blu3dragon

iTrader: (1)

The stock throttle map has been bugging me for a while. It feels progressive at first, but then quite aggressive, and the 2nd half of the pedal travel does not do anything, particularly at mid rpms (3500-5500). Modifying this was actually one of my justifications for buying ME in the first place, but 2 years then went by with me not really being able to figure out the throttle maps...

In that time I had a couple of spins on track in low grip scenarios out of slow corners:
Turn 2 at Sonoma in the dry on factory tires, and then a couple of times in the wet in different places. Particularly in the wet, it feels like there is a point where 1mm of pedal travel simply has too big an effect on the power output.

Finally I decided to record some data and came up with these graphs:

Note: Throttle pedal position % is along the bottom (x-axis), g/s % is along the side (y-axis). g/s% is the measured g/s as a % of the max achievable g/s for the rpm at the time.









Aha!.. proof that I am not crazy. Or at least some kind of validation. Looking at the relationship between measured g/s and pedal position, g/s ramps from close to minimum at 10% pedal position to close to max by 40% pedal position, and is basically at the max at 50% pedal for all the rpm ranges I measured.

After staring at data (Accelerator pedal position, throttle and absolute throttle position) and the available maps in MazdaEdit for some time I finally decided to modify the "Desired Engine Torque" table.

My theory being that the "throttle duty" tables are used to calibrate the throttle (they alter the amount of throttle requested for a given pedal position), while the "Desired Engine Torque" table modifies the amount of torque (engine output) the ecu tries to achieve for a given throttle.

Here is the stock "Desired Engine Torque" table. You can see that it seems to ramp pretty close to max by 39 throttle (which is a little under 50% of the 84 max).



So, I made the "Desired Engine Torque" table linear, with the hope of getting a linear engine torque output to throttle pedal position.



I have yet to re-generate my graphs, but this seems to have had the desired effect.

I then altered my throttle duty tables a little bit to make the pedal progressively more responsive with the aim of improving control for small amounts of throttle. Some of this is already build in to the stock map. I just smoothed it out a bit, to end up with this for all the variants (A, B, C, etc):

logalinipoo

There is no need for you to change the 250 and 500 RPM range in the throttle tables. You risk not having the full throttle cut off(for deflooding) And having that smoothed It might give more chance for flooding when starting.

I also like to modify mine a little. at 100% throttle from 4000 rpm to max I set it to 0.10 So if someone gets in my car and floors it it falls on it's face. Then I set the 750 rpm to 3500rpm at 50% throttle.

I change the 90% to 98% and the same as the stock 100% Then averaged. So you lift off the floor just a little and you have full throttle.

When I let the nebighor boy drive it he came back pissed saying it runs like ****. I took him for a spin and he couldn't figure it out.

blu3dragon

iTrader: (1)

I had not thought of that, although I did not modify any of the values for 0% throttle position, so it should not make any difference. I'd actually expect the ecu to ignore the throttle position during starting as well and just do it's own thing, but maybe not.

There are a couple of slightly odd entries in the factory "desired engine torque map"; The negative values which I'm guessing represent engine braking, and at 3500 and 4000 rpm the 3% throttle position column has a lower value than the 0% which is a bit odd. I used the 0% throttle values as my minimum.
There's also a ridge at 6500 rpm - perhaps a feel thing, or it could be meant to compensate for some hole in the engine output. In either case, it does not seem significant enough for me to worry about. Nor did the subtle differences between the multiple Throttle Duty maps, which is why I simply set them all the same.

logalinipoo

The different throttle maps are different gears and neutral. So I can't imagine it will make a huge differance. I don't have the requested torque with accessport.

blu3dragon

iTrader: (1)

Since I had a question via pm, here is my edited "Desired Engine Torque" table in text format:

<table>
<tr><td></td><td>0</td><td>3</td><td>5</td><td>7</td><td>9</td><td>12</td><td>15</td><td>18</td><td>21</td><td>30</td><td>39</td><td>48</td><td>57</td><td>84</td></tr>
<tr><td>500</td><td>0.62</td><td>3.21</td><td>4.94</td><td>6.67</td><td>8.39</td><td>10.99</td><td>13.57</td><td>16.17</td><td>18.75</td><td>26.53</td><td>34.30</td><td>42.07</td><td>49.84</td><td>71.92</td></tr>
<tr><td>750</td><td>-16.19</td><td>-11.76</td><td>-8.80</td><td>-5.85</td><td>-2.89</td><td>1.54</td><td>5.97</td><td>10.41</td><td>14.84</td><td>28.14</td><td>41.43</td><td>54.73</td><td>68.03</td><td>107.92</td></tr>
<tr><td>1000</td><td>-19.28</td><td>-14.28</td><td>-10.95</td><td>-7.62</td><td>-4.29</td><td>0.70</td><td>5.70</td><td>10.70</td><td>15.70</td><td>30.69</td><td>45.68</td><td>60.67</td><td>75.66</td><td>120.63</td></tr>
<tr><td>1500</td><td>-21.38</td><td>-15.45</td><td>-11.50</td><td>-7.55</td><td>-3.59</td><td>2.34</td><td>8.27</td><td>14.20</td><td>20.12</td><td>37.91</td><td>55.70</td><td>73.48</td><td>91.27</td><td>144.63</td></tr>
<tr><td>2000</td><td>-23.38</td><td>-16.98</td><td>-12.73</td><td>-8.46</td><td>-4.20</td><td>2.20</td><td>8.59</td><td>14.98</td><td>21.37</td><td>40.55</td><td>59.73</td><td>78.91</td><td>98.09</td><td>155.63</td></tr>
<tr><td>2500</td><td>-25.38</td><td>-18.38</td><td>-13.70</td><td>-9.04</td><td>-4.37</td><td>2.63</td><td>9.64</td><td>16.64</td><td>23.65</td><td>44.66</td><td>65.67</td><td>86.69</td><td>107.70</td><td>170.73</td></tr>
<tr><td>3000</td><td>-27.59</td><td>-20.18</td><td>-15.24</td><td>-10.30</td><td>-5.37</td><td>2.04</td><td>9.45</td><td>16.85</td><td>24.26</td><td>46.49</td><td>68.71</td><td>90.93</td><td>113.15</td><td>179.82</td></tr>
<tr><td>3500</td><td>-30.38</td><td>-22.70</td><td>-17.58</td><td>-12.46</td><td>-7.34</td><td>0.34</td><td>8.02</td><td>15.70</td><td>23.37</td><td>46.41</td><td>69.44</td><td>92.48</td><td>115.52</td><td>184.63</td></tr>
<tr><td>4000</td><td>-34.09</td><td>-26.24</td><td>-20.99</td><td>-15.76</td><td>-10.52</td><td>-2.66</td><td>5.20</td><td>13.06</td><td>20.92</td><td>44.49</td><td>68.06</td><td>91.64</td><td>115.21</td><td>185.93</td></tr>
<tr><td>4500</td><td>-38.49</td><td>-30.22</td><td>-24.71</td><td>-19.20</td><td>-13.69</td><td>-5.41</td><td>2.85</td><td>11.12</td><td>19.39</td><td>44.20</td><td>69.01</td><td>93.80</td><td>118.61</td><td>193.03</td></tr>
<tr><td>5000</td><td>-42.28</td><td>-33.99</td><td>-28.45</td><td>-22.92</td><td>-17.39</td><td>-9.09</td><td>-0.80</td><td>7.50</td><td>15.80</td><td>40.69</td><td>65.58</td><td>90.47</td><td>115.36</td><td>190.03</td></tr>
<tr><td>5500</td><td>-47.49</td><td>-38.76</td><td>-32.94</td><td>-27.11</td><td>-21.29</td><td>-12.55</td><td>-3.83</td><td>4.91</td><td>13.64</td><td>39.84</td><td>66.04</td><td>92.23</td><td>118.44</td><td>197.03</td></tr>
<tr><td>6000</td><td>-53.78</td><td>-44.93</td><td>-39.03</td><td>-33.13</td><td>-27.24</td><td>-18.39</td><td>-9.55</td><td>-0.70</td><td>8.15</td><td>34.69</td><td>61.23</td><td>87.77</td><td>114.31</td><td>193.93</td></tr>
<tr><td>6500</td><td>-56.09</td><td>-47.42</td><td>-41.63</td><td>-35.85</td><td>-30.07</td><td>-21.40</td><td>-12.73</td><td>-4.05</td><td>4.62</td><td>30.63</td><td>56.65</td><td>82.66</td><td>108.68</td><td>186.73</td></tr>
<tr><td>7000</td><td>-62.09</td><td>-53.30</td><td>-47.43</td><td>-41.57</td><td>-35.71</td><td>-26.92</td><td>-18.13</td><td>-9.33</td><td>-0.54</td><td>25.85</td><td>52.22</td><td>78.61</td><td>104.98</td><td>184.13</td></tr>
<tr><td>7500</td><td>-65.28</td><td>-56.79</td><td>-51.13</td><td>-45.47</td><td>-39.81</td><td>-31.32</td><td>-22.83</td><td>-14.34</td><td>-5.85</td><td>19.62</td><td>45.08</td><td>70.55</td><td>96.02</td><td>172.43</td></tr>
<tr><td>8000</td><td>-69.19</td><td>-60.80</td><td>-55.21</td><td>-49.62</td><td>-44.03</td><td>-35.64</td><td>-27.26</td><td>-18.87</td><td>-10.48</td><td>14.67</td><td>39.83</td><td>64.99</td><td>90.15</td><td>165.63</td></tr>
<tr><td>8500</td><td>-71.59</td><td>-63.51</td><td>-58.13</td><td>-52.74</td><td>-47.36</td><td>-39.29</td><td>-31.21</td><td>-23.13</td><td>-15.06</td><td>9.17</td><td>33.39</td><td>57.62</td><td>81.85</td><td>154.53</td></tr>
<tr><td>9000</td><td>-71.99</td><td>-63.90</td><td>-58.50</td><td>-53.11</td><td>-47.72</td><td>-39.63</td><td>-31.54</td><td>-23.45</td><td>-15.36</td><td>8.91</td><td>33.18</td><td>57.45</td><td>81.72</td><td>154.53</td></tr>
</table>

NotAPreppie

The former genchem TA in me is screaming about taking off points for not labeling your axes directly on your plots...

Is it possible that your data points just don't extend far enough up the RPM range? I mean, you've only gotten to about half way to redline.

blu3dragon

iTrader: (1)

Sorry, that was just me being quick and lazy in excel

My data did enough to convince me of my theory without looking at the higher rpm range, which is harder to get data for on the street (and I was also breaking my engine in at the time). Just by feel it is a little more progressive from the factory at higher rpms and you can see that on the Desired Engine Torque table. However, it is still not linear (which you can also see on the factory table).

Having now driven the car on track for 5 days in total, and daily on the street since making this change, I can say I am very happy with the throttle map now. I have not driven it hard in slippery conditions yet, but I have hung the tail out quite a bit in the dry and it feels very controllable on the throttle.

This is all presented here as is, so you can draw your own conclusions, or use as a basis to experiment further :-)

paimon.soror

iTrader: (15)

hrmm... subbed for interest

paimon.soror

iTrader: (15)

So im not sure if i missed something ...

was there any mathematical algorithm that you used to taper your graph to be linear, or did you straight eyeball it all

jasonrxeight

I probably won't like the edited table because I would find it very slow and sluggish to drive.
The factory pedal chart is clearly designed for quick throttle response.
Also spinning out in slower corners? You have other issues rather than throttle response. I've floored in 1st gear in slow corners the back end would just stick to the ground like glue.
You can clearly see the factory chart is compensated for the torque curve because your pedal is usually how much torque you want to get and you are expecting the same amount of torque throughout the rpm at a set gas pedal position.
With your edited curve, you will experience weird things like the car would accelerate faster and harder as the rpm climbs when holding a steady throttle pedal because its not torque curve compensated. In other words, you will be backing off the pedal as the rpm climbs in order to get a more linear torque curve.

logalinipoo

I run a liner throttle map and it changed the response to Rev faster. The first part of the map has very little throttle opening per pedal movement. The biggest thing that happened is no more bucking at slow speeds.

blu3dragon

iTrader: (1)

I used excel to calculate the numbers for a straight line increase in tq from a throttle position of 0 up to 84 (max). The 3D plot is visually deceptive since the throttle position axis is not linear. Here is a better graph:



EDIT: if you mean the Throttle Duty graph, that one I eyeballed based on the stock map, and iterated a couple of times based on driving feel. It is also linear once you go past 10% pedal position.

blu3dragon

iTrader: (1)

It might feel like the car is slower at first. As soon as you get used to having to press the pedal a little more, then you will realize the car is just as quick, and responds just as quickly. However, if you don't need the extra control, and prefer the car to dart forward with just a little pedal movement, that is fine. Many people do, and lots of tunes (for other cars) actually go in this direction from stock. There are also many "throttle response boost" products on the market that do just this.

My car will stick fine when I am cornering below the limit of grip. I did this for control when cornering at the limit on track. I posted some recent video (using this map) here: https://www.rx8club.com/rx-8-racing-...3/#post4715919 A year ago I would have spun the car at least 180 degrees when the rear came out as much as it did in two of those videos. My driving has improved since then, but I think an easier to control throttle is part of that. Of course, now I am driving faster since I don't spin so much, and so I still make these mistakes

I went for 50% throttle = 50% of the available tq at that rpm. 100% is 100% of available tq at that rpm. I figured this would allow the best control in all situations. As for experiencing weird things, I actually found the opposite. With the factory map I would have to consciously move the pedal if I wanted to maintain constant acceleration at certain points. My map reduces this need to the point where it is not really noticeable (but does not eliminate it if you really are trying to keep constant acceleration).

Here is what the factory map looks like on a 2d plot (4k rpm line is in red). You are giving up a lot of control for that "fast" response.


I'm not sure exactly what you did, but it sounds like the opposite of what I did here.

logalinipoo

Yeah it kind of does I like how you did it, but don't have Mazda edit. I've also reduced my max throttle opening since lm fi. Because once you hit boost you loose a lot of the top of pedal play so I reduced it to 60 max at 100% accelerator position.

miro279

Interessing thread, does anyone know how the TBW actually works? What I mean: there is a steady state of variables and then the divers makes an input with the pedal.... how does the ECU calculate the change at the throttle body?

From the maps my guess would be there is a pedal position that corresponds more or less linear to the throttle (4 maps throttle duty where 0% pedal is 0,1 throttle and 100% pedal is 84 throttle.
The ECU than looks in map called "desired engine torque" and translates the throttle value to a desired torque.

There is als a map called "relative throttle plate position" having "rpm" and "calculated torque" as input variables.

If anyone can point me to the whole picture, I'd appreciate.

ZenVirZan

You might be interested in my findings in my Versatuner thread here: Zen's Cable Throttle tune

This tune is for S1 but the same concept should apply to S2

TeamRX8

iTrader: (25)

I modified mine long ago and can’t remember much other than making it more linear and never having any low rpm bucking issues ever again

just because it doesn’t do much wrt airflow at certain opening % doesn’t mean you want the blade to make big opening jumps. It may not seem like much, but when you’re on the traction limit those fine changes actual provide a lot of control even though intuitively it may not appear that way.

I suspect it’s one of the reasons that people who switched to a cable throttle long ago felt like they had a lot more control on the track.

.

Meat Head

I’m just adapting your values to suit my 2010. Will let you know how it goes. Strangely the number of fields I have available to edit are less than what you have shown on the VersaTune website. Maybe another difference between the earlier and later cars?

ciprianrx8

Here is what I found that suits me:

(all 4 similar tables changed, no other changes made in DBW tables which I suspect actually dont do anything - used for traction control perhaps ?)
What I wanted to achieve is remove the dead last half of accelerator pedal travel. In stock form I've found that pressing the gas 50% has the same result as flooring it. I wanted more control.

Will feel damn slow the first time you try it... but then you see there is a point in using the entire pedal travel. There is a noticeable difference between going 50%, 75% and 90% on the gas now - though the last 10% hardly make a difference. Didn't want to leave any of those already lacking HPs...

Meat Head

I noticed yesterday morning on the drive to work the traction control fuel cut(?) feels more drastic since modifying the throttle. Still happy with the result. Will post my version of the tune to VersaTune site tonight.

ciprianrx8

There's no such thing. TC will only retard ignition timing AFAIK and under extreme conditions will reduce the opening of the throttle body.

Meat Head

I guess that’s what I interpreted as fuel cut - throttle closure. Steep learning curve for me and I’m skinning my nose on it.