|
nyet
|
 |
« Reply #15 on: May 16, 2011, 12:47:44 PM »
|
|
|
HOLY CRAP!
Bad news man, get a proper velocity stack before doing any tuning. Your MAF is reading garbage w/o anything there to straighten the flow and allow it to properly develop.
You need at LEAST 1 foot of ... something before the MAF sensor.
I can't stress this enough. FULL STOP. Do not do any further tuning until you get your intake sorted.
|
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
 Offline
Posts: 1056
|
|
« Reply #16 on: May 16, 2011, 01:50:26 PM »
|
|
|
HOLY CRAP!
Bad news man, get a proper velocity stack before doing any tuning. Your MAF is reading garbage w/o anything there to straighten the flow and allow it to properly develop.
You need at LEAST 1 foot of ... something before the MAF sensor.
I can't stress this enough. FULL STOP. Do not do any further tuning until you get your intake sorted.
Really? I don't see how this is any different then the stock MAF connected to the top of the airbox... it doesn't have any velocity stack either... Anyway, here is my current setup. Still not ideal, but better then before obviously. BTW my 85mm MAF housing has the flow straightener in it:   |
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #17 on: May 16, 2011, 02:31:26 PM »
|
|
|
Really? I don't see how this is any different then the stock MAF connected to the top of the airbox... it doesn't have any velocity stack either...
If you're running an airbox you don't need a stack... the airbox takes care of letting the flow form. If you have a short intake to a cone filter you need a stack or at least some straight pipe. |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #18 on: May 16, 2011, 07:02:03 PM »
|
|
|
If you're running an airbox you don't need a stack... the airbox takes care of letting the flow form.
If you have a short intake to a cone filter you need a stack or at least some straight pipe.
I see... well what I've got right now (above) seems to be working for me. I don't see anything strange going on in the logs in regards to the MAF values anyway. What should I be looking for if anything? Thanks again! |
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #19 on: May 16, 2011, 07:08:52 PM »
|
|
|
I see... well what I've got right now (above) seems to be working for me. I don't see anything strange going on in the logs in regards to the MAF values anyway. What should I be looking for if anything?
basically, you'll find that the less linear the MAF response is, the weirder the KFKHFM required to compensate (i.e. make your WB readings match your req AFR *and* have your trims, both short and long term, near zero). |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #20 on: May 16, 2011, 07:42:02 PM »
|
|
|
basically, you'll find that the less linear the MAF response is, the weirder the KFKHFM required to compensate (i.e. make your WB readings match your req AFR *and* have your trims, both short and long term, near zero).
Alright, well I'll def. keep an eye on things... I think with the silicon and elbow it's around 8-10 in total right now before the MAF. |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #21 on: May 17, 2011, 06:10:53 AM »
|
|
|
Ugh, the drive to work today was pretty crappy... I seem to have more lean spots now during cruise. It's pretty annoying. And of course I forgot my WBO2 @ home again...
I also think I'm a little lean in under boost/open loop now...
The worst part is the oscillating of the AFR under cruise... it mainly seems to run on the leaner side and then all of a sudden the ECU correct and the car picks up power and surges forward.
Exhaust note changes and everything...
|
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #22 on: May 17, 2011, 12:27:21 PM »
|
|
|
Just thinking about geometry here. It seems to me that the sensor opening is pretty far from exact center of maf housing, am I right?
The further it is from the center, the less air speed it will register at the same air flow as in the exact center. You might have to compensate for that in your MAF table or by KRKTE.
Sorry just noticed this post now... You are correct that it is not 100% in the center, I would think that since the MAF sensor is a fixed length going to a larger diameter that almost every MAF housing used on Stg 3+ cars would have a similar issue. I'm going to put my wideband in tonight and see what I can do... Because the car really drives like poo under partial load... Either that or I might just invest in the MAF/sensor that you have. If I do go that route, would you help me with interpolating the MAF values as you did with yours? |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #23 on: May 17, 2011, 09:49:47 PM »
|
|
|
Well installed the wideband today, and obviously the wideband shows completely different stuff then the ECU's NBO2's... it's constantly richer then what the ECU thinks it is. That said, I started to play around again with TVUB values... and I just can't wrap my head around it. Now I'm at a point where ultra light (20-23 in vac) load is stupid rich, idle is somewhat ok sometimes, lean when idle and voltage drops (ie. when the a/c and/or the EFK is on). Med load 5-15 in. vac is ok ish for the most part, sometimes too lean. And boost is again stupid rich now... Anyway where are some logs from tonight. Admittedly they're the best logs of the night...        Also the oscillation problem I've mentioned before doesn't (always?) seem to be related to AFR. My WB shows the AFR's to re relatively steady (0.1 AFR difference) while I can still feel the car pulling/not pulling sometimes during certain cruse ranges. It's really frustrating (to me) that it takes so long to make minor changes, especially since you cannot see what exactly is going on (not live). IMO that is the biggest draw back to trying to learn ME7 tuning like this (drive/log, tune, flash, drive again to log, analyze log, etc, etc...) I think I just need sleep  |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #24 on: May 18, 2011, 06:55:03 AM »
|
|
|
Question... KFKHFM adjustments...
Larger numbers are to correct for rich spots? or vice versa?
So larger # = leaner or richer?
|
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #25 on: May 18, 2011, 07:17:15 AM »
|
|
|
Question... KFKHFM adjustments...
Larger numbers are to correct for rich spots? or vice versa?
So larger # = leaner or richer?
It is a multiplier to amount of fuel. So anything below 1.0 will decrease amount of fuel and vice verse. I looked at that AFR graph. Is that from wideband? You know that your target AFR for WOT is AT MOST 11.8 but for a sake of safety (getting a batch of crap fuel) and reducing knock overall, it would be nicer to be at 11.5-11.6 across the board. If you jump over that 11.8, you will no doubt get a lot of timing retard/fuel dump which will result in quite perceivable slow downs. To massage AFR at wot you can use either KFKHFM or KFLF (despite the name of table it controls fueling across the board) or FKKVS (modifier for injectors RPMs/Pulsewidth). |
|
|
|
« Last Edit: May 18, 2011, 07:29:44 AM by julex »
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #26 on: May 18, 2011, 07:30:26 AM »
|
|
|
It is a multiplier to amount of fuel. So anything below 1.0 will decrease amount of fuel and vice verse.
Thanks! Also here is a screen of a problem area under cruise... You can see the AFR is all over the place even through am at a constant throttle angle.  |
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #27 on: May 18, 2011, 08:52:36 AM »
|
|
|
You noob!  That's how narrowband sensor operates lol. It can only measure exact 14.7 AFR/1.0 lambda -+ like 0.5AFR (maybe even less, it is very little usable range since the further you deviate from 14.7 the less resolution that sensor has) so it is constantly doing wild swings around and ECU tracks the pattern and is changing mixture so that oscillations are within expected range. I think it measures 0.5v or so around 14.7AFR with voltage swinging around a lot even with 0.1 - 0.2 change. It is completely normal. You should see your wideband sitting comfortably at target AFR of 14.7 while the O2s are on the roller coaster ride. |
|
|
|
« Last Edit: May 18, 2011, 08:55:31 AM by julex »
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #28 on: May 18, 2011, 09:03:00 AM »
|
|
|
You noob! That's how narrowband sensor operates lol. It can only measure exact 14.7 AFR/1.0 lambda -+ like 0.5AFR (maybe even less, it is very little usable range since the further you deviate from 14.7 the less resolution that sensor has) so it is constantly doing wild swings around and ECU tracks the pattern and is changing mixture so that oscillations are within expected range. I think it measures 0.5v or so around 14.7AFR with voltage swinging around a lot even with 0.1 - 0.2 change. It is completely normal. You should see your wideband sitting comfortably at target AFR of 14.7 while the O2s are on the roller coaster ride.  I'm just trying everything to understand the oscillation issue... so when I see that, and then further down the log (scroll to the right if you haven't seen the rest of the log because of the forum layout) the AFR's are steady what am I supposed to think lol.. anyway... I have a feeling I need to start from scratch again. Maybe even from a base M-box and just copy the linearization tables for the MAF, etc over and just start from the very beginning... Calculate a KRKTE, start with some close TVUB's and see what happens. What do you think? |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #29 on: May 18, 2011, 11:17:43 AM »
|
|
|
Checked my LTFT's after the drive to work this morning... partial is pegged @ -25%, even though @ partial the car is stupid lean for the most part (except at super light loads like @ ~20in vac).
LTFT's for idle is @ 0%.
STFT's are pegged at -25% as well.
Tried to get some logs this morning through ECUx, but lately it seems to be choosing what to log and what not to log... Not sure if this is due to the VM or not, either way I'm missing the entire 4k RPM section in two different logs (Logs fine from 3k to 3.9K, skips all of 4K and resumes at 5K RPM, time line is perfect), so they're useless.
Everything the logs are telling me makes no sense to me.
It shows the ECU is pulling fuel when it's already lean, the car jerks around like no ones business even when the AFR's are stable, anything below 4k rpm and from say -10 in vac to 5psi boost is stupid lean (15.0+ AFR).
|
|
|
|
|
Logged
|
|
|
|
|
