logged some 1st gear throttle lag
 

1st gear, rolling start at 10mph, ambient temps were about 80deg, 200 coolant temp, 200 oil temp, VDC off.

you can see a full 1 second delay before the accelerator pedal voltage reaches 4.6v, after that an incredibly boring 3 seconds until 4k rpm is reached.

my throttle maps are still stock, and I havent yet logged when the engine is cold which is next on my lest. I'm also going to double check the break pedal switches to ensure they are fully depressed when the break is released, which has been known to cause similar throttle behavior.


http://i.imgur.com/N3G4FnT.png

Of topic question, what are you using to run Uprev/cipher on your Mac? I ended up buying a cheap PC laptop just to run those programs. :icon17:

it's a windows XP VM (virtual box)

Try again starting at 2k RPM. IIRC, no lag.

Saw your PM this morning. The delay in the accelerator pedal reading is likely just due to the OBD2 lag, but there shouldn't be a big lag in the throttle, I'm typically seeing something like 300ms from pedal to throttle body. I can try to repeat your tests with my logger if you'd like.

BTW have you tried with TCS off? sometimes it cuts throttle in 1st/2nd before long before the slip indicator comes on.

VDC was off, logging was done via uprev which most definitely uses canbus (my pm was about something else entirely). the delay is not only very clear in logging, but its extremely noticeable when driving, which is why I started logging to begin with. it seems that the behavior is not consistent from car to car, so your logs may not show anything near what my car does. further, when the car is cold the pedal and throttle lag are almost entirely gone, but i use the term "cold" loosely here as the water and oil were already warmed up since its summer, so I have no idea as of yet what is causing the lag to occur after a bit of driving.

my next step is going to be disabling ETC entirely and seeing if the lag is gone, which would mean the issue could be solved with the throttle maps.

Today I disabled ETC (electronic throttle control) entirely, unfortunately this is not map switchable and requires reflashing the ecu. For track days ETC off is ideal as the ECU never interferes with the throttle but since the ecu cannot control the throttle itself, the downsides are numerous:

- no throttle cut at rev limiter (safety before fuel cut)
- no SRM
- no cruise control
- no throttle control for VDC, in theory with VDC on it would still control the brakes but not the throttle which could lead to some nasty conditions

There is a huge difference in throttle response, and I get full TPS voltage with one exception. it seems that below 2000 RPM the throttles don't like to be fully open regardless of pedal position, once you get past 2k you get full TPS voltage. So it would seem that my throttle issues are solved with ETC off, and that means theres hope to eliminate the weird throttle issues by adjusting the Throttle maps in uprev.

that, my friend, is due to the high intake air temperature.
104*F or 40*C is high in my books and will no doubt cause that huge lag.
Water temp at 199/92c might also have played a role here but first blame is the air temp pulling timing.

What intake did you have on at the time?

When I researched how VVEL works I found this, VVEL seminar PDFthis would check out with the delay in throttle position sensor and why it is at max voltage. When you disable ETC it just lowers the rpm where vvel takes over airflow.

The thing is, most people have been misled to believe that the throttle bodies on our engines are wide open all the time. This is definitely NOT true as i can watch the throttle position sensor angle live on my Scangauge and anyone with an obd2 reader can too.

I wonder what would happen if the throttle bodies were removed from our cars and a dummy fitted to make up the gap - obviously the real throttles would need to remain connected, just zip tied to the side.

I believe most of the throttle plate actuation is at idle and low load --- after that, VVEL handles most of it. Disabling ETC probably just turns all but idle and dashpot (off throttle, coast-down) correction to VVEL.

Immediate throttle response -- tip in, especially off idle -- is probably primarily handled by the throttle maps.

I think someone recently posted some graphs that overlay TPS and VVEL that I think more or less aligns with that interpretation... but someone correct me if I'm mistaken.

Keep in mind the timing inaccuracies of virtual machines.

http://www.vmware.com/files/pdf/Time...alMachines.pdf