one here, our closed loop to open loop transition with delay, base pulse width and when we open that, we can see the axis is defined there as base pulse width in milliseconds. |
| 22:05 | So we want to make an appropriate change here based on the injector size change we've made and we know that we've gone from 550 to 1000 cc injectors so we've essentially almost doubled our injector size and on face value you'd think that obviously if we halve the numbers in this table that would be a good place to get started. |
| 22:25 | That actually seems to be a little bit overkill though, goes a little bit too far so instead what I'm going to do is make a slightly smaller change of about 30% here. |
| 22:34 | Obviously you can fine tune this to suit your requirements once you get up and running. |
| 22:37 | However we're going to start by highlighting the entire table here and to make that reduction of 30%, we can use the multiplication function. |
| 22:46 | You can either enter the multiplication symbol or go to the edit menu, click on multiply data and to reduce this by 30%, we want to multiply it by 0.7. |
| 22:58 | Essentially 1 minus 0.3. |
| 23:01 | So let's go ahead and do that now. |
| 23:03 | Alright let's close that table down now, we can close down our menu structure to keep things nice and simple here and the next change we're going to make is to our cranking fuel here and we've got a few parameters here, what we're going to do is basically make the same change to all of these tables that are pulse width based. |
| 23:21 | We do have some compensations down here, these are percentage based compensations, we don't want to touch those, if we open this one here for example, again we can see that the axis is in milliseconds so again, we're going to make that same reduction of 30% or multiply by 0.7. |
| 23:38 | Let's go ahead and change all of these tables now. |
| 23:52 | Now just a point on our cranking tables while we are modifying these here, we can see we've got a number of tables listed as A through to H however if we look at the numbers in these tables we can see that it would appear at least G and H which I'll open both of these here, are not being used. |
| 24:09 | Even in the stock form before I made any modifications we can see all of the numbers in these two tables are set to 0.01 miliseconds. |
| 24:17 | There's a bit of a red flag when we compare this to one of the other tables where we can see the sort of shape of the table, the trends of the table, these make sense as cranking enrichment tables so in this form here we can see it's pretty unlikely that tables G or H are actually being used. |
| 24:32 | Not uncommon in a factory ECU to have multiple tables, some may not be used, we're not always going to know which tables are used either, which is why we go through and make the same percentage changes to all of them anyway. |
| 24:45 | Let's move on. |
| 24:46 | We're now going to deal with our tip in or acceleration enrichment, again a lot of parameters here and this is one that is quite often overlooked when fitting larger injectors to factory ECUs, factory engines like our Subaru here. |
| 25:00 | And the engine will run OK but you're going to have poor transient control of the fuelling so what we're going to do here is make some changes to our throttle tip in enrichment A, our throttle tip in enrichment B, I'll just move those across so we can see both of those there, both of them exactly the same numbers and we also need to change our minimum tip in enrichment activation. |
| 25:23 | Now this one is relative again to pulse width which is important, we also have another one here which is listed relative to throttle position which we're not going to change so again we're going to make that same 30% reduction to these three tables so let's go ahead and get that done. |
| 25:40 | Alright with our changes made we can now save our file and obviously we want to give that a name that's going to make sense, clearly in this instance, we're going to just remove our stock title there and we're going to add ID1000. |
| 25:56 | With that done, we can now flash those changes into our ECU, let's go ahead and do that now. |
| 26:04 | Alright we're now ready to start our engine for the first time, let's get it up and running. |
| 26:10 | Now if we've made all of our changes as I've just discussed, our engine should start and run just like it is now, pretty much just like normal. |
| 26:17 | Granted our fuelling right now isn't going to be quite right but it should be close enough at least to allow our engine to run which is really the whole purpose of our first round of changes. |
| 26:27 | We do however need to allow the engine to come up to a normal operating temperature and we also need to allow the ECU to drop into closed loop operation before we can actually start gathering data and actually see what we need to do for our next round of changes, so we'll allow that to happen now. |
| 26:42 | Alright our engine's been running for around about 10 or 15 minutes now, all of our temperatures are nicely stabilised and I've allowed it to idle here for long enough for our trims to stabilise as well so we've got a really good snapshot of what's actually happening so let's have a look at that data and we can see our combined short term and long term fuel trims there, obviously they are always moving around a little bit but around about 5.5% long term, around about 1% on our short term so 6-7% is our combined trim there and it's a positive trim meaning that we are naturally a little bit lean, we're not supplying enough fuel there. |
| 27:20 | OK so that's our trim at idle, remembering at idle we are much more affected by our latency so let's get ourselves up to 3200 RPM cruise, our second data point and see what our trims are like under those conditions. |
| 27:37 | Alright we've allowed our engine to run at 3200 RPM for long enough to allow everything to stabilise and just for our trims to sit reasonably consistent and stable so let's have a look at what we've got going on here. |
| 27:49 | And at the moment we can see we've got a positive trim of around about 3-4% and we've got a negative long term trim of -3.9%, let's call it -4%. |
| 27:59 | So essentially these are cancelling each other out, yes they are moving around a little bit but for all intents and purposes, our overall trim is within about 1% of 0 so even that first change that we've made there, we're very very close. |
| 28:13 | If I wanted to be really fussy, our overall trim is possibly very marginally negative overall, maybe -1%, I'm certainly not going to be worrying about that. |
| 28:24 | So let's come back to idle here, we'll shut our engine off and we'll talk about what we can learn from those numbers. |
| 28:30 | First of all, what's worth understanding here is even with that one change just to our scaling and not adjusting the latency at all, the trims are actually pretty much within the realms of what I'd consider acceptable, many tuners would just tune the car based on the scaling that we've just changed now and not worry anything more about the latency but there are some hints here that our latency does need a little bit of work. |
| 28:56 | What we know is at 3200 RPM, our overall trims were essentially zero so our scaling data at this stage, I'm actually pretty happy with, that first change that we made is pretty much on point, however we had a positive trim at idle and at idle remember we are more affected by our deadtime or latency value, the latency makes up a bigger proportion of our overall injector pulse width so this is why we've got an error here which isn't really showing itself at 3200 RPM where our overall injector pulse width is much greater. |
| 29:30 | So we know that our trim at idle was positive and what we want to do to correct this is increase the latency value. |
| 29:39 | And the reason for this is when we increase the latency value, what it's doing is telling the ECU that the injector needs to be opened for longer in order to get the correct amount of fuel out which is exactly what we want. |
| 29:51 | A longer pulse width is going to provide more fuel, more fuel means that our natural trims will be less, hopefully closer to 0. |
| 30:00 | So let's jump into our reflash software and we'll see how we can make that change. |
| 30:06 | Right in our EcuFlash software, we want to come down to our injector latency and we want to make a change to this table so what we're going to do is highlight our entire table here, we'll use our multiplication symbol and we're going to enter a value of 1.05 and that's going to increase our values by 5% there. |
| 30:28 | Now I'll just quickly talk about the fact that I've made this overall change and this isn't necessarily going to be quite right. |
| 30:36 | We can see that our axis here of course is battery voltage and we may find that when we have our values correct at our normal running voltage of 14 volts, while we're retaining the general shape, let's actually have a look at the shape of that graph now, when we're retaining the general trend or shape of that graph, we might find that our latency down at maybe 9 or 11.5 volts, our lower voltage break points, that might need some more specific tweaking. |
| 31:07 | Now personally I'm not going to worry about this. |
| 31:11 | Generally what I find is as long as I follow the trend of that table, keep that same shape and make a percentage change to the entire table, we're going to be there or there abouts with those lower voltage and even the higher voltage of course. |
| 31:23 | The other aspect that really plays into this is that when the engine is running, if everything is operating as it should be, the alternator should be retaining our battery voltage at or very very close to 14 volts, generally 13.8 to 14 volts so this cell or break point here, this is really the most important one, it's going to influence our operation. |
| 31:46 | If you want to get really granular with this and go a little bit above and beyond then once you've got everything dialled in at 14 volts then you could of course disable the alternator and then that's going to reduce the charging voltage and that'll allow you to look down at maybe this 11.5 volt. |
| 32:02 | Again it really depends on your personal preference and how much time you're prepared to put into this. |
| 32:07 | The technique that I'm demonstrating, this is absolutely adequate in my opinion. |
| 32:12 | Alright what we're going to do now is save this, flash it back into the ECU, we'll come back once we're up and running with our engine at normal operating conditions and see the effect of that change. |
| 32:24 | Alright, again back up and running with our latency change made and I've allowed everything to stabilise, we're back in closed loop at idle and straight away we can see that our combined short term and long term trim, very very close to the mark, just with our quick change there of 5%. |
| 32:42 | We've got a long term trim of -2.3, a positive trim of 0.78 for our short term. |
| 32:48 | So we're essentially around about 1.5% negative total. |
| 32:53 | Absolutely clearly within our defined limits of what's acceptable there. |
| 33:00 | However that change is currently just at idle, that's all we're looking at there and of course when we change the latency, it's still going to have an impact, albeit a smaller one on our higher RPM, higher pulse width operation. |
| 33:15 | Remember we've increased our latency by 5%, now that actually made a change at 14 volts of about 0.04 milliseconds so it's adding that same 0.04 milliseconds at higher pulse widths as well so we're going to expect that we might have a knock on effect of being a little bit richer under our higher RPM operating point, let's get up and running there and have a look and see how we are. |
| 33:42 | Alright so we're up and running and we can see our trims are moving around, we've got a negative 3.9% trim on our long term and our positive trim between about 0 and maybe 2% on our short term so we're showing a slight negative trim of maybe around about 2, maybe 3%, obviously it's changing as we speak but exactly what we expected, we can see that we have gone slightly negative in our trim there. |
| 34:09 | It's absolutely still within the realms of what I'd consider to be acceptable though so at this point I'd be happy to now move on and tune this vehicle. |
| 34:19 | Before we move on, I just want to cover off a few elements and we'll start with a high level look at what we've just done or what we're trying to do with our injector scaling. |
| 34:28 | Now we're using our combined closed loop trims, our short term and our long term fuel trims to help guide us, we're looking at two areas, one at idle where we've got a small pulse width going to the injector and our latency makes up a larger proportion of that pulse width. |
| 34:43 | Hence our trims will be affected more by our latency. |
| 34:46 | We're also looking at a higher RPM, slightly higher load area, 3200 RPM where our pulse width is larger, hence the latency has less of an effect. |
| 34:56 | Now let's understand what our trims are telling us. |
| 34:59 | If our combined trims under both situations are positive, what this means is that our natural fuelling, our scaling is lean. |
| 35:08 | We're not supplying enough fuel and our closed loop system is making up that shortfall. |
| 35:13 | We can fix this primarily with our scaling number and we do this by reducing the scaling number. |
| 35:20 | So essentially if we've overall got a trim of positive 10% let's say, what we'd do is we'd start by reducing our scaling number by a factor of 10% as well. |
| 35:31 | So if we were at a value of 1000 for example, we would reduce that by 10% which of course is a factor of 100 so we'd take our scaling from 1000 down to 900 and test again. |
| 35:45 | And it's just an iterative process of adjusting that scaling until our trim's at least at our higher RPM point, very close to 0. |
| 35:54 | Now let's talk about our latency, remembering we're looking primarily at our idle operation to help guide us with our latency. |
| 36:00 | Now under those conditions if our trims there are positive, again what this means is that the trims are adding fuel because we're naturally too lean and under those conditions, we can fix this with our latency by increasing the latency value. |
| 36:17 | Remember when we increase the latency value, it extends the injector pulse width, hence adding more fuel. |
| 36:23 | But the guideline here is that how do we know whether we fix the situation with scaling or latency? Well if our trims at 3000 RPM, 3200 RPM are very close to 0 but at idle they are positive, we want to fix this with our latency. |
| 36:42 | If our trims are essentially very similar, maybe positive or negative at both points, our idle and our higher RPM point, we're going to fix that with our scaling. |
| 36:52 | OK so that's the first element I wanted to talk about. |
| 36:54 | The next is before you take the numbers that we've just developed here for our ID1000s and go and blindly drop them into your own car, understand that there are a number of factors that will actually affect your final result. |
| 37:08 | The first of these, which we've already discussed in the first part of this example is we are relying on our mass airflow sensor calibratioin being correct. |
| 37:19 | That's not always going to be the case so sometimes you can end up essentially putting a bandaid on an incorrect MAF calibration with your injector scaling and latency. |
| 37:30 | If we've gone through this process like I've demonstrated here, we're already starting by proving that's correct but I just want to mention, this is a trap that will catch some people out. |
| 37:39 | The other element though is that our injector scaling and latency values will depend on our fuel pressure. |
| 37:46 | Now in a stock car, you're not going to know what the fuel pressure is. |
| 37:50 | We can assume the factory 3 bar, 43.5 psi, that's not always going to be correct. |
| 37:55 | Particularly if you've fitted a larger volume of fuel pump, you may find that your actual fuel pressure jumps up very slightly. |
| 38:03 | Sometimes you can essentially overwhelm the factory fuel pressure regulator, it can't quite cope on the return flow and hence the overall fuel pressure may jump up a little bit. |
| 38:12 | That is going to affect your required fuel scaling and latency values. |
| 38:16 | The other element and again one that's really easy to overlook is ethanol content. |
| 38:21 | Now if you've got ethanol in your fuel, maybe E85, most people would understand that's going to have a dramatic effect on our required injector scaling and latency values. |
| 38:32 | However, a lot of pump gas these days still may have an ethanol content of 5-10% so again this will skew those results and as a point here, if you were developing a dedicated E85 map then you would change your scaling values to account for the E85 content. |
| 38:52 | Again the process is exactly what we've looked at, just need to understand that the numbers are going to look nothing like the flow and latency values from our Injector Dynamics data. |
| 39:02 | It's worth mentioning that while for our demonstration here, we've dialled our injector scaling and latency in with just two changes, this is rarely going to be the case, it is an iterative approach and normally you'll need perhaps 3-5 iterations in order to get everything dialled in. |
| 39:20 | Also as I've mentioned, when we do change our latency values, this can have a knock on effect of needing a further change to our scaling if we end up affecting our trims at our higher RPM point so it is a bit of an iterative change, just requires an understanding of what you're trying to achieve which hopefully this module has explained and a little bit of patience. |
| 39:42 | The last point that I'll make here is that our tuning is not complete. |
| 39:47 | At this point we have only dialled in our injector scaling and from here of course, we actually need to tune the engine and make sure that our fuelling is on point, particularly in the higher load, higher RPM areas of operation where we're operating under open loop conditions. |
