Practical Reflash Tuning: Step 6: Confirm Calibration on the Road
Watch This Course
$229 USD $114.50 USD
-OR-
Or
8 easy payments of only $14.31
Instant access. Easy checkout. No fees.
Step 6: Confirm Calibration on the Road
23.47
| 00:00 | - For the 6th and final step of our 6 step process, we've got the car off the dyno and we're here at Highlands Motorsport Park to put it throuhg its paces and see if everything that we saw on the dyno stacks up under real world conditions. |
| 00:12 | Now while we are privileged here to be able to use a racetrack which gives us the benefit of knowing we're not going to come across oncoming cars and there's definitely not going to be any pedestrians out there, everything that I'm going to do today can be done safely out on the road. |
| 00:27 | I do urge you if you are going to be doing your tuning on the road though to make use of a helper, someone who can drive the car, concentrating solely on the driving process, that will leave you freed up to concentrate on running the laptop. |
| 00:41 | This way your attention is not going to be divided and this is going to ensure safety. |
| 00:45 | Having said this, when it comes to road tuning with the reflashing process on GM vehicles, we actually do have a few advantages here because we will be concentrating on using the scanner to gather the data so while we're actually driving out on the road or track we don't really need to be watching the laptop screen, we can concentrate solely on driving the car then once we're back in the safety of the pits or pulled over to the side of the road, then we can analyse our data and see what changes are needed so it is actually a relatively safe approach. |
| 01:17 | What I would recommend when it comes to road tuning, particularly for our wide open throttle full power runs, we should really have access to a wideband air/fuel ratio meter, somewhere within the peripheral vision of the driver and this just lets the driver immediately abort the run if the air/fuel ratio is excessively lean or for that matter excessively rich. |
| 01:38 | It's always safest to abort, come back, make some changes and then attempt again. |
| 01:42 | Now before we head out onto the track I just want to talk a little bit about how we're going to be gathering this data and some of the parameters that we will be looking at. |
| 01:51 | So let's dive into the laptop software for a moment. |
| 01:54 | We've got our engine idling at the moment and if we have a quick look at our short term fuel trims, remembering our long term fuel trims in this instance are disabled. |
| 02:02 | We can see that our trims are sitting pretty close to 0, as I've been speaking our short term fuel trim on bank 2 has gone a reasonable way positive but if we look at the trend, at the start of this graph it was much closer to 0 and it will probably settle down again. |
| 02:18 | The important point I want to make here is twofold. |
| 02:21 | First of all, with our closed loop enabled and our short term or short term and long term fuel trims operational, these are basically a snapshot into the quality and accuracy of our volumetric efficiency table. |
| 02:33 | Remembering in the perfect world we'd have these essentially always sitting at 0. |
| 02:38 | Obviously when we move into power enrichment they're going to be of no use to us, they're going to be sitting at 0 but under closed loop conditions, this will give us a snapshot into how good our tune is. |
| 02:48 | So the other aspect of this though is right now if we look, our short term fuel trim back 1 sitting basically on 0, bank 2 sitting 3-4%, starting to move positive again. |
| 02:59 | So the point here is we've got a discrepancy back to bank and with the reflashing on GM vehicles, it's very difficult for us to accurately do much about that. |
| 03:08 | So what we're going to want to do really is split the difference with our trims bank to bank and basically adjust our volumetric efficiency tables so that the average trim is as close to 0 as we can and that's what we're going to be showing you how to do. |
| 03:23 | We also need to talk about the variety of ways that we can basically correct our VE table, now that we're out on the road or racetrack. |
| 03:31 | So obviously on the dyno we disabled our trims and we used our wideband input and we used our math channel for our equivalence ratio error. |
| 03:40 | We can also totally do that on the road or racetrack as well. |
| 03:44 | The option I am going to do here is to leave my short term fuel trims enabled and instead we're going to be using our short term fuel trims to track what changes are necessary. |
| 03:54 | If you are using short term and long term fuel trims, understandably you need to use the total of the short term plus long term fuel trim to get an accurate representation of that error. |
| 04:05 | So let's have a look at how we're doing this so we'll just get our chart logger out of the way here. |
| 04:12 | And we've got another histogram that we've set up here which you can see I've labelled short term fuel trim error. |
| 04:19 | So really it's set up almost identical to the graph that we used on the dyno for our steady state tuning although this time instead of our equivalence ratio error channel we're using our short term fuel trim error so let's have a look at that. |
| 04:32 | We'll go to our graphs layout and have a look at how it's set up. |
| 04:35 | Coming down to the one that we're interested in, you can see the parameter here that we've chosen, so this is where we need a little bit of understanding. |
| 04:43 | If we click on this, and if we type in STFT which stands for short term fuel trim, we'll see the options that are available. |
| 04:51 | First of all we've got this set of parameters here that we can choose. |
| 04:55 | You can see that these are bank separated though so we could choose bank 1 or we could choose bank 2 but that's not going to really take into account what's happening on the other bank and as we've just seen, there can be a bank to bank discrepancy. |
| 05:08 | We also have, if we come down here some pre defined math channels which are quite useful for long term plus short term fuel trims so if you are still using your short term and long term fuel trims, you can choose those, it'll sum those up which is going to give you the full amount of error. |
| 05:24 | What I've actually done though is I've created this user defined channel here called STFT average. |
| 05:29 | So as its name implies, it's simply the average of the trim on each bank. |
| 05:34 | Let's see how I've created that. |
| 05:36 | So we can go to our tools menu and we want to go here to our math parameters and click on that. |
| 05:43 | We've only got one available that I have set up here which is STFT average. |
| 05:48 | It's really simple here, what we can do is just click here to add new variables into this expression and that expression there is simply short term fuel trim bank 1, plus short term fuel trim bank 2, we're dividing that by 2 to get the average and then at the end here I am also multiplying by 100 to turn it into a percentage. |
| 06:08 | That's as simple as it is. |
| 06:10 | So that's the parameter we are using and we'll just open our graphs layout again, have a quick look at that again and the rest of this is exactly the same as what we've already looked at inside of the worked example. |
| 06:21 | We've got our column axes and our row axes basically replicating our VE table. |
| 06:26 | Close that down and we are now filling out this table with the average short term fuel trim. |
| 06:33 | So that's got us what we need as a parameter to enter. |
| 06:36 | Now obviously again, this is only going to be useful during closed loop operation so we're still going to approach this in 2 ways, we're going to go out onto the track, gather some data under steady state cruise conditions and fill in as much of the histogram as we can using that short term fuel trim average. |
| 06:53 | Once we're happy with that and made any adjustments necessary to our VE table, we're then going to repeat the process under wide open throttle acceleration conditions and that's going to allow us to use our wideband then and we'll see how closely we're tracking to our commanded equivalence ratio. |
| 07:10 | So really we're replicating exactly what we did on the dyno with some steady state tuning, some wide open throttle ramp run tuning. |
| 07:16 | Before we do head out on the track, there's one last aspect we need to consider here and that is that this is a situation where if we collect garbage data, we're going to get a poor result so the quality of the data depends on the way we drive the car. |
| 07:29 | And a little bit of common sense is required here. |
| 07:31 | First of all when we are gathering data into our histograms, we want to be very smooth with our throttle inputs. |
| 07:37 | This is going to avoid any instance where we end up with acceleration enrichment affecting our results. |
| 07:45 | It's also a good idea to stay away where possible from overrun fuel cut where the injection is disabled because this will read excessively lean so little bit of common sense there. |
| 07:55 | And the other aspect that is really easy to overlook when we are road tuning is we want to make sure that we allow any heat soak to dissipate before we start gathering data. |
| 08:06 | This is a real problem with road tuning because what we're doing is periods of driving the car, then we'll be shutting the car down where the engine bay is obviously hot, we've got no airflow so over the period where we're analysing data and making changes, everything heat soaks in the engine bay. |
| 08:20 | If we were just to restart the car, head straight back out on the track and gather data again, we're going to have a minute or 2 of unrealistic results so really important to give the car a couple of minutes of just normal road driving at a reasonable speed to get airflow through the engine bay and get rid of that heat soak. |
| 08:38 | So that's what we're going to do to get started. |
| 08:41 | Now we've got a bit of an idea of what we're going to do, let's head out onto the track and gather our first set of data. |
| 08:47 | Alright so while we are letting the heat soak dissipate we can also take the opportunity here to just get a bit of a feel for how the car's driving. |
| 08:56 | And again with a cam this is one of those areas that can be quite tricky to get dialled in. |
| 09:02 | Particularly as we've discussed during the worked example as the cam becomes larger we get into a sitiuation where we need to raise our idle speed enough that the car does push so this gives us an opportunity as well particularly in a carpark or when the car is stationary, to see how much push is being produced when we pull the car into drive and need to sit on the brake. |
| 09:26 | And this can be an instance where it's definitely worth having a bit of a play with the idle speed, see if you can get away with dropping it by 50 to 100 RPM, if you are having an instance where the car is pushing a little bit too much for your liking. |
| 09:40 | Alright so we're out on track now, we've got rid of our heat soak, we can make sure that our temperatures are at normal operating conditions, we're still a little bit hot in terms of our air temperature sitting at the moment 41°, 96°C coolant temperature but that's close enough for us to start gathering some data so what I'm going to do here is I'm actually just going to stop our scanner and we'll restart it and that will just get rid of any of that data that we've gathered while we were sitting stationary in the pits and while we've been heading out here on track. |
| 10:12 | Now just looking at the data now as we're driving and we're seeing the 3 cells that I've got access to at the moment, we're 2-3% so everything's starting to look pretty good already. |
| 10:24 | But we'll just continue through here. |
| 10:26 | Remembering being nice and smooth on the throttle and we want to make sure that we vary both our manifold pressure as well as our engine RPM as much as we can to sort of fill out that range that we're going to be spending the most time in when we're actually cruising. |
| 10:40 | So I'm going to do this for about a lap, we'll speed this process up now. |
| 11:32 | Alright at this stage we've got a reasonable amount of data captured so I'm just going to stop our scanner here and let's head back to the pits, we'll have a look at that data, see what we can take from that and what changes if any are going to be required. |
| 11:44 | Alright looking at our data at a glance, first of all everything does look reasonably good. |
| 11:50 | We're seeing for the most part our trims are within that plus to minus 2-3% range which is about as good as we can realistically hope for and after all, this is why we do have our short term and long term fuel trims there to pick up the pieces for small variations like this. |
| 12:06 | What we're looking for is any real outliers and about the only one that we've really got is this area out here where we've got a positive 8%, a 5.6% and then a negative 5.1% trim. |
| 12:16 | So these are all at very low RPM, in this case we're talking 1000 RPM and 80 kPa. |
| 12:23 | So it's an area that is difficult to get to at best and this could have been the result of flare on a shift or something of that nature so I wouldn't be taking too much from this just on face value. |
| 12:35 | What we can do of course is have a look at our count number and see how many hits we were getting. |
| 12:40 | So we do actually have a reasonable number, 92 in the particular cell that was sitting around 8%. |
| 12:46 | This one here though for example we can see we've got a count of 2 in that cell, another one here with only 2 in it and we do need to be mindful of this because if we've got a cell that only has maybe 50 or less hits, it's very hard to take that as reliable data. |
| 13:03 | We'll go back to our average though and so we can see that this cell here does look like we've probably got enough data points in it to be concerned about and of course if we do want to make any changes here we've got the same options that we've already looked at during the rest of the worked example, we can highlight this particular cell, we know that that's 80 kPa and 1000 RPM, we can come back to our editor and we can find that same cell, so 80 kPa and 1000 RPM, so we can take that particular cell there and we can make a change of 1.08 multiply that and that will apply that error. |
| 13:43 | We do need to be a little bit mindful of just making sure that we're not creating anything that's not particularly smooth. |
| 13:49 | What I'd probably do here, just looking at the surrounding numbers, I would probably also extend that particular change down a little bit as well, maybe not quite as dramatic, maybe we can add 5% just a little bit of hand smoothing there. |
| 14:02 | So this is the sort of work that we're going to be doing based on our results. |
| 14:06 | If we head back to our scanner though, I'm not really too worried about any of the data in there. |
| 14:12 | If I'm within plus or minus 2 or 3%, I'm personally pretty happy. |
| 14:15 | Depends how fussy you want to be but remember there's not a lot of point trying to achieve a perfect result because it's essentially impossible, there's always going to be some variation. |
| 14:27 | We can see there's an area down here that we didn't gather a lot of data so that's another area we might want to go back out and gather a little bit more data here, maybe extend ourselves out to sort of 3200 RPM or thereabouts. |
| 14:41 | When we are doing this, what we're trying to do though is gather the data in the places that we are actually going to be spending time when we're driving the car. |
| 14:48 | So there's not a lot of point being super fussy at 5000, 6000 RPM and 40 to 60 kPa in the vacuum area, we're simply not going to be there so realistically the VE table, there's not a lot of point dialling that in to absolute perfection there. |
| 15:03 | So at this point we're pretty comfortable with what's been going on with our VE table. |
| 15:08 | We've seen that everything's pretty much stacking up but while we're here the other aspect that we can look at is whether we've got any knock retard occurring and that's something we didn't really look at too much on the dyno when we were in steady state to what we'll do is come up here to our spark retard table and this is a histogram that just logs our knock retard activity. |
| 15:28 | And we can see everything's looking pretty good, we've got one cell here that has pulled 1° of timing. |
| 15:35 | And I'm personally again not particularly worried about that, that happened at 2600 RPM and by the looks of it, 0.52 grams per cylinder. |
| 15:44 | But if we do have repeated knock retard in there, particularly if we're starting to get significant numbers, 2, 3, 4°, in that case we can use this to help guide us with where abouts we need to retard our timing. |
| 15:56 | In this instance though, everything's looking pretty good with both our fuelling and our ignition timing under cruise conditions, car's driving nice, I'm pretty happy with everything, so we're going to go out now and basically replicate our full throttle ramp runs that we were doing on the dyno. |
| 16:11 | To do this we're going to use manual mode with our 6 speed automatic transmission. |
| 16:14 | I'm going to hold third gear and we're going to just accelerate from low RPM, as low as we can get in the RPM, down the longest straight we've got on our racetrack here and again we're going to gather data with our scanner while we're doing that so let's head out on the track and gather that data. |
| 16:29 | Now just like gathering our last piece of data, we've just come out of the pits, we've been sitting stationary for a while, we're going to just give the engine a few minutes here, or half a lap or so to let any potential heat soak dissipate so let's go through that now. |
| 16:59 | Alright we're coming up to the section of track that we're going to use here and when we are choosing the gear that we want to use, it's a bit of a fine line between choosing a low gear that will result in a low terminal speed but may result in too much torque multiplication and potential wheel spin with a more powerful car. |
| 17:17 | So we're going to try 3rd gear here, we're just coming up to our long section of straight and we'll see if this will let us get through to our rev limiter or close to. |
| 17:26 | So what we want to do here is just start down at around about 1500 to 2000 RPM, basically replicating what we saw on the dyno, just coming around this last corner now, I'll just roll smoothly onto the throttle with our scanner operating. |
| 17:58 | Alright so we've managed to gather our data there. |
| 18:00 | Quick glance at it, looks reasonably good but let's head back to the safety of the pits and we'll have a better look at that data and see what changes may be required. |
| 18:09 | Alright so let's have a better look at our results and first of all what we want to do is analyse the air/fuel ratio. |
| 18:15 | So if I just drag our chart logger all the way back up to the top, we're obviously looking at our commanded lambda vs our measured lambda. |
| 18:21 | Now we do have a little area here where we are a touch lean, if we click through here, you can see that probably from around about 2600 RPM through to about 3500 RPM, we are around about 3, in this case about 4% lean so that's something we definitely want to address. |
| 18:42 | For the rest of it though, we can see that we're mostly within about 1% of our target. |
| 18:46 | Although we are erring on the side of being a little bit leaner than our target. |
| 18:50 | So we'll see how we can address this. |
| 18:53 | While we're on our chart logger we can also see that we've got a little bit of knock activity and we've got a little bit of that background activity that I've already talked about. |
| 19:01 | Have got a point here though, 5600 RPM where we've got a degree of knock retard. |
| 19:05 | So that's something that I will want to address. |
| 19:08 | Anyway let's have a look at what we can do with these 2 aspects. |
| 19:11 | I will just mention as well with our fuelling we can see we've got this lean blip here but I'm not going to be too worried about this, you can see that this is as I'm tipping into the throttle. |
| 19:21 | It's not enough that I could feel this, this is almost certainly just an aspect of our transient enrichment which just goes to show how important it is when we're gathering that steady state data to make sure we're not being affected by our transient enrichment. |
| 19:34 | Alright so what we're going to do for a start is just refresh our minds here, 2600 RPM to 3500 RPM is where we're a little bit lean and we also want to check here our manifold absolute pressure, we're 95, 96 kPa. |
| 19:47 | So let's head over to our scanner, our editor again and we'll make those changes. |
| 19:53 | So we're about 95 kPa and we are 2600 RPM so I'll just find that zone which is there, through to 3600 RPM which we are there. |
| 20:05 | Now we're 95 kPa, 96 kPa but the changes I'm going to make here, I'm going to extrapolate out to the right here and in this case we were 3% lean in that area so I'm just going to multiply by 1.03. |
| 20:18 | Just come back to our scanner, we can see for the rest of the run we were still probably about 1, maybe 2% leaner than I'd like. |
| 20:28 | So again what I'm going to do is come back to our editor and I'll extrapolate that change, we're actually pretty good at 3600 and 3800, it wasn't until we got up to about 3900, I'll make this change from 3800 and above and we'll go all the way out to 6800 here and we'll make a change of 1.015, 1.5% there. |
| 20:50 | So we'll make that change there, head back to our scanner and what I'm going to do as well, we can see, we're probably a little bit lean down in that bottom end as well, about the same, about 1, 1.5% so we'll just extrapolate that change down further as well. |
| 21:05 | And what we can then do of course, the same that we've been doing right through our worked example, we'll copy the changes we've made to that table and we'll paste those into our other 2 tables, control C and control V, will allow us to quickly and easily do this. |
| 21:23 | Alright so from here what we would do is obviously flash that back in and confirm that the changes have given us the result. |
| 21:30 | It's iterative, it's exactly the same as what we've already looked at so I'm not going to go through that again. |
| 21:34 | Let's have a look at our ignition though. |
| 21:36 | So we'll head back to our scanner. |
| 21:37 | So we can see the knock retard there is occurring at 5600 RPM, 94 kPa but remembering it's grams per cylinder that we're actually interested in here. |
| 21:47 | Now what we can do is just slide these out of the way for a moment. |
| 21:50 | And we can come back to our spark retard table, histogram and we can see we've got our 1° occurring here, it's actually registered it at 6000 RPM and we can see that we were at 0.88 grams per cylinder. |
| 22:04 | So what I'd probably do here though is I would remove a degree of timing from 5600 and 6000 RPM just to be safe so le'ts go ahead and we'll see how we can do that. |
| 22:15 | We'll come to our high octane table. |
| 22:17 | So we were 0.88 grams per cylinder so we'll just make sure we are in the right zone there. |
| 22:24 | And we'll come across to 5600 and we actually may be better off making this from a 0.84 grams per cylinder. |
| 22:32 | Just basically removing a little bit of timing from around that area just for safety's sake. |
| 22:36 | So we'll highlight 5600 and 6000, 0.84 grams per cylinder and above and we'll make a change there, we'll press the minus key which will remove 1° of timing. |
| 22:48 | From here we'd flash that in and it's an iterative process just basically checking our work, making sure that we've made the necessary changes to our VE table to have our air/fuel ratio tracking our target nicely. |
| 22:58 | Remembering we're going to see some fluctuation but if anything we'd like to see that fluctuation erring on the side of a little bit richer than our target than our target rather than leaner which is what we're seeing there. |
| 23:08 | That being said, still for our first ramp run or test run out there on the road, we're pretty close based on what we saw on the dyno. |
| 23:14 | Then the same with our ignition timing, we want to make sure that we're not seeing repeated knock retard occurring in the same zones. |
| 23:21 | We will always get some sporadic knock retard over a long log file but what we're looking for and trying to prevent is consistent and repeated knock retard that's happening in the same zones, that's something that we will need to address. |
| 23:34 | So we're going to finish our worked example here. |
| 23:36 | If you do have any questions on this worked example, please feel free to ask those in the forum and I'll be happy to answer them there. |
