Customizing the DBW Throttle mapping
#1
Former Sponsor
Thread Starter
Customizing the DBW Throttle mapping
I've used my throttle map quite a bit to simplify the part throttle tuning process (discussed here), but I haven't really tried optimizing it for daily driving. I've always thought that the throttle is way too sensitive at low RPMs, which causes me to rev way more than I'd like before engaging the clutch from a stop in 1st gear. It makes me look like a novice with a manual transmission. However, I've noticed that the other DBW cars in FlashPro Manager have signfiicantly different target throttle maps that I think might help with this issue.
For reference, here is the "stock equivalent" target throttle map for the S2000. After studying my datalogs extensively, it's apparent that these numbers aren't really 1:1, since 66% on the throttle pedal causes the throttle plate to go to almost 100%. Basically, the top 1/3 of the throttle pedal is all 100%, and only the bottom 2/3 of the throttle pedal travel causes the thortlte plate to move. I'm not certain if the other DBW cars behave the same, but I suspect they do.
I see how each of the maps below would be better than the generic flat maps the S2000 have, but I really wonder how these were derived, or which approach might be best for the S2000. Have any of you guys played around with it? I'm sure many of you know that I'll do my usual geek-out thing on it eventually, but my FlashPro needs to be sent back for repair.
S2000 target throttle plate.
Civic Type R - Euro
Civic Type R - Asia
Civic SI - US
Civic R18 - US AT NOTE: I think this map is different from the MT version due to Honda wanting to minimize the throttle sitting in a specific zone where the automatic might "hunt".
Civic R18 - US MT
For reference, here is the "stock equivalent" target throttle map for the S2000. After studying my datalogs extensively, it's apparent that these numbers aren't really 1:1, since 66% on the throttle pedal causes the throttle plate to go to almost 100%. Basically, the top 1/3 of the throttle pedal is all 100%, and only the bottom 2/3 of the throttle pedal travel causes the thortlte plate to move. I'm not certain if the other DBW cars behave the same, but I suspect they do.
I see how each of the maps below would be better than the generic flat maps the S2000 have, but I really wonder how these were derived, or which approach might be best for the S2000. Have any of you guys played around with it? I'm sure many of you know that I'll do my usual geek-out thing on it eventually, but my FlashPro needs to be sent back for repair.
S2000 target throttle plate.
Civic Type R - Euro
Civic Type R - Asia
Civic SI - US
Civic R18 - US AT NOTE: I think this map is different from the MT version due to Honda wanting to minimize the throttle sitting in a specific zone where the automatic might "hunt".
Civic R18 - US MT
#4
Former Sponsor
Thread Starter
I started getting corrupted datalogs a few weeks ago, and FlashPro Manager started thinking that it needed to apply the same firmware updates over and over. Hondata says it probably has a bad flash ram chip, so I need to send it in for repair. I don't know what the warranty was on it, but it's apparently expired. The repair will cost anywhere from $50 to $250, depending on what they find.
#7
Former Sponsor
Thread Starter
I've already tried a few different throttle maps, and have definitely improved the pedal response at low RPMs / low speeds. I'll post the the map when I'm done with it.
Trending Topics
#8
Registered User
recently i was thinking about it.
what bothers me is s light delay in throttle response.
in FPM i see ~0.2sec delay between throttle pedal change and "throttle plate" change....
what bothers me is s light delay in throttle response.
in FPM i see ~0.2sec delay between throttle pedal change and "throttle plate" change....
#9
Former Sponsor
Thread Starter
Are you saying that the datalogs show a delay? If so, then I suspect much of that delay is really just a latency in the data. My datalogs show at least .04 seconds between unique values for TPlate or Tpedal, so I'm sure the CAN bus could account for about 0.1 seconds of that delay.