00:00 |
Ok, we can now move into the key step of this process , which of course is finding and defining the maps that we've already highlighted in the last step that we need to find for our stage one tune.
|
00:11 |
So, we're going to work through this essentially in a bit of a logic flow order, the EDC17 and we're going to start by finding our driver's requested torque table or tables as the case will actually be.
|
00:24 |
So, what we'll do is we'll start by moving across and having a look at a sample of what our driver requested torque tables will look like.
|
00:31 |
So, we can see this sort of sawtooth pattern here and of course we've got our axes here to the left of this table.
|
00:39 |
Also important to make sure that we've got OLS set up correctly so we want this in 16 bit, low, high and decimal.
|
00:48 |
So, we need to make sure we've got that so that we're actually going to essentially match the shape of this particular table.
|
00:54 |
Alright, so let's head back into OLS and we'll start by making sure we have in fact got that set up correctly and we can see that we do.
|
01:02 |
Also we may want to consider changing our magnification on our X axis.
|
01:06 |
I've got this set to 200%, if we haven't got it exactly set the same as our screenshot, everything's going to look a little bit more expanded or compressed, but this will be close enough for our purposes, so what I'm going to do now is simply skip through until I find something that matches that pattern.
|
01:23 |
And you'll see at the moment our cursor is set to just the left hand side of the map data or calibration data area so we don't necessarily know where this is going to be, can be time consuming but let's just start by clicking the left mouse arrow or button to the right hand side of our map data and that will just jump us a screen at a time.
|
01:43 |
So, let's go through until we find these tables.
|
01:46 |
Alright, so we've got at the moment four of these tables on our screen.
|
01:56 |
We can see that OLS has actually found the first two of these but not the remainder, we'll see why that's the case in a moment.
|
02:03 |
We can also see interestingly that these tables that have been found, one is 9x16 and the other's 8x16 so these aren't actually all the same size.
|
02:13 |
Now, if we just move across to the right a little bit further, we'll see we've got a few of these that repeat again, OLS hasn't found these.
|
02:23 |
Look like we've actually got two here that have all zeros in the tables as well.
|
02:31 |
If we just keep going I think we've got one more to the right as well.
|
02:34 |
Now, a key element when we're looking at these as well is the actual values in the table.
|
02:39 |
So, we can see we've got two more out here to the right, which OLS has also found.
|
02:45 |
So, the key aspect I was mentioning is the raw values or the height of these values.
|
02:50 |
So, we can see for example the tables out here, maximum values sort of look like they come out around about 9000.
|
02:58 |
If we come back across to the left here though, we've got another one here, kind of about the same height there, about 9000.
|
03:08 |
And then we've got all of these other ones that are much lower.
|
03:11 |
So, it's always worth just understanding what this might mean.
|
03:14 |
These are going to be requested torque tables, what we've got is three that look like they are significantly higher raw values than the others and these are probably most likely the tables that the ECU is actually going to be operating off under normal conditions.
|
03:27 |
These lower ones are probably more likely to be limp home mode driver requested torque or maybe for snow or something like that.
|
03:37 |
So, it's important just to have an understanding of what those raw values look like and what they actually mean.
|
03:43 |
Let's not get too far ahead of ourselves though, let's get into the actual map definition process and obviously it makes sense to start with the left hand most table which OLS has found.
|
03:55 |
So, I'll just double click on this and open it up and see what we're looking at here.
|
03:58 |
So, driver requested torque, we know that this is going to be engine RPM versus driver accelerator pedal position.
|
04:06 |
And if you're looking at these numbers, you should already be starting to see some familiarity.
|
04:12 |
For example that value 8192, we know this is one of our logical number sequences so of course if I divide 8192 by itself, we're going to get a factor of one or 100, it's just what we want obviously for accelerator pedal position.
|
04:26 |
If we look at our vertical axis, we know that this should be engine speed, obviously 10710 RPM, a little bit optimistic for any diesel engine.
|
04:35 |
The Bosch EDC17 controller usually uses a factor of two or divide by two to actually scale the RPM axis.
|
04:44 |
And then we've got values in the table itself, our Z axis data or our map data is of course our requested torque, the EDC17 controller requests a specific torque value in Newton metres.
|
04:57 |
Now, how we scale that is going to require a little bit of understanding of the torque the engine is going to produce but let's get into this anyway.
|
05:04 |
We'll start by looking at our X axis here, our accelerator pedal.
|
05:09 |
So, we'll give the map a description for a start.
|
05:15 |
Alright, so now we're going to set up our scaling factor here so we'll come down and use our little function button and we'll click on that and we will select our output.
|
05:24 |
Let's move this, actually we can see that just fine.
|
05:27 |
So, we know our factor is going to be 8192 or some combination, so if we put in 81, it gives us to a maximum value there of 101.
|
05:37 |
So, that's what we want so now it's going to be point decimal place and then 92, that'll get us to 100.
|
05:44 |
So, we can click OK there.
|
05:45 |
That's got our accelerator pedal axis scaled, we can always add a little bit of precision if required, not really essential in this case and we can click ok.
|
05:54 |
Now, we're going to click on our Y axis here, which again we already know is engine speed.
|
05:59 |
This is of course in revs per minute and in order to scale this we can either multiply here by 0.5, we can see that gives us a maximum value of 5355, pretty realistic there.
|
06:16 |
Or we can achieve exactly the same thing by again using our little function button and of course multiplying by 0.5 is the same as dividing by two.
|
06:25 |
So, two different ways of achieving the same result.
|
06:28 |
Again we can add some precision if we want but not really necessary.
|
06:31 |
Ok, our map data, so we need to understand here that we already know the stock engine is rated at 800 Nm, so we should be expecting numbers somewhere in the vicinity of 800.
|
06:45 |
And at first glance we don't have anything that looks remotely like that.
|
06:49 |
But let's just close this map down for a moment.
|
06:52 |
This really comes back to again what I was talking about with the height of these bars.
|
06:56 |
So, these ones we know were about 3400 give or take.
|
07:01 |
We can see this map , which is one of those three maps with the significantly higher values, we know this is around about 9000 so we can just have a quick look, we'll skip ahead, have a quick look at this one and we can see that yeah under full throttle, we're sort of starting with a target of 9000.
|
07:17 |
If we know that the engine produces 800 Nm, well 9000, if we divide that by 10, that's 900, that's a number that feels like it's probably in the ballpark.
|
07:28 |
We also know from Bosch description files that the factor they do use here is 0.1.
|
07:33 |
So, just important to understand that, because if we were only looking at that first map we looked at, just escape out of this and go back to it, it can be a little bit hard to understand what that factor would be given again this is going to be a significantly reduced torque request given it's probably a limp home mode or something similar.
|
07:51 |
Ok so, let's click into our Z axis data and we're going to give our map a description first.
|
07:58 |
Alright, we've given the map a name so what we'll do is Ctrl A to highlight all of that, Ctrl C, tab down and Ctrl V, we'll paste that into the description and of course this time we are looking directly at torque in Nm so we'll choose that as our unit.
|
08:18 |
Then we can come down to our factor and offset and what we want to do here is multiply by 0.1 and we can see that gives us maximum values here in this particular map of 340 Nm.
|
08:30 |
So, we'll click ok and that's our first map there now, found scale, defined correctly, everything's looking good.
|
08:38 |
Alright, let's close this down and we'll move across to our second high torque map.
|
08:43 |
Of course, we can do exactly the same process here and define this by hand.
|
08:48 |
Noting that the axes are slightly different here, we've got one additional row in our y axis here so we need to keep that in mind.
|
08:59 |
But let's close this down, we'll speed up the process by using the map properties and copying and pasting those.
|
09:06 |
So, let's close this down, go back to our first map so we can do that.
|
09:08 |
So, with our defined map now open again we can right click and we want to come down to copy map properties.
|
09:15 |
So, what we want to do is make sure that we are copying the name, description, units, the ID, the factor, offset, precision, etc and we want to make sure that we are copying this from the X, Y and Z or map axes.
|
09:29 |
To the right we can see we've got the option to also select our data source and address, we don't want to be doing that so we can untick that.
|
09:38 |
We'll click ok so that's copied everything there from this map that we want.
|
09:43 |
We'll press escape to close that down, we'll come back across to our next map and CTRL V will paste that in.
|
09:50 |
And an interesting change from the first map we defined, we can see that extra row for our engine speed there at 5500 RPM, the torque request is all set to zero, so this kind of acts a bit like a rev limiter of sorts.
|
10:04 |
Obviously if we go beyond 5250 RPM, we're dropping our torque request down to zero, hence controlling and limiting the engine speed.
|
10:13 |
So, that's our two first maps now defined, we'll escape out of that.
|
10:17 |
Alright, now we're going to define the remainder of these driver requested torque tables.
|
10:22 |
And we know that OLS hasn't found these remaining tables despite the fact that on face value they do look really similar.
|
10:29 |
It's because of a small inconsistency in the axis data.
|
10:33 |
So, what we're going to do is use the map data but we're going to copy the axis data from our existing map here.
|
10:41 |
So, what we can do to start with is just click somewhere near to the end of our data.
|
10:46 |
And we'll convert here to text view.
|
10:49 |
So, this is where we are at the moment.
|
10:51 |
We know that this is an 8x16 table so what we want to do is make sure that we've got our width correctly set , which we by fluke have actually got that set to 8.
|
11:02 |
Otherwise we can use the M and W keys to increase or decrease that.
|
11:07 |
Remembering what we're looking for is the visual pattern out on this right hand side.
|
11:12 |
So, we can see at the moment with the width incorrectly set, we've got this diagonal sort of trend here so if we bring this back down by using the W key, when we get to 8, we can see everything jumps into alignment, we've now got this nice vertical line, that's what we're looking for.
|
11:29 |
We can also see that the shape of the defined table up above is nice and consistent, that's what we want to see, we definitely don't have that in this data so what we want to do is use the Control key and the left or right arrow keys and just move along until we've got everything into alignment there, which we've got.
|
11:50 |
Alright, now what we're going to do is come to the top left corner, the information above this is essentially map axis data so this is our first value, we know that this is 8x16 so we're going to hold down the shift key and we'll come across 8 so 1 2 3 4 5 6 7 8 and then down 16 so 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 and we'll press K to mark that table.
|
12:16 |
So, straight away we can see the values here make a bit of sense, 3399, essentially what we saw in that first mathem, as we've already discussed we know that this is most likely going to be some kind of a limp home map, so it's not really that critical to us anyway, but let's have a look at what we can do here.
|
12:35 |
So, we'll come back to the last map, which is the same size, 8x16, and what we're going to do is copy our map properties.
|
12:43 |
So, this time what we want to do though is also copy the axis address.
|
12:48 |
So, we're going to use the axis data from this map and we'll just copy that into our remaining maps.
|
12:55 |
Alright, we come across to our data and this time we want to tick the little box here for our data source plus address.
|
13:02 |
And we want to copy that from , which areas? In this case of course just the X and the Y axes, we don't want to do this for the map data so we'll click ok.
|
13:13 |
Alright, we can press escape to close that down and we'll come back to our next map and Ctrl V and that applies all of the information we need so we've now got axis data that makes sense, exactly the same as what we have in our other defined maps and of course our torque values in here also make sense.
|
13:33 |
I should mention at this stage we could also add some precision with another decimal place here in our Z axis torque data.
|
13:41 |
Personal preference here, doesn't really make too much of a difference and you can always come back and change that at a later point.
|
13:47 |
Alright, so at this stage we have now found three of our maps, we're going to go ahead and repeat this process.
|
13:55 |
So, we can see that we've got exactly the same data out here so same sort of deal goes here, we'll click somewhere near the end of that data range, we'll go to our text view and we can see, we just move down a little bit here, we can see we're already set up on the width of eight , which we know is what we require so we're going to use our Ctrl and arrow keys here just until everything comes into alignment there.
|
14:20 |
We've got that done now.
|
14:22 |
We'll come up here to our first value and we're going to go across 8 by 16 down.
|
14:26 |
Press K to mark that table, all the values look right, Ctrl V will paste the information again into our, or our map properties again into our new map, press escape, job is done.
|
14:40 |
So, we're just going to repeat this process here and I'll jump ahead, because it is just exactly the same that you've seen.
|
14:47 |
So, let's jump ahead until we've got all of these tables marked.
|
14:58 |
Alright, so we've now got seven of our tables marked over here, we can see that those are shown in our My Maps folder.
|
15:04 |
Now, I did mention that we do potentially have two more here, I'll just scroll over so we can see these.
|
15:13 |
Notice that we've got this gap right here in the middle of the tables that we've just defined and we've got those same similar looking axis values.
|
15:21 |
The fact that these are all set to zero, again obviously we're not going to be using these when the engine's running so little point in actually going ahead and defining those two tables as well.
|
15:32 |
So, seven of our driver requested torque tables are found and defined, what we're now going to do is just put these into a folder to make it a little bit neater as we go through and define more of these tables.
|
15:41 |
So, in order to create a folder, you'll remember from the body of the course what we're going to do is come across and we'll right click on the first of these tables and we want to click on copy name.
|
15:51 |
We're going to then do exactly the same, right click again and we're going to go to the new folder option.
|
15:58 |
So, Ctrl V will just paste that name that we copied so driver requested torque just makes everything a little bit quicker and easier than typing that in yourself.
|
16:07 |
Then what we want to do is click on this little button here, so if the name contains the following text, of course Ctrl V, that will automatically put these into the folder and we can click apply rules, click OK and now we've got a nice neat folder with our seven driver requested torque tables.
|
16:26 |
Alright, now that we've got our driver requested torque tables found and defined, we'll move on and the next logical set of tables that we're going to need to find are our torque limitation tables.
|
16:36 |
The Bosch EDC17 controller actually has a range of torque limitation tables as well as some single value torque limits, which can be quite tricky to find.
|
16:46 |
We're going to start here with our torque limitation versus barometric air pressure.
|
16:51 |
Get a sense of what that actually looks like.
|
16:54 |
This is the shape of these tables.
|
16:56 |
So, we can see we've actually got two of them, one here and another one here.
|
17:00 |
So, that's the shape that we're looking for and this is again in 16 bit low high decimal, so we need to make sure that we are set up like that.
|
17:11 |
Alright, so let's go back to OLS here and we'll scroll through and try and find those tables.
|
17:18 |
At this stage of course I don't necessarily know whether these torque limitation tables are going to be to the right or the left of where we left off with our driver requested torque, but let's move to the right here and see what we can find.
|
17:32 |
OK so we've got our two tables there and we can see that the OLS software has actually already found these so let's click on the first of these and we'll open it up.
|
17:48 |
Alright, let's have a look at our raw data for the moment and we know that this table is going to be engine RPM versus barometric air pressure.
|
17:55 |
The map data of course is going to be our torque limitation in newton metres.
|
17:59 |
Looking at the data straight away we can see the y axis here, 700 through to 950.
|
18:04 |
These numbers suspiciously look like barometric air pressure raw with no scalar, 950 this would be millibars.
|
18:12 |
Then we can see we've got our horizontal x axis here ranging out to 10,000, this is going to be our RPM and then our newton metres in our map data.
|
18:23 |
So, let's start scaling this, we'll start with our x axis here and we're going to of course call this engine speed, units of revs per minute.
|
18:34 |
And again with the EDC17 controller our RPM is typically going to be multiplied by 0.5 or alternatively divided by two, same result.
|
18:45 |
Let's click ok and remembering again we can move around between the x, y and map data and the axes inside of that little box are just closed down, just for simplicity, particularly while you are learning these skills to make sure you don't make any mistakes, it's always easiest to do this by clicking directly on the axis that you want to change.
|
19:02 |
So, we just double click on that.
|
19:03 |
So, our y axis here is barometric air pressure.
|
19:07 |
And we will use a unit here of millibar.
|
19:12 |
And again this is a little unusual but we are not applying a scaling here, this is a raw value, so we'll click ok.
|
19:19 |
Now, we can double click on our map data and give our map a name, this is torque limitation versus baro.
|
19:28 |
Again control A will copy that or highlight that I should say, control C will copy, tab will move down and then control V will paste that in.
|
19:37 |
So, our torque limitation here, our units are in newton metres of course and we already know here that we need to multiply this by a factor of 0.1 or divide by 10, it's going to give us the same result, 0.1 in here and now we can see the result here.
|
19:54 |
So, this is important to understand because this particular barometric air pressure torque limitation table, we can see that the peak values in here through the midrange are sort of in the high to mid 700 newton metre values.
|
20:08 |
If you'll remember that our driver requested torque peak value was 900 newton metres, straight away we can see that this table along with some others is actually going to be limiting our torque below that.
|
20:19 |
So, we can actually make quite big improvements in our torque just by essentially matching the max requested torque to our torque limitation values.
|
20:27 |
Of course, we do need to consider engine strength as part of this, that's not necessarily going to say that that's going to be a great idea for every application, but we can just see how these torque limitation tables can actually go over and above what we're actually requesting from the driver's accelerator pedal position.
|
20:44 |
Now, the next table is going to essentially be the same as this so before we close this down, we'll right click here and we're going to copy our map properties, so we don't have to do all of this again.
|
20:54 |
This case we don't want to be copying our data source and address, so we'll untick that little box here, we're just going to be copying our name, description, ID, factor, offsets etc and we want to make sure again we're copying these from X, Y and Z axes.
|
21:09 |
We'll click OK, press escape to close that down and because OLS has found the second torque limitation versus barometric air pressure map, we can just double click on that, Ctrl V and our job is done there.
|
21:21 |
So, that's our two torque limitation versus barometric air pressure maps found and defined, but our job is not complete just now.
|
21:30 |
We've also got a range of torque limitation tables versus gear so let's have a look and see what these tables look like.
|
21:38 |
Alright, so what we can see is this is the area where our torque limitation versus gear tables are, we've got this sort of jagged sawtooth appearance to it.
|
21:47 |
So, we're now going to scroll through in OLS and keep going until we find these.
|
21:54 |
Alright, we've found a section of the map data that has that same matching pattern here.
|
21:59 |
And by looking through this we can see a pattern that kind of repeats and it looks like we've got 1, 2, 3, 4, 5, 6, 7, 8, 9, potentially 10 of these maps.
|
22:11 |
To make our life just that little bit more difficult, OLS clearly as we can see has not found and marked these maps, so it's going to be all up to us.
|
22:20 |
So, what we want to do is find out for a start how big these maps are and again just to add a little bit of complexity, they aren't all the same.
|
22:29 |
These are 2 dimensional maps, they only have one axis, which in this case is going to be RPM and then the map data of course is our torque limitation.
|
22:37 |
So, let's just start with the very start, our left most map here.
|
22:43 |
And what we'll do is we'll scroll across so the flat section that we can see just in here, in fact let's just make this a little bit easier to see, we'll zoom in a little bit.
|
22:54 |
This flat section in here is going to be our torque or map data.
|
23:00 |
So, we can see if we're looking at the values here with the cursor, 8000, we already know we've got a 0.1 multiplier there, divide by 10 so 800 newton metres we know, we can sort of see those numbers, that'll make sense.
|
23:14 |
Now, if we just use our left arrow key and move across, we see we have another value of 8000, nothing's changed, we'll move across one more time, we've got 8000, nothing's changed.
|
23:23 |
Now, we move to a value of 10,000, again we'll move to the left, a value of 1000 and then finally a value of 2.
|
23:30 |
If we go to the left again, we jump back up and things have all changed.
|
23:33 |
So, what we want to do, and it's a little bit harder when it's such a small table, what we want to do is take note of this 2.
|
23:39 |
This is the axis size so in this case it's a 2x1 table.
|
23:43 |
And we know that after this 2, the next 2 values are going to be our X axis values.
|
23:48 |
So, looking at those again, 1000 and 10,000, finally that will be our map data, which we know is going to be too long as well, 8000 and 8000.
|
23:58 |
So, we know that's likely to be our first table.
|
24:01 |
Let's just jump to our text view and see what that all looks like.
|
24:05 |
Ok so, at the moment the width of our text view is completely unhelpful for what we're trying to find, we've got that set to 2.21 so we'll just use our W key, narrow that down until we get to our width of 2.
|
24:19 |
So, we've got here our value for our axis 2 so we know that the next 2 pieces of data are going to be our axis values and then this is going to be our first actual map data point here, so what we'll do is use the CTRL and arrow key to move that across and we'll highlight that and we'll press the K key to mark that axis.
|
24:43 |
So, we've got our first table found here, we know of course engine RPM as we already discussed so let's double click on the X axis and we're going to call that engine speed.
|
24:54 |
We're going to give that a unit of RPM and we are going to then apply our multiplier, in this case again we know that 0.5 is the multiplier that EDC17's used so now our data will be sensible, 500 to 5000 RPM, we'll click ok, we'll double click into our map data and we want to call this torque limitation by gear, so let's go ahead and do that.
|
25:26 |
Alright, we've given it a title, we'll CTRL A, CTRL C to copy that, tab down to our next line for our description, CTRL V and the units of course here are newton metres.
|
25:37 |
And again we already know that our multiplier here is 0.1 so we've got numbers that make sense.
|
25:45 |
Now, the key to this as well is seeing that the limitation here is 800 newton metres, goes back to what we were saying with our previous tables, the torque limitation values are below what we were actually requesting with our driver requested torque, so this table here would come in and sort of spoil all the fun if we're trying to make more torque.
|
26:03 |
Let's click OK there, we've defined the first of those torque limitation by gear tables.
|
26:09 |
What we're going to do though is right click here and copy our map properties.
|
26:14 |
Again all we want to do at this stage is copy our X, Y and Z axes, obviously we don't have a Y axis in this particular case, we could unclick that, we'll click ok.
|
26:25 |
Let's press escape to close that down, we'll go back to our two dimensional view and we can now carry on and define our next table.
|
26:34 |
What we're looking for here is essentially this repeated trend or shape, so if we move across to the right we can see that we've got another piece of data here that looks the same and we can see that our cursor values that I've just clicked on there, 8000, again it's all making sense, this looks like a bit of a rinse and repeat of the table we've just found.
|
26:54 |
Let's move to the left, still got a value of 8000, we'll move to the left one more time, still got a value of 8000, left again, got the same table axes here now, 10000, 1000, if we go one more to the left, we're probably going to see this is going to show us a size of two, so basically exactly the same as the table we have just found.
|
27:14 |
Let's switch to our text view and have a look at this.
|
27:17 |
So, there's our two, that's our axis size, we know that again the numbers to the right of this is going to be our x axis so let's just use the Control key and use our cursor to just get everything aligned there, so here is our axis, 1000 and 10000 and here is our data ,so we'll just highlight our data, mark the table using our K key and then Control V and we've just sort of sped that whole process up.
|
27:45 |
So, now we've found two of those tables.
|
27:47 |
I did mention however that they're not all 2x1 tables, there is a little bit of complexity so let's jump back to our 2D view and we'll carry on.
|
27:55 |
Alright, so in our two dimensional view, if we're looking for that same repeating pattern, it looks like we've got something similar out here so let's have a look at that.
|
28:04 |
And again yep we've got those values of 8000, so we're already sort of prepared that we're looking for these and as we keep coming across, now we jump up to that 10000 again, we should be recognising this but this time if we go to the left again, we see now we drop down to 4300, 4100, 3000, 2000 and finally we get to a value of 5.
|
28:24 |
So, this one's a bit of an outlier, this means that the width of this table, it's a 5x1, not a 2x1 , which we've been dealing with so far.
|
28:32 |
No problem though, let's jump across to our text view and we'll see how we can deal with that.
|
28:36 |
First of all of course we want to change our settings here so that we have a width of 5.
|
28:42 |
We can see we are currently set up with a width of 2 so we can just press the M key 3 times and that'll bring us out to our correct width.
|
28:50 |
So, we are currently on this value here, 5, again we know that this is telling us that the axis is 5x1 so the values to the right of this are going to be our axis data values.
|
29:04 |
So, 1, 2, 3, 4, 5 and then our data will start here, they're all 8000s, no big surprise here, hold down the shift key and we'll highlight those, press the K key to mark that table and then if we press CTRL and V, again that's going to apply our scaling and our data for our axis there, job is done, press escape and that's going to close that down, let's head back to our 2D view.
|
29:30 |
Now, interestingly that is the only 5x1 table, now we're back to 2x1 tables, let's just move ourselves across to the right here and we'll have a look at the remainder.
|
29:40 |
So, again we've got this same repeating pattern.
|
29:42 |
So, let's click over here to the right, we've got 8000, 8000, we've got our 10,000, 1000 and 2 so exactly the same as what we've already seen.
|
29:54 |
Let's switch across to our text view, we're going to use our W key now to narrow back down to a width of 2, we've got our axis data here and here and then we've got our map data here and here, so we'll shift and arrow key to highlight those, K to mark the table, CTRL V and our job is done.
|
30:15 |
This really is just a rinse and repeat so we'll just speed this process up and get the rest of our tables here marked for our torque limitation versus gear.
|
30:33 |
I will just pause here because this particular table differs from all of the others that we've just defined in that we can see this actually has a 650 newton metre limit as opposed to all of the other tables , which are set to 800.
|
30:46 |
Now, without diving a lot deeper into the description file, I can't specifically say why this table is any different, it might be some kind of limp home mode, maybe it's limitation for maybe reverse gear or something of that nature, but we do need to understand that this one is different and that would be something that we'd need to take note of when it comes time to actually tune.
|
31:08 |
Let's press the escape key to close that down and we'll go back to our 2D view here and we can see if we move further to the right, we lose that pattern, so we'll just close that, narrow the view down a little bit and we can see we've got all of our torque limitation versus gear tables now defined.
|
31:25 |
Now, we're not quite done with our torque limitation yet, we've also got to find our single value torque limit and this can be tricky, because it is just a single value.
|
31:35 |
What this requires is either a knowledge of where to find this, which is what we're going to be using, or alternatively a description file.
|
31:44 |
Now, we do know that in the EDC controller, this single value torque limit is placed just to the left of our gear based torque limits , which we've just gone through.
|
31:54 |
So, jumping back into our view here, let's just get a little bit more detail and what we'll do is we'll move across to the left here.
|
32:02 |
So, this is the first of those gear based limits that we found and defined and if we place our cursor just to the left here, and what we want to do is look down at our value at the cursor here, this is currently 2, we already know this was the size of our table.
|
32:17 |
If we move the cursor to the left one, we can see we've got this value 8500, which we should also recognise already because we saw that inside of our gear based limitation tables.
|
32:28 |
So, this is our single value torque limit and we now need to actually turn this into a table.
|
32:36 |
We can have a look at this in TextView, got our value there of 8500, of course it is a 1x1 table if you want to think of it like this, so what we need to do is use our W key and narrow our view down and now that we've got it down to a single value, we can press the K key and mark that.
|
32:54 |
So, we're going to call this single value torque limit.
|
32:59 |
And as usual, Control A, Control C, tab and Control V, we can paste that down and we'll select our units in newton metres and we'll apply our same scaling of 0.1, 850 newton metres, everything's looking good.
|
33:14 |
We can now press escape to close that down and essentially at this point we've found all of the torque limits that we're going to need for our stage one tune.
|
33:21 |
The only thing that really remains here is a little bit of a clean up just to add some folders to make everything look a little bit neater.
|
33:28 |
So, let's go ahead and do this now.
|
33:30 |
So, start up the top with our torque limitation versus barrow, right click on that and we want to copy name, right click again and we're going to come down to new folder.
|
33:41 |
And we're going to Control V to name that folder and then we will add the name again and it will automatically apply that rule and put all of those into a folder.
|
33:53 |
We've got our single value torque limit, we can put this in a folder, it is only one single table or value, but for the sake of completeness let's do that.
|
34:03 |
So, we'll copy our name, right click again, new folder, Control V and Control V again, apply that rule, ok.
|
34:14 |
Now, we'll do the same thing for torque limitation versus gear, so again same process here, copy our name, right click, new folder and then Control V, Control V again, apply that rule, put everything into our folder.
|
34:32 |
Ok now, for the sake of completeness we're going to create another folder called torque limitation and we'll put all of those folders into it.
|
34:39 |
What we'll do is we'll come across to MyMaps, right click here and we'll create a folder and we'll call this torque limitation.
|
34:47 |
We'll click ok to create that folder.
|
34:50 |
Now, that we've created that folder what we can do is simply drag and drop these other folders into it.
|
34:57 |
And once we've completed that, if we expand that folder out we'll see we've got our other folders within.
|
35:02 |
So, now we've got everything nice and neatly arranged we can move on and find our next map.
|
35:06 |
We're going to follow the logic flow in the ED17 controller in the next set of maps that we've defined and define are our NM conversion tables and these simply take our requested torque after we've gone through our torque limitation tables and then convert this into an injection quantity, essentially how much fuel do we require in order to create this requested torque.
|
35:28 |
If we look at our laptop screen here, this is the general shape that we're looking at for this table or set of tables and we can see it's got a very obvious shape with this sort of hollow in the middle and we can see here that we've got five of these tables that we're trying to find.
|
35:44 |
So, should be really quite easy for us to identify these in the map.
|
35:47 |
Let's head back into OLS now.
|
35:50 |
So, we're leaving off from our single value torque limit here and again we might not necessarily know whether these tables are to the left or the right of our current cursor position, we can see we're sort of somewhere in the middle of the calibration data.
|
36:04 |
So, they should be to the left but either way we're just going to go from one side to the other until we can find it.
|
36:09 |
So, what we'll do is we'll jump back here until we find this set of tables, again should be very obvious when we see them.
|
36:14 |
So, 10 screens to the left of where we were with our single value torque limit, we've found these tables, they're quite clear and quite obvious and we've got five of them already identified by OLS.
|
36:28 |
And what we can see is that these are different sizes, the first one on the left is a 16x16 table, the remainder are 17x16.
|
36:36 |
I'm going to move a little bit out of order here just for a bit of clarity so we'll actually start by double clicking on our second table and we'll define this.
|
36:45 |
To start with we really need to understand what is this table representing.
|
36:50 |
Well, it's NM conversion, so one of the inputs to this table is going to be our requested torque and newton metres.
|
36:57 |
This is a three dimensional table so our other axis is going to be engine speed and then the output of this table is going to be our requested injection quantity and this is in milligrams per engine cycle.
|
37:10 |
Now, it's going to be a little bit tough, if you're not familiar with tuning diesel engines, sort of having an understanding of what kind of magnitude you'd expect to see from injection quantity, this is something that comes with time, we're going to help you there a little bit as well.
|
37:23 |
So, looking at this table we can see our X axis ranges up to 9000 and this should already stick out, because we know that we've been dealing with torque values up to around that 900 newton metre vicinity.
|
37:36 |
So, there's a bit of a chance there that we are looking at our torque request.
|
37:40 |
However, we've also got the Y axis going up to 9000 as well.
|
37:45 |
In this case this is more likely, given the values in it, to be our engine RPM.
|
37:50 |
So, let's start with that engine RPM, we'll double click on it and of course give it a description.
|
37:58 |
The units for this of course are going to be in revs per minute and we already know that the multiplier we've been using in the EDC controller for engine speed is multiplied by 0.5.
|
38:09 |
We'll do that and yeah these numbers now make sense, 580 up to 4500.
|
38:14 |
We can now click ok and close that down, let's deal now with our X axis , which we're expecting here is going to be our requested torque.
|
38:25 |
And of course our unit for this is newton metres and as usual our multiplier, 0.1, just the same as what we've been using for the remaining factors.
|
38:35 |
We can give this a degree of precision as well if we decide we want to.
|
38:39 |
So, everything's looking quite good there, we can click ok and close that down.
|
38:43 |
Now, if you're still a little confused because the maximum value for both of these axes was 9000 in raw value so I've sort of jumped ahead and said that Y axis is RPM and the X axis is torque request, but you can always get a bit of a sense for this as well.
|
38:59 |
We can see that our torque request value runs at the moment from 0 out to 900, so it sort of ticks the box of what we're expecting.
|
39:06 |
Let's just double click here on our RPM axis and if we were to apply that same scaling of 0.1 for torque, we can see that this doesn't really make sense.
|
39:16 |
Yes, the maximum values do stack up, but we sort of start this axis at 116 newton metres of torque , which isn't realistic.
|
39:24 |
So, just a little bit of a sanity check there if you're a bit confused.
|
39:27 |
Do these numbers make sense? In this case these only make sense as RPM, so we'll click ok.
|
39:33 |
Now, we've got our injection quantity to deal with so let's double click on our map data and we're going to call this NM conversion.
|
39:43 |
Control A, Control C, we'll tab down and Control V to fill that in.
|
39:48 |
And we are dealing here with milligrams per engine cycle.
|
39:53 |
So, again not necessarily going to know the magnitude of these numbers, but in this case I do know that we're going to end up with a factor here of 0.01 and in this case we definitely would want to add a couple of degrees of precision to this.
|
40:08 |
The numbers here, again if you are familiar with tuning diesel engines, these are going to make sense as our injection quantity values.
|
40:15 |
So, we can click ok here and we've defined our first table.
|
40:19 |
Well actually technically it's our second table, but we'll get to that.
|
40:23 |
Before we close this down, let's just right click and we'll copy our map properties, making sure that we are copying from our X, Y and Z axis data, we'll click ok, escape, we'll close that down, we've got our first table defined.
|
40:37 |
We'll come across now to our first table, the left one, and we'll double click on that.
|
40:42 |
Now, the reason I did go a little bit out of order here is if we actually look at the numbers on the axes here, this one could be a little bit trickier to actually work out.
|
40:51 |
Given that the maximum value on our Y axis, which we now know is torque, actually only goes out to 5500, which is going to scale to 550 newton metres.
|
41:01 |
So, nowhere near the 850 to 900 that we're expecting so this could throw you off.
|
41:06 |
And likewise we've got values out to 8800 on our RPM axis.
|
41:11 |
Now, I'm not 100% sure why this is a different table, this could be for some kind of limp home functionality or a different mode, however the table is essentially still exactly the same, so CTRL V will paste our description and our scaling data into this, we can press escape, close that down and of course we're going to go ahead and do this for our remaining three tables now.
|
41:36 |
Ok so, there's our five NM conversion tables found and defined.
|
41:41 |
Again, what we're going to do now is just put these into a folder to make it nice and neat and tidy so we're going to click on our first of those, we're going to right click and we're going to copy name and we're going to right click again and we're going to create a new folder, CTRL V will name that folder and we're also going to automatically copy all of the existing tables that we just found into that and our job is done.
|
42:07 |
The next set of tables we're going to find and define are our smoke limitation tables and there's two types of table here, we have a single table, which is smoke limitation by manifold absolute pressure and then we've got a set of tables, which are smoke limitation via lambda.
|
42:24 |
So, let's start by having a look at the pattern that we'll be trying to search for.
|
42:28 |
So, this is the first of those tables, this is our smoke limitation by manifold absolute pressure.
|
42:34 |
So, this is the sort of trend that we're looking for, axes to the left here.
|
42:38 |
So, let's head back into OLS and we'll go ahead and find that.
|
42:43 |
And again we could be looking to the left or the right here, I know that we're going to be a reasonable distance to the right of our cursor at the moment, so let's get ahead and find that now.
|
43:01 |
Alright, so looks like we've got a pattern here that looks the same, we can head back across and just have a look.
|
43:08 |
And yep that looks exactly the same here.
|
43:10 |
So, we know that we've found our smoke limitation by map table and we can see obviously OLS has also identified this as a 13 by 16 table, let's double click on it and see what we're looking at here.
|
43:22 |
Alright, so the axes on this are going to, no surprise, be manifold absolute pressure on one axis and engine RPM on the other.
|
43:31 |
Let's just start by having a look at our engine RPM axis.
|
43:34 |
We already know that this is going to be multiplied by 0.5, so let's go ahead and set that up.
|
43:43 |
And that all looks good, we've got our axis scale from 600 to 4000 RPM , which is starting to make sense.
|
43:50 |
That obviously means that our X axis here is our manifold absolute pressure.
|
43:55 |
Now, here this is defined in millibars so there is actually no scaling applied, this is a one to one, which is nice and convenient for us so let's go ahead and enter our description here.
|
44:12 |
And lastly we've got our map data to define, so we'll double click on our map data and this of course is smoke limitation by map, so we'll enter that now.
|
44:26 |
The units for this, the actual axis data here is our fuel mass limitation, so we will be using here milligrams per cycle for our axis and the value here, the scaling factor is a factor of 0.01.
|
44:42 |
So, as usual, it's not necessarily going to be that you're going to be able to guess these numbers, once you're a little bit more familiar with the magnitude of the numbers that you're going to be seeing for fuel mass, this does make sense, we've got values there scaling up to about a maximum of 80, 82 milligrams per cycle.
|
45:00 |
Nice to add a little bit of precision here, so we'll add two decimal places, click ok and that's the first of our smoke limitation tables found and defined.
|
45:10 |
Now we're going to be looking for our smoke limitation by lambda tables and we've got seven of these tables and these can be a little bit tricky to find.
|
45:19 |
They also share an axis, which we're going to find at the end of the table so a little bit trickier than what we've done so far but if we jump across to my laptop screen, this is what we're looking for.
|
45:30 |
And again quite a subtle shape here, we've got, this is actually the last of these maps and then the one to the left.
|
45:38 |
Looking at the numbers, which we'll see, these are probably most likely the two of the seven maps that are actually utilised, but for the sake of completeness, we will go through and find them all.
|
45:49 |
So, that's the shape we're looking for.
|
45:51 |
What we can't really see here is that out to the left, we've got this just a straight line essentially and this is the other five maps.
|
45:59 |
So, we're looking for this long straight line followed by these two maps, these little bumps, which we can see.
|
46:06 |
So, let's head back into OLS and we'll see about finding those.
|
46:10 |
And again I am cheating a little bit, I know that these maps are going to be to the right of our cursor position, so let's move through and see what we can find.
|
46:20 |
OK so this is the shape we're looking for here, we can see we've got one of those maps right here and we've got another one here to the left.
|
46:28 |
So, actually not too far from our smoke limitation by map, which kind of really makes a bit of sense.
|
46:34 |
So, let's go back a little bit and we'll see what I was talking about here with this long straight line.
|
46:39 |
So, we've actually got five maps that we're going to have to find in there, but let's just start with the last of these maps, the last two of these maps and we'll see what we can find.
|
46:51 |
So, what we actually know is that this is going to be the axis here for all of these maps, so it's going to be a shared axis and if we look at the bottom number of these, we'll come over here and if we look to the left, so we've got 16 and then we go to the other axis and again we've got 16.
|
47:12 |
So, we know that this is going to be a 16 by 16 map.
|
47:15 |
And OLS has actually identified this as the axis for the map to the right.
|
47:21 |
So, what we're going to do is right click on this and we'll delete that potential map.
|
47:26 |
So, let's come over here to the end of our data area and we'll click on our text view here.
|
47:34 |
So, we already know that our table is 16 by 16 , which of course means that we need to set our width to 16 and in this case we're already starting there.
|
47:43 |
If not of course the W and M keys will get our width set to 16.
|
47:48 |
So, what we want to do for a start is just get our alignment of our data area, which we can see graphically here so that that all makes sense.
|
47:56 |
So, we'll just use the control and right arrow key, or left arrow key I should say, and move that all over so that that all looks how it should.
|
48:04 |
Now, the number that I'm on at the moment, the one I've got highlighted, this is the size of the axis.
|
48:10 |
So, our axis starts straight after that so our data area is immediately above this.
|
48:14 |
I'll click on that bottom left corner of our data here and holding down the shift key we'll come across 16 and then we'll come up 16 as well.
|
48:30 |
And then we'll press the K key to mark that map.
|
48:33 |
So, at the moment we've got no axes and the numbers in this are going to be lambda values.
|
48:38 |
So, looking at these and knowing diesels we're always running on the lean side of stoic.
|
48:44 |
So, looking at these, the majority of the map has values of 1300 in here.
|
48:49 |
And this makes sense because 1.3 lambda is probably about as rich as we'd typically run before we start seeing noticeable smoke.
|
48:57 |
So, we can double click into this map here and we're going to give it a name for a start, smoke limitation by lambda.
|
49:07 |
And again as usual, control A, control C, we'll tab down, control V, and of course our units here will be lambda.
|
49:15 |
And our multiplier here, our factor, let's just see how this looks.
|
49:20 |
If we enter a value of 0.1, no, 0.01, 13, 0.001.
|
49:29 |
Now, I've got just a table full of ones, but of course this is going to make a bit more sense when we start adding a little bit of precision so we'll add three decimal places there, click ok.
|
49:40 |
And there we've got our table set up and this again all makes sense.
|
49:44 |
If we come back to our text view here, we can see at the moment our cursor is highlighting this value of 16, which we know is the size of the axis.
|
49:51 |
So, the number directly to the right of this, this is the first value for in this case the y axis.
|
49:58 |
So, we'll right click on that and we will set that as address as y axis of the previous map.
|
50:04 |
We've got this other value here, 16, at the end of that axis.
|
50:09 |
That means that our next axis is also 16 long, we'll click on that value, right click on that and we'll set that as address of x axis.
|
50:17 |
Now, what we can do is double click on our map and we've now got our axis values available, of course we do need to still scale them.
|
50:25 |
Let's start with the simplest one, which of course is in this case our RPM axis on the y axis.
|
50:32 |
We'll double click on that and we'll give that a name.
|
50:35 |
And of course our unit here of revs per minute.
|
50:39 |
Now, as usual, the factor here for scaling our RPM in the EDC17, which we've already seen, 0.5, 500 to 4000, all making good sense, we'll click ok.
|
50:50 |
Next, we can double click on our x axis and this is our mass airflow and in Bosch EDC terms this is defined in units of milligrams per hub.
|
51:02 |
And this is a difficult one, you're not necessarily going to guess this, you're going to need a little bit of experience to sort of know the range of numbers that we'll see here, but actually if we close this down for a minute and we scroll across to the right we see we've got maximum value of 13000.
|
51:16 |
Now, the actual scaling for this is multiplied by 0.1, so description file will obviously give us this information or if you've got experience in seeing the sort of scale numbers or log these values, you'll also get an indication of what these sort of values should range to.
|
51:35 |
So, let's just enter that data now.
|
51:42 |
And if we change our scale to 0.1, we're now scaling out to a maximum of 1300 and we can click ok.
|
51:51 |
So, that's our first map defined there.
|
51:54 |
Now, what we're going to do as usual here is right click and copy our map properties.
|
51:58 |
And this time we want to also make sure that we are copying our data source and address for our X and our Y axis, remembering that the axes are going to be shared with our remaining tables.
|
52:11 |
So, that's done, now we've got that copied to the clipboard.
|
52:14 |
Now, what we're going to do is find our next map, if we go back to our 2D view, we can see that that starts somewhere around about here.
|
52:23 |
And again, there is a table that has been found, we'll just delete that potential map.
|
52:28 |
And let's just go back to our text view here.
|
52:32 |
And we'll make sure that we are in the right spot here, our first value there is 1300.
|
52:39 |
So, let's just use our CTRL and arrow keys and we'll align our data until everything's looking about right.
|
52:47 |
And again we know that this table is going to be 16x16.
|
52:51 |
So, we'll just hold down the shift key, come across to the left there, obviously our width is set to 16 and we'll also come up 16.
|
52:57 |
So, one, two, three, four, five, six, seven, eight, nine, ten, eleven, twelve, thirteen, fourteen, fifteen, sixteen.
|
53:04 |
And we'll press K to mark that table.
|
53:07 |
Now, we're going to just CTRL V and that will copy our axis data and our scaling and description job is done there.
|
53:15 |
Looking at these values here, these also are predominantly filled with values of 0.3 so very similar to the last table, we do have some richer targets down here, the lower airflow and lower RPM regions.
|
53:28 |
So, we'll press escape and close that down, we've found our second table.
|
53:31 |
Let's just go back to our 2D view here and we'll move across to the left.
|
53:36 |
We are sort of doing this in reverse order, but as you can see there with the axis at the end, the fact that we need to copy that data and the fact that those last two tables, which I'm pretty confident in saying are the ones that are relied on the most, actually have a bit of shape to them, these other remaining tables are a little bit tricky to find.
|
53:55 |
So, what we'll do is we'll just put our cursor at the far side of the data so we should be there or thereabouts.
|
54:01 |
We'll click on our text view and we can see we've got our cursor sitting here.
|
54:04 |
So, we need to just take a little bit of a look at these numbers and see what makes sense, the actual point where I put the cursor, we've got values of 2 and 2.
|
54:14 |
Now, just to the left of that we've got a value of 1020 and we already know that if we multiply that by our lambda factor or scalar, that would be lambda 1.02.
|
54:24 |
So, that makes sense, the two values of 2 do not so what I'm going to do is just use the Ctrl and arrow key to get those out of the way.
|
54:32 |
And we're going to start here in the bottom left, we're going to then simply shift and across to highlight out 16 and then shift and up to highlight 16 as well.
|
54:45 |
And ok we'll mark that table, looking at the numbers, everything makes sense there, Ctrl V will paste our axis and scaling data in, escape and our job there is done.
|
54:55 |
And we can see that when I do this, OLS sort of tries to guess ahead, it's moving down the table or the data range instead of up, we're going in the opposite direction so delete will unmark that next section.
|
55:08 |
So, what we're going to do now is just jump ahead and I'm going to mark the remaining four of these tables, it's just exactly the same process that we've looked at so far.
|
55:26 |
I said we had seven tables here, but if we actually look, it looks like we've got another table up here as well so we'll mark that one and that makes for eight tables for our smoke limitation by lambda.
|
55:41 |
Alright, all of our tables are marked now and of course as usual what we want to do is just tidy things up a little bit by adding all of these tables into a folder.
|
55:53 |
So, let's go ahead and do that now.
|
55:56 |
What we're going to do is right click and come down to copy name, we will right click again and click on new folder and we'll name that folder, I've copied the name from our first table, which we'll remember was smoke limitation by map, so I'm just going to remove the by map and we'll put all of these tables into the same folder.
|
56:17 |
So, Ctrl A, Ctrl C and then we'll come down here and this will automatically apply this rule to all of our smoke limitation tables and our job is done there, if we open our smoke limitation we can see we've got all of those.
|
56:31 |
Smoke limitation is now found and defined, let's move on with our next set of tables.
|
56:36 |
We're now going to find and define our start of injection tables, which define when the injection event is going to start.
|
56:44 |
One of the aspects with increasing the power on diesel vehicles is generally this is done by increasing the fuel delivery and when we do this, this does require the injection timing to be adjusted as well.
|
56:57 |
If we jump into our laptop screen here, this is the general shape of the table we're going to be looking for and we've got a range of these tables, there's quite a number of them that we're going to have to find.
|
57:08 |
So, we'll see this pattern essentially repeat.
|
57:11 |
So, let's jump back into OLS and we'll go ahead and find that.
|
57:16 |
Now, again I'm cheating a little bit, I know that the tables that we're looking for here are going to be to the left of our cursor, so we're going to jump ahead that way until we find them.
|
57:30 |
Alright, so we're seeing these tables here and if we just check back, yeah this generally shows the same sort of pattern.
|
57:43 |
Head back to OLS and as I mentioned we can see there are quite a number of these tables.
|
57:47 |
They just keep repeating, so it's not just the pattern that we want to be taking note of, it's also the repetitive nature of the pattern, that's the two things that really help us find this essentially what is a needle in a haystack.
|
58:01 |
We can also see that OLS has marked all of these tables, which is somewhat helpful.
|
58:06 |
So, let's just start at the left hand side here with the first of these and we'll double click on it.
|
58:11 |
Of course, in order to define this table we need to know what the axes are going to be and in this case our start of injection table is going to have RPM on our y axis and the x axis is going to be injection quantity in milligrams per stroke or milligrams per cycle.
|
58:26 |
The values in it are degrees and the scaling factor for this one is quite unique.
|
58:31 |
This really does require a description file, you are absolutely not going to be able to just stumble upon this and guess it based on logical values, but let's get into this and start with the simplest , which is our engine speed axis.
|
58:45 |
So, let's add a description here.
|
58:48 |
And as usual our factor for the RPM axis is 0.5, giving us 1000 to 4500 RPM all makes sense.
|
58:56 |
We'll just click here to the x axis and this one, as I mentioned, is injection quantity.
|
59:01 |
And you'll quite often see this as IQ as well, it's actually just a shortened form.
|
59:08 |
And we are looking here for milligrams per cycle.
|
59:13 |
Ok, the factor for this, and again this really comes down to understanding the sort of values or magnitude of the values we're likely to see.
|
59:21 |
At the moment we've got 6500 is the largest value and the actual scaling factor for this is 0.01, giving us a maximum value of 65 milligrams per cycle and again what we can do is add a little bit of precision to this.
|
59:35 |
Ok so, that leaves us with the map data so we'll double click on that and this map is start of injection.
|
59:44 |
Again Ctrl A, Ctrl C, tab, Ctrl V and this one, the units are degrees.
|
59:51 |
Now, the scaling factor as I mentioned here, it's quite a long one, so I'm just going to enter that, it is 0.0219736525.
|
01:00:05 |
So, again no surprises, you weren't going to exactly guess that.
|
01:00:08 |
And we'll give this a couple of degrees of precision as well.
|
01:00:12 |
So, there we've got our degrees before top dead centre for our injection event to start and again assuming that you have had some involvement with diesel engines, these numbers will start to look reasonably sensible.
|
01:00:25 |
Alright, so now we've got our first map defined, what we're going to do is of course copy our map properties, this time we want to make sure that we've unticked our data source box here, because each of these tables have their own axes.
|
01:00:37 |
We'll click OK, escape to close that.
|
01:00:40 |
And now I'm just going to jump ahead because this is just a case of rinse and repeat and we do have a number of these tables so let's just speed this up.
|
01:00:59 |
Alright, so we've defined at this stage 15 of these maps and on face value if we scroll out to the right, we don't seem like we've got any more of that repeating pattern that we can see.
|
01:01:11 |
So, on face value this looks like the remainder or the balance of our start of injection maps, we actually do have a few more to go but these are removed a little bit, we've got exactly the same pattern that we're looking for, but they are a little bit further out to the right so we'll just scroll through until we get to those now.
|
01:01:30 |
Alright, so we can see we've got this map here, which OLS has already identified, this one is 11x13, worth noting here that these tables do change in size, but it's got exactly that same pattern.
|
01:01:49 |
We've got these other tables out here as well, this one and this one and obviously this repeats, they on face value look quite different, but we can change the way we're viewing OLS and if we view this as signed values, now we start to see that everything looks a little bit more familiar.
|
01:02:06 |
So, let's start again with the first of those maps, we'll double click on that, Ctrl V, we'll paste in our scaling and information there.
|
01:02:14 |
Looking at these numbers, yep that all looks logical, so we'll press escape.
|
01:02:18 |
Now, we'll click on the next one, so this is the first of those that had a fairly unique shape.
|
01:02:24 |
If we apply, well first of all look at our numbers, obviously these are signed values and if we just simply Ctrl V and apply our existing scaling factors, we can see that it obviously takes care of our X and Y axis but the map data is obviously incorrect, we've just got some random numbers that don't make a huge amount of sense.
|
01:02:43 |
So, if we double click in here, the reason for this is these are now signed numbers, we know that we've got positive and negative values so what we need to do is come down here and click on this little sign button, click ok and now everything's looking much better.
|
01:02:57 |
The remainder of these tables are signed, so what we'll do is right click and we'll now grab our new map property so that'll just take into account that signed value.
|
01:03:07 |
We'll press escape and now we can carry on, we'll speed up again and we will define the remaining tables.
|
01:03:15 |
And that looks like we've found them all, we'll just scroll up again and we've got 24 and I know that's what we should be looking for.
|
01:03:27 |
Of course, this does beg the question of with 24 maps, which ones should we be adjusting, but that's a tuning aspect, not necessarily a map definition aspect.
|
01:03:37 |
As usual what we're going to do is just clean things up a little bit so we'll right click and copy name, right click again and create a new folder and we'll just move all of those into our folder.
|
01:03:51 |
Alright, our job's done there, onto our next map.
|
01:03:53 |
One of the tuning changes we're almost inevitably going to make when tuning turbo diesel engines is to increase the boost pressure in order to provide more airflow so that we can then add more fuel.
|
01:04:04 |
To do this we're going to need to raise our requested charge pressure targets, as well as potentially change our charge pressure limits.
|
01:04:13 |
Let's start by looking for our charge pressure target or requested charge pressure tables and looking at my laptop screen, this is the general shape of these tables.
|
01:04:21 |
Relatively easy to find and we're generally going to have four of these tables that we're looking for and these are 10x16.
|
01:04:30 |
So, let's jump into OLS and we'll go ahead and find these.
|
01:04:34 |
Seeing as we've just come off the back of finding our signed start of injection tables, we want to go ahead and put it back to our non signed view and we're going to move to the right here until we find our charge pressure tables.
|
01:04:54 |
Alright, we've found the first of these tables and we can just have another look at our image.
|
01:04:59 |
Worth noting here that the X and Y magnification is set different to OLS, so it does look a little bit different, but yeah, generally, we've got that shape so we know, we're confident we've got that table and of course we can see OLS has actually found this table.
|
01:05:12 |
Let's double click on it and see what we can learn.
|
01:05:15 |
So, this is going to be charge pressure or requested charge pressure.
|
01:05:19 |
And we already know from previous axes that use manifold absolute pressure that the EDC controller works in absolute pressure in millibars and we're seeing numbers around the 2600 mark so 2600 millibars as an absolute boost pressure target, yep that makes sense.
|
01:05:38 |
The axes here are going to be RPM on our Y axis and then we're going to have injection quantity on our X axis.
|
01:05:46 |
We already know the scaling factors for essentially all of these, so let's go ahead and start with our Y axis.
|
01:05:53 |
Of course, engine speed.
|
01:05:56 |
And a factor of 0.5 takes us out to 5000 RPM.
|
01:06:00 |
We'll click here on the Y axis and as I mentioned, injection quantity.
|
01:06:05 |
And our units there, again milligrams per cycle.
|
01:06:10 |
And our multiplication factor for this or scaling factor is 0.01, giving us a maximum value of 70 milligrams per cycle and again we'll give a little bit of precision there with one decimal place.
|
01:06:23 |
So, let's just click across to our map now and this is requested charge pressure.
|
01:06:31 |
And units there are millibar and there is no scaling factor for this, it is just a 1 to 1.
|
01:06:38 |
So, we'll click OK and our table is defined.
|
01:06:41 |
What we'll do is right click and copy our map properties, we're copying our description and factors for the X, Y and Z axis, we can click ok, press escape and we'll close our first table down.
|
01:06:53 |
Let's scroll across to the right and find our remaining tables.
|
01:06:58 |
Alright, we've got another one here, again OLS has found that so we'll double click on that, Ctrl V, our scaling is applied, we'll press escape, close that down.
|
01:07:07 |
And let's see if we've got any more out here to the right.
|
01:07:12 |
Another one here, again double click on it, Ctrl V to apply our scaling and we can close that down.
|
01:07:17 |
We've come across to the right here, we've got another map that on face value looks similar, but notably a little bit different, we also have a different map size.
|
01:07:27 |
This one's a 10x14.
|
01:07:29 |
So, this is one of our requested charge pressure tables and what that means is at the moment we've only defined three of these tables and I know there are four.
|
01:07:37 |
So, probably what this means is that I've simply missed the first of these.
|
01:07:41 |
So, let's just scroll back across, so here's the three tables that we have already defined.
|
01:07:47 |
Let's just left click there.
|
01:07:49 |
This is the first of them that we defined and let's just scroll to the left of that and I'm picking that I missed the left most one so we'll keep coming and here we go.
|
01:07:58 |
This is the fourth, well actually the first of these tables, again double click, Ctrl V to apply our scaling, escape to close that down and we've got our requested charge pressure tables found.
|
01:08:12 |
Alright, so what we're going to do again is just add these into a folder.
|
01:08:18 |
We'll copy our name here and we will create a new folder.
|
01:08:23 |
What I'm going to do is I'll paste that in but we'll just call this charge pressure for the moment, because we are going to put our charge pressure limitation tables in here as well.
|
01:08:33 |
We will put our requested charge pressure, well actually I'll just change this to charge pressure as well and that will put all of those into our charge pressure folder.
|
01:08:44 |
Alright, let's move on and we'll find our charge pressure limitation tables and we're going to find there's two different ways that the charge pressure limitation is achieved.
|
01:08:53 |
One is via barometric air pressure and the other is via intake air temperature.
|
01:08:59 |
Technically, you wouldn't necessarily need to find both of these but for the sake of completeness, let's go ahead and do that.
|
01:09:06 |
So, this is the shape for our charge pressure limitation via baro, we can see that shape there.
|
01:09:12 |
And it looks very similar to the table to the right of our charge pressure, requested charge pressure tables.
|
01:09:19 |
So, let's go back into OLS and we'll go ahead and find that.
|
01:09:22 |
So, we were out to the right here, let's continue out that way.
|
01:09:30 |
Alright, so that's our table there.
|
01:09:33 |
We just head back and have a look at our image.
|
01:09:35 |
Yep, despite the fact that our magnification scale is a little bit different, that is the table, it is a 10x14 16 bit.
|
01:09:43 |
And if we head back to OLS we can see that OLS has already found this and again 10x14 and 16 bit so let's double click on this.
|
01:09:52 |
And the numbers in this should be pretty obvious by now.
|
01:09:56 |
Of course, on our vertical axis we've got our engine speed.
|
01:10:03 |
And on our X axis, so this is charge limitation via baro, so of course barometric air pressure is going to be the other axis.
|
01:10:12 |
If we're looking at these numbers, 450 to 1000, these are going to be raw values that just have no scaling applied and these are barometric air pressure in millibar.
|
01:10:25 |
Alright, so now we've just got our Z axis or map data and of course we can pretty much figure this one out.
|
01:10:32 |
So, this is charge pressure limitation via baro.
|
01:10:36 |
Control A, Control C, we'll tab down, Control V and the values here of course are our boost pressure targets in millibars absolute and again with no scaling applied.
|
01:10:46 |
So, that's the first of our charge pressure limitation tables found.
|
01:10:50 |
Let's have a look now at the shape of our charge pressure limitation via intake air temperature.
|
01:10:56 |
The last of our charge pressure tables is not actually a limitation table so much as a correction table, so this corrects the requested charge pressure based on the current intake air temperature and looking at the laptop screen at the moment, this is the shape that we are looking for.
|
01:11:11 |
Let's head back into OLS and we'll scroll through until we can find it.
|
01:11:21 |
Alright, that looks like we've got it right there.
|
01:11:25 |
And OLS has also found this map, let's have another quick look over at our little JPEG and make sure that it does look the same and obviously with the scaling being slightly different, there are some subtle differences but that is the table we're after.
|
01:11:39 |
Alright, so looking at this table we need to understand what we're expecting here.
|
01:11:43 |
So, it's a charge pressure correction versus intake air temperature.
|
01:11:48 |
Obviously, one of these axes is going to be intake air temperature.
|
01:11:52 |
Not immediately obvious but we can also have a look at our y axis, the number 2700 should already stick out there as a requested charge pressure in millibar, so this axis here, our x axis is going to be our inlet air temperature.
|
01:12:08 |
So, let's just enter that.
|
01:12:13 |
Now, if we actually multiply this by 0.1, what we're going to do is find that straight away these numbers all look sensible as a degrees C.
|
01:12:24 |
So, it's a bit of a case here of understanding what that scaling's going to be in this case.
|
01:12:28 |
You're not necessarily going to be able to guess this if you understand what you're looking at, we know that it's air temperature.
|
01:12:34 |
So, this is something that generally if you don't know already, you're going to need to find from a description file, but we're making that process a little bit easier.
|
01:12:43 |
We'll also change our units there to degrees C.
|
01:12:46 |
Let's head over to our y axis and that's going to be our charge pressure.
|
01:12:53 |
And of course in millibars and we don't need to apply any scaling.
|
01:12:57 |
And last of all our actual map will give us a map description, which is going to be charge pressure correction via intake air temperature.
|
01:13:10 |
And our job's done.
|
01:13:12 |
So, at this stage we can close that down by pressing escape, we've got all of the tables that we're going to need there to properly manipulate the charge pressure.
|
01:13:19 |
I will mention that if you're going a little bit further than a stage one tune, there may be benefits also in having access to the base VNT tables, but for our stage one tune, that's not going to be essential so we'll keep this nice and simple and we won't reference those tables.
|
01:13:35 |
As usual we'll just have a little bit of a clean up here and put those tables into our charge pressure control.
|
01:13:41 |
Now, we're going to move on and look at some of the controls for our common rail diesel fuel pressure and raising the fuel pressure is quite a common modification when we are retuning these vehicles.
|
01:13:53 |
So, let's jump into our laptop software, again this is the general shape of the tables that we're looking for so pretty obvious and we're going to be looking for three of those or three of them, one beside the other just like that.
|
01:14:06 |
So, let's head into OLS and we'll get those found or located.
|
01:14:11 |
Now, in this case I'm cheating a little bit, I know that these tables are going to be to the left of where I am so we'll head over until we find them.
|
01:14:21 |
And that looks like we've got them right there so let's just scroll over, find the first of these tables and again we can see that OLS has already found these tables as a matter of course, let's double click on the first of these.
|
01:14:35 |
So, what are we looking for here? Well, this is our target fuel pressure and again we need to have some understanding of what sort of pressures we're likely to see.
|
01:14:45 |
Generally, we're talking in bar and it wouldn't be uncommon to see values around 2000 bar.
|
01:14:51 |
So, we can see in the bottom right hand corner we've got values of 20,000 so probably, pretty likely that this is our table and the scaling factor on that's probably going to be 0.01, sorry 0.1, let's do our maths properly.
|
01:15:06 |
We're going to have on our y axis here, engine speed and our other axis, our x axis is going to be our injection quantity so let's go ahead and add some scaling to this.
|
01:15:18 |
Engine speed as usual, we already know we're going to be using a factor of 0.5, so let's apply that.
|
01:15:27 |
Yep, 4800, that all looks pretty good to me.
|
01:15:30 |
Our x axis, injection quantity, this will be as usual in milligrams per cycle and we already know that our scaling for this is 0.01 and we're going to add a little bit of precision to that as well and let's click on our map.
|
01:15:50 |
So, this is going to be requested rail pressure.
|
01:15:56 |
And this is in bar.
|
01:16:00 |
And our scaling factor as already discussed, 0.1, yep that gives us values that make sense for common rail fuel pressure.
|
01:16:08 |
So, we're going to have three of these tables as discussed so we can right click, copy our map properties, click ok, escape to close that down and we'll repeat this process for our remaining two maps.
|
01:16:22 |
Alright, our three requested rail pressure maps have been found and defined and again we'll just quickly tidy these up by adding them into their own folder.
|
01:16:35 |
Now, along with our requested rail pressure, we're also going to have a rail pressure limitation or a couple of rail pressure limitation tables.
|
01:16:42 |
Looking at our screenshot here, this is the shape we're looking for, quite an obvious shape with this sort of stepped mountainous transition.
|
01:16:50 |
So, let's head back into OLS and we'll locate that particular table or those two tables.
|
01:17:00 |
Ok, and straight away we can see those tables pop up just a little bit to the right of our requested rail pressure and it makes sense that these would be located pretty close to each other.
|
01:17:11 |
So, we've got those two tables, again OLS has already found them so let's double click on those and define them.
|
01:17:16 |
So, again what are we expecting here? Well we're probably going to expect RPM versus injection quantity and of course the numbers in the table are our rail pressure limit and we can see again we've got the same 20,000 values that we saw before so safe to assume here that our same factor of 0.1 is going to sort that out.
|
01:17:36 |
So, let's start with our y axis and we will define that.
|
01:17:43 |
Now, we'll move across to our x axis and this again is injection quantity.
|
01:17:52 |
And the map itself as discussed is our rail pressure limitation.
|
01:18:00 |
Alright, we've got one more of these tables so again we'll right click on the table, copy our map properties, click ok, escape to close that down and we can double click on the table to the right, Ctrl V to apply our scaling and our job's done there.
|
01:18:15 |
What we can do, just as a matter of completeness, is add a little bit of precision to the injection quantity axis there, we can come back and do the same for this one.
|
01:18:25 |
We can see that the scaling or range of the two tables is quite different.
|
01:18:31 |
This one running up to 88 mg per cycle whereas if we go back into our second table, we can see that second table only ranges up to 60 mg per cycle.
|
01:18:41 |
So, on the home stretch here we've essentially defined most of the tables that we are going to need to find for our stage one tune.
|
01:18:49 |
We've got one more that I'm going to add into the mix here, which is our main injector on time table.
|
01:18:54 |
So, let's jump into our laptop software and see what that table looks like.
|
01:18:58 |
Alright, so we can see the shape of that table there, sort of typical sawtooth pattern that is diminishing as we move from the left to the right.
|
01:19:07 |
And it's also worth noting here, this is actually quite a large table, it's 22x16.
|
01:19:13 |
So, let's jump into OLS and we'll locate that.
|
01:19:20 |
Alright, so we've found that table there, pretty obvious, pretty easy one to find.
|
01:19:25 |
Let's double click on that and add some scaling.
|
01:19:28 |
So, what are the axes for this? Well, we're going to have injection quantity versus rail pressure.
|
01:19:36 |
And the numbers in the actual map axis there are our injection on time in microseconds.
|
01:19:43 |
Looking at the numbers, hopefully it should be pretty apparent.
|
01:19:47 |
We're ranging out to 20,000 here, this should stick out as a rail pressure value.
|
01:19:52 |
Out to 9500 here, this is going to be injection quantity.
|
01:19:58 |
So, let's go ahead and add some scaling.
|
01:20:05 |
And for our Y axis we're going to call this injection quantity.
|
01:20:13 |
And this is in milligrams per cycle as usual and of course we already know that our scalar there is 0.01 and we can add a little bit of precision.
|
01:20:23 |
And then our map data is our main injection on time.
|
01:20:31 |
And this is in microseconds as I mentioned here.
|
01:20:34 |
So, this is actually a one to one table as well.
|
01:20:38 |
And again you might not necessarily recognise these values or the magnitude of these values until you've been tuning diesels but that's our table found and our job essentially there is complete, we'll press escape to close that down.
|
01:20:52 |
So, that's essentially all of our tables found and defined for our stage one calibration.
|
01:20:57 |
Only remaining task left to do here is just to add some more folders for those last tables we've just found so we'll go ahead and do that now.
|
01:21:12 |
And our job's complete, we've got all of our tables defined, we can now move on with the next step of our process.
|