evank

iTrader: (2)

I'm posting this on the advice of AbeFM, who looked at my datalog and decided something is amiss.

The log is at http://www.snarc.net/log-oct-2.xls and the msq is at http://www.snarc.net/evan_hr_64.msq.

Abe wrote:

Wow, you've got something bad-wrong. I don't know what yet. I It looks like you're constantly blipping the throttle. I was thinking it was all in my head, since the longest in the ENTIRE log it was on any single value other than 0 was... 3 seconds? Maybe 2. You drove for how long going "blip blipblipblip blip blip"?

On the other hand, the manifold pressure seems to follow it ... It looks WAY weird.

Then he added:

Something is odd, though, why does the MAP line bounce with the throttle. That makes it look real. I'm thinking the thottle might be loose? Or, shoot, I don't know. The fact that the two move together....

A hose or the crankcase breather could make the MAP fluctuate, but not the TPS. A bad sensor or bad ground could make the TPS look funny. But for both to move at once.. Does the car lurch? (Evan here again: yes, it does lurch.) Could you have a setting wrong so it's double using a pin? (Evan here again: Beats me! DIY built this - 3.57, parallel - and Paul installed it. All I know how to do is tune, but it's plausible that I fucked up something.) That can happen on MS-II, you tell it to turn the fans on pin8 and read the baro on pin 8 and it goes crazy whenever the fans are on .... post a thread up, this is worth it.

Suggestions anyone? I wasn't driving erratically when I datalogged.

Things felt MUCH better when I accepted the changes using this datalog and MLV's VE analyzer. Even the idle was better, despite the fact that I told VE analyzer not to change any cells below 30 MAP. The car was smoother and had more power; that msq is at http://www.snarc.net/evan_hr_65.msq but I haven't yet logged with it. I did notice that the new VE table (using MegaTune's default RPM columns, vs. my old custom ones) has a lot more fuel in the middle cells. This was on the "normal" setting in VE analyzer. However Abe advised that I seek and fix the problem before using this new msq.

Big caveat, which I forgot to mention to poor Abe: my laptop recently died, so I got a new one, and obviously had to reinstall MT and MLV. Maybe some important underlying setting changed, for better or worse? I used the original project files (rescued by putting my old laptop's HD into a USB enclosure.)

Off to sleep now. Looking forward to seeing replies tomorrow!

jbresee

Where is the vacuum line connected for the MAP sensor?
Do you have a filter in MAP line?

Mine was bouncing between accell / decell and the MAP was bouncing with it until I moved the line and put the fuel filter on it. That dampened the MAP oscillation.
Threads should show up in search under "map pressure wave" I think.

evank

iTrader: (2)

Latest datalog: problem remains. Also car is very rich now. I don't know why MLV did that.

http://snarc.net/datalog200910051740.xls

evank

iTrader: (2)

Paul informed me that I grossly misunderstood Abe's comments.

I thought Abe was saying that MAP and throttle are NOT supposed to move together. Apparently he was only saying they don't normally move THAT MUCH, but the "together" part is what's supposed to happen.

I also always thought that when taking a datalog, you're supposed to drive aggressively, in order to reach as many cells as possible. Do you guys have other approaches?

Now I understand why Abe also told me to datalog while driving like I'm going to the store for an errand .... i.e. not like I'm tuning for max performance. Doh!

I'll get on that .... tomorrow.