Smooth Throttle for UpRev

Awgd8 · 09-08-2014, 07:02 PM

Quote:

Originally Posted by wstar

Random notes from testing: the Throttle Position (as in butterfly) data in OBD-II may be faked, but it does track something roughly equivalent in VVEL if so. I did some logging on the curves while trying to roll through the throttle smoothly and consistently using different test maps. One was my standard smooth map to 3800, and the others were modified to smoothly reach 3200 and 4200. The funny thing is that in spite of the large change in the wide-open numbers, it was hard to detect any real difference in the graphs or the driving feel.

The 4200 map did tend to ramp out a bit quicker at the end (which just makes fine-grained control with the pedal more difficult), and the 3200 map did seem a little smoother at the top, but the differences were pretty trivial. It seems like the numeric maximums are largely irrelevant within reason. What causes throttle behavior changes is the shape of the curves (since they all started at the same low-end values, the 4200 map had a sharper slope than the 3200 map).

Also, based on testing, and based on reading some google search results of UpRev guys posting on other forums, reaching WOT at any RPM was never an issue on this car. Even on my stock throttle map, I can get the TPS output to peg when I stab the throttle from a stop in first gear. Apparently running a map like mine does remove some throttle latency in lower speeds/gears though.

What's left to test really is playing with the curvature. Right now I'm running linear (by percentage) from 18.8% to 96.9% on the table, and the TPS outputs still look fairly non-linear when rolling through the pedal smoothly. By that I mean, it goes relatively smooth for a while, but in the upper region it tends to just skyrocket relative to pedal position (i.e. more in the shape of an asymptotic graph), which I guess means I need to run the numbers up a little quicker near the bottom and flatten out the deltas in the upper-mid part. Going to try some other curves on the data and see what makes it smoother.

My AGWD8 Blend 2/3 happens to have 4200 max on the throttle table, so wstar originally experiment those 2 values (3200 and 4200) and did not feel a difference. I think this proves that in order to feel a significant increase in throttle response is to give more pronounce curvature on the graph and the right location where you needed it to kick in. Too smooth of a curve feels less torque all through out the powerband.

Now I wonder why the tuner (wstar post#1) graph has significant steep curvature from low end, mid and top... (see the graph he posted LIGHT BLUE LINES)

DIGItonium · 10-07-2011, 08:52 PM

No edit capability here... not without a license upgrade, but I don't know what I'm doing either.

Wow that's very interesting stuff. So there must be a lot going on with the throttle software... perhaps doing a bunch of SQL queries?

It's good to know about the response with ETC disabled, so it appears the VVEL actuators may have the mechanical capability to change position quickly.

True... in a sense "throttle position" is probably not relevant with our engines versus accelerator position. It's now VVEL actuator position... so what parameters are you able to log and monitor in real-time? Do you think you can record various parameters while making the video? Imagine overlaying the plots with your video. Someday I'll fetch a laptop and run Cipher to log stuff on the way to work and home.

Can't wait to see some videos! Be safe!

wstar · 10-07-2011, 10:14 PM

For these tests I'm just monitoring throttle position ong raphs generated by Torque on my phone while i drive. I could probably log that, pedal position, and some random VVEL parameters in Cipher alongside the usual rpm/speed info, but I'm not sure how I'd sync that with video.

DIGItonium · 10-08-2011, 12:34 AM

I've only spent 5 minutes looking at Cipher. So I guess you can preselect up to 20 parameters to record. When you're ready to test, start recording on the camera, then start log recording. Stop recording on both when you're done with the test and before proceeding with the next test.

What I did was ghetto-rig my digital camera to the steering wheel with masking tape. In fact, my camera was upside down and I rotated it using the video editor (MS Movie Maker). I don't know how to overlay image over video with Movie Maker, though.

If you can't overlay the image and video, separate post is fine.

wheee! · 10-13-2011, 08:46 PM

updates?

wstar · 10-16-2011, 09:30 AM

Top post updated with new results, and a new, better map. Read that post for all the updates...

cdoxp800 · 10-16-2011, 10:10 AM

Impressive work wstar

Zat_Zuma · 10-16-2011, 11:02 AM

Excellent work ......repped

N8GTOL · 10-16-2011, 11:34 AM

Nice write-up

DIGItonium · 10-16-2011, 05:17 PM

Great work! I'd love to see some videos. No wonder I had issues getting boost to build up at times especially when starting out at very low revs, and it takes forever to rev up to redline in 2nd as if the throttle is barely opening up.

I'm sure you've given tuners a great starting point.

Even though I can't edit the ROM on my own, perhaps Sam can take a look at this and send me a new flash. Maybe I can keep the standard response for daily driving and this one can be the "sport" map.

wstar · 10-16-2011, 06:11 PM

Well, even the "naive" approach will open that up for you, supposedly. The point of this exercise is to get that benefit while retaining a smooth, predictable response. I imagine some people will actually feel like their car is slower on my map, because in theory it should smooth the acceleration curve, resulting in less Jerk and/or Jounce which, I believe, is what leads people to believe the car has a very "snappy" throttle response based on Seat-of-the-Pants. I prefer predictability.

I have some more datalogging to do to back this up, and to make some estimation about how the 3800's concept helps low RPM driving force in general (smoothed out or not).

My plan for those charts is to log basically just rpm-vs-time (plus a few other parameters for my own validation) as an accurate vehicle speed metric, do it for a 2nd gear pull flooring the pedal from 1500 RPM on a few different maps (assuming I can do that without wheelspin, otherwise I'll go to 3rd again), and then derive accel -> jerk -> jounce data from that, or at least to whatever practical limit there is in the data.

wstar · 10-16-2011, 08:15 PM

I did some data logs between the good curve and the stock map on raw acceleration with a stomped full throttle from 2k -> top in 2nd gear, and to be honest at first glance the difference isn't huge.

Of course this isn't engine tuning, this is throttle tuning, we don't expect to make the engine accelerate any faster in most of the range. What I expected was a notable difference in the acceleration in the low RPM range though, basically from allowing the virtual pedal to fully open like the real pedal did, and what I found was a relatively small one (in the right direction). Subjectively, the effect seemed a lot more noticeable though. For the short time window that matters we don't have a ton of sample points either, and there is a tiny bit of sensor -> cipher lag as well. All of that could be dampening the results.

Don't have time to run the numbers on the numbers, but I'll do that later and make some graphs.

wstar · 10-17-2011, 08:46 AM

Well, for now I'll stick by my subjective impression (which was fairly strong) that UpRev and the tuners are right: maps with a solid column (or more) of 3800 values at the end do get on the throttle quicker at lower RPMs when you stomp the pedal all the way.

The MPH and RPM data via Cipher simply isn't good enough to infer g's of accel at the fine scale needed to tell the difference between the stock and non-stock maps at low RPMs in the scenario above, and adding to the error is my inability to get the graph start points lined up precisely (same starting pedal pos + rpm, leading to same timestamp for first full-throttle sample).

I'll post up this chart here anyways so you can see what the data looks like, but the bottom line to me is "too noisy and inaccurate to tell anything from". MAP4 is stock, MAP3 is one of the 3800 maps (the 2300 one in this case, not that it matters which at full throttle):

DIGItonium · 10-17-2011, 12:44 PM

Have you driven a DE/HR car lately? I wonder if you can get your impressions with those cars in comparison to your map.

So how's daily driving coming along?

wstar · 10-17-2011, 01:40 PM

No, I've never driven the DE or HR motors. Daily driving on the "curve 2300" map is great, but so was the "curve 2700" as long as you didn't go full throttle, and the original "linear" was pretty decent as well, although it seemed a little snappy or sensitive in the upper parts. Really even the stock map is reasonable too, so long as you're always at say ~3.5K+ RPM. The tuner map was kinda awful though, lots of odd unpredictability at partial throttle.

Sorry I haven't posted a video. I actually have a good setup for shooting video out the front windshield (which I used for this), but not for recording my dash dials while driving. I'll take some video on stock vs curve2300 sometime this week though, using aLapRecorder and virtual gauge overlays for RPM and TPS, like I did for the track stuff. I just need to get it all set up again after (yet again) updating my beta android firmware.

10-07-2011, 08:52 PM	#2 (permalink)
DIGItonium A True Z Fanatic Join Date: Jan 2009 Location: Kansas Posts: 3,800 Drives: 09 Z34-TT 6MT Rep Power: 43	No edit capability here... not without a license upgrade, but I don't know what I'm doing either. Wow that's very interesting stuff. So there must be a lot going on with the throttle software... perhaps doing a bunch of SQL queries? It's good to know about the response with ETC disabled, so it appears the VVEL actuators may have the mechanical capability to change position quickly. True... in a sense "throttle position" is probably not relevant with our engines versus accelerator position. It's now VVEL actuator position... so what parameters are you able to log and monitor in real-time? Do you think you can record various parameters while making the video? Imagine overlaying the plots with your video. Someday I'll fetch a laptop and run Cipher to log stuff on the way to work and home. Can't wait to see some videos! Be safe! __________________ Last edited by DIGItonium; 10-07-2011 at 08:57 PM.

10-07-2011, 10:14 PM	#3 (permalink)
wstar A True Z Fanatic Join Date: Mar 2009 Location: Houston, TX Posts: 4,024 Drives: too slow Rep Power: 3595	For these tests I'm just monitoring throttle position ong raphs generated by Torque on my phone while i drive. I could probably log that, pedal position, and some random VVEL parameters in Cipher alongside the usual rpm/speed info, but I'm not sure how I'd sync that with video. __________________ 7AT Track Car! Journal thread / Car setup details

10-08-2011, 12:34 AM	#4 (permalink)
DIGItonium A True Z Fanatic Join Date: Jan 2009 Location: Kansas Posts: 3,800 Drives: 09 Z34-TT 6MT Rep Power: 43	I've only spent 5 minutes looking at Cipher. So I guess you can preselect up to 20 parameters to record. When you're ready to test, start recording on the camera, then start log recording. Stop recording on both when you're done with the test and before proceeding with the next test. What I did was ghetto-rig my digital camera to the steering wheel with masking tape. In fact, my camera was upside down and I rotated it using the video editor (MS Movie Maker). I don't know how to overlay image over video with Movie Maker, though. If you can't overlay the image and video, separate post is fine. __________________

10-13-2011, 08:46 PM	#5 (permalink)
wheee! A True Z Fanatic Join Date: Sep 2009 Location: Mod Capital of Canada Age: 58 Posts: 7,354 Drives: cars. lots of cars. Rep Power: 40182	updates? __________________ H&R / Stillen / Berk / HKS / Swift / Michelin / UpRev / SPC / Amuse Rep / Seibon / Fringe XVO / TWM / more... My Journal

10-16-2011, 09:30 AM	#6 (permalink)
wstar A True Z Fanatic Join Date: Mar 2009 Location: Houston, TX Posts: 4,024 Drives: too slow Rep Power: 3595	Top post updated with new results, and a new, better map. Read that post for all the updates... __________________ 7AT Track Car! Journal thread / Car setup details

10-16-2011, 10:10 AM	#7 (permalink)
cdoxp800 A True Z Fanatic Join Date: May 2009 Location: FL Posts: 1,337 Drives: 2011 Nismo Rep Power: 19	Impressive work wstar __________________ 2011 Brilliant Silver NISMO 370z. #: 0046. PPE LTH \| MillSpec Cats \| F.I Exhaust \| Akuma Intakes \| Rotora 6/4 Brakes

10-16-2011, 11:02 AM	#8 (permalink)
Zat_Zuma A True Z Fanatic Join Date: Nov 2008 Location: Edmonton, Alberta ...eh Posts: 1,171 Drives: 2017 Touring auto Rep Power: 19	Excellent work ......repped __________________ [B]2009 & 2017 370Z Touring Sport Autos SOLD

10-16-2011, 11:34 AM	#9 (permalink)
N8GTOL Base Member Join Date: May 2010 Location: San Diego Posts: 228 Drives: and flies Rep Power: 22	Nice write-up

10-16-2011, 05:17 PM	#10 (permalink)
DIGItonium A True Z Fanatic Join Date: Jan 2009 Location: Kansas Posts: 3,800 Drives: 09 Z34-TT 6MT Rep Power: 43	Great work! I'd love to see some videos. No wonder I had issues getting boost to build up at times especially when starting out at very low revs, and it takes forever to rev up to redline in 2nd as if the throttle is barely opening up. I'm sure you've given tuners a great starting point. Even though I can't edit the ROM on my own, perhaps Sam can take a look at this and send me a new flash. Maybe I can keep the standard response for daily driving and this one can be the "sport" map. __________________

10-16-2011, 06:11 PM	#11 (permalink)
wstar A True Z Fanatic Join Date: Mar 2009 Location: Houston, TX Posts: 4,024 Drives: too slow Rep Power: 3595	Well, even the "naive" approach will open that up for you, supposedly. The point of this exercise is to get that benefit while retaining a smooth, predictable response. I imagine some people will actually feel like their car is slower on my map, because in theory it should smooth the acceleration curve, resulting in less Jerk and/or Jounce which, I believe, is what leads people to believe the car has a very "snappy" throttle response based on Seat-of-the-Pants. I prefer predictability. I have some more datalogging to do to back this up, and to make some estimation about how the 3800's concept helps low RPM driving force in general (smoothed out or not). My plan for those charts is to log basically just rpm-vs-time (plus a few other parameters for my own validation) as an accurate vehicle speed metric, do it for a 2nd gear pull flooring the pedal from 1500 RPM on a few different maps (assuming I can do that without wheelspin, otherwise I'll go to 3rd again), and then derive accel -> jerk -> jounce data from that, or at least to whatever practical limit there is in the data. __________________ 7AT Track Car! Journal thread / Car setup details

10-16-2011, 08:15 PM	#12 (permalink)
wstar A True Z Fanatic Join Date: Mar 2009 Location: Houston, TX Posts: 4,024 Drives: too slow Rep Power: 3595	I did some data logs between the good curve and the stock map on raw acceleration with a stomped full throttle from 2k -> top in 2nd gear, and to be honest at first glance the difference isn't huge. Of course this isn't engine tuning, this is throttle tuning, we don't expect to make the engine accelerate any faster in most of the range. What I expected was a notable difference in the acceleration in the low RPM range though, basically from allowing the virtual pedal to fully open like the real pedal did, and what I found was a relatively small one (in the right direction). Subjectively, the effect seemed a lot more noticeable though. For the short time window that matters we don't have a ton of sample points either, and there is a tiny bit of sensor -> cipher lag as well. All of that could be dampening the results. Don't have time to run the numbers on the numbers, but I'll do that later and make some graphs. __________________ 7AT Track Car! Journal thread / Car setup details

10-17-2011, 08:46 AM	#13 (permalink)
wstar A True Z Fanatic Join Date: Mar 2009 Location: Houston, TX Posts: 4,024 Drives: too slow Rep Power: 3595	Well, for now I'll stick by my subjective impression (which was fairly strong) that UpRev and the tuners are right: maps with a solid column (or more) of 3800 values at the end do get on the throttle quicker at lower RPMs when you stomp the pedal all the way. The MPH and RPM data via Cipher simply isn't good enough to infer g's of accel at the fine scale needed to tell the difference between the stock and non-stock maps at low RPMs in the scenario above, and adding to the error is my inability to get the graph start points lined up precisely (same starting pedal pos + rpm, leading to same timestamp for first full-throttle sample). I'll post up this chart here anyways so you can see what the data looks like, but the bottom line to me is "too noisy and inaccurate to tell anything from". MAP4 is stock, MAP3 is one of the 3800 maps (the 2300 one in this case, not that it matters which at full throttle): __________________ 7AT Track Car! Journal thread / Car setup details Last edited by wstar; 10-17-2011 at 08:55 AM.

10-17-2011, 12:44 PM	#14 (permalink)
DIGItonium A True Z Fanatic Join Date: Jan 2009 Location: Kansas Posts: 3,800 Drives: 09 Z34-TT 6MT Rep Power: 43	Have you driven a DE/HR car lately? I wonder if you can get your impressions with those cars in comparison to your map. So how's daily driving coming along? __________________

10-17-2011, 01:40 PM	#15 (permalink)
wstar A True Z Fanatic Join Date: Mar 2009 Location: Houston, TX Posts: 4,024 Drives: too slow Rep Power: 3595	No, I've never driven the DE or HR motors. Daily driving on the "curve 2300" map is great, but so was the "curve 2700" as long as you didn't go full throttle, and the original "linear" was pretty decent as well, although it seemed a little snappy or sensitive in the upper parts. Really even the stock map is reasonable too, so long as you're always at say ~3.5K+ RPM. The tuner map was kinda awful though, lots of odd unpredictability at partial throttle. Sorry I haven't posted a video. I actually have a good setup for shooting video out the front windshield (which I used for this), but not for recording my dash dials while driving. I'll take some video on stock vs curve2300 sometime this week though, using aLapRecorder and virtual gauge overlays for RPM and TPS, like I did for the track stuff. I just need to get it all set up again after (yet again) updating my beta android firmware. __________________ 7AT Track Car! Journal thread / Car setup details

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
370Z FI Throttle Response: Boost vs. Throttle Position	DIGItonium	Forced Induction	9	01-23-2011 08:49 PM
Adjusting Throttle Response with UpRev?	AlphaSnacks	Tuning	4	12-05-2010 03:40 AM
2nd gear not as smooth as the others???	VQinside	Engine & Drivetrain	14	11-03-2010 09:22 PM
UpRev Osiris/Technosquare users- throttle mapping?	SE	Tuning	3	08-29-2010 05:03 PM
Smooth launch in 1st Gear?	BrianMSmith	Nissan 370Z General Discussions	22	03-26-2010 10:40 PM