|
julex
|
 |
« Reply #45 on: May 24, 2011, 01:41:23 PM »
|
|
|
MAF is always l/h inside ME tables. VCDS/EcuC reads g.s. Conversion factor is 3.6, eg. 1g=3.7l/h
Having said that, these two maps for MAF are very far off, by 25%. Without graphing them I cannot tell if one is simply the first one multiplied by 125% or something else is going on there.
I think you need to scratch the table and go back to basics.
Take M-box maf table which is a good basic shape for Hitachi MAF (you're not running Bosch by any chance?) and use it to calculate theoretical MAF curve for your housing.
You need to figure out what the multiplication/division factor is to transform stock MAF to your 90mm (right?). It isn't hard:
Cross Section Area = Pi * (r^2)
Calculate both for stock and your housing. Divide resulting number for your housing by stock housing number and you will get the multiplication factor. Multiply all cells in the MAF table by that number, there, you should have something you can work with.
If this still doesn't work for you you probably have some serious issues with flow inside that MAF housing.
|
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
 Offline
Posts: 1056
|
|
« Reply #46 on: May 24, 2011, 02:02:48 PM »
|
|
|
Well... you just made me think of something.
I'm using a Bosch MAF housing modified for the Hitachi sensor. In my mind I figured an 85mm MAF housing is an 85mm MAF housing... but I guess there is more to it then that.
2nd. If I have two programs, both for 85mm housings, both M-boxes, but one drives fine at part throttle and idle and the other doesn't (nefmoto) I tend to lean away from it being a hardware issue to be quite honest. And when I say the other file drives fine, I mean the car drives like stock pretty much. I just don't want to run an underscaled MAF and obviously a few other reasons to not use that file.
|
|
|
|
|
Logged
|
|
|
|
Matt Danger
Full Member
Karma: +17/-1
Offline
Posts: 116
|
|
« Reply #47 on: May 24, 2011, 04:51:19 PM »
|
|
|
I'm using a Bosch MAF housing modified for the Hitachi sensor. In my mind I figured an 85mm MAF housing is an 85mm MAF housing... but I guess there is more to it then that. My Hitachi is molded into the RS4 Bosch housing very similar to the way you did yours without problems. |
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #48 on: May 24, 2011, 04:58:47 PM »
|
|
|
Hm. Well if you don't have a boost leak, i'd drive around randomly at different load/rpm operating points and log STFTs (make sure your ltfts are stable! if not, your stft data might not be accurate!)
Then make a scatter plot and see if you can see any patterns.
If you want, you can post the log here and I can try to post up some 3d gnuplot visualizations. I have a couple scripts I used for it.
|
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #49 on: May 24, 2011, 08:35:41 PM »
|
|
|
^^^ Thanks Nye, so on my last revision for today my LTFT's were stable and I logged the following STFT's & MAF/RPM readings via VCDS after getting off the highway and stopping for gas. It was a drive through the city trying to hit and hold different load points.
Now I noticed that VCDS would stop/freeze during logging and then resume so if you notice "blocks" of data and then nothing or the same values that would be it.
I'm assuming my cheap china VCDS cable is to blame.
BTW, With my current tweaks (some very minor ones I did blindly to the MAF compensation) and just by TVUB I've gotten the car to drive very smooth... it still has some issues here and there... but it's already much better.
|
|
|
|
|
Logged
|
|
|
|
Jason
Hero Member
Karma: +38/-0
Offline
Posts: 500
Breaks everything!
|
|
« Reply #50 on: May 24, 2011, 11:37:50 PM »
|
|
|
I have the freeze issue occasionally with even my newest, latest and greatest vag-com cable. I think it's ECUx to blame, lol.
|
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #51 on: May 25, 2011, 06:37:12 AM »
|
|
|
I have the freeze issue occasionally with even my newest, latest and greatest vag-com cable. I think it's ECUx to blame, lol.
Yes quite possible  but this time it was VCDS and not ECUx.... I tried ECUx after logging with VCDS and ECUx wouldn't connect at all... Then again this all could be because of the VM... I will try later on to see if I have these issues on a pure WinXP machine. No VM. |
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #52 on: May 27, 2011, 02:23:26 PM »
|
|
|
Sorry man. I looked at your logs, and i can't see any patterns to your STFTs  In my case, it was pretty obvious, there were rich/lean spots that were obvious at several load/rpm points, but it was mostly deterministic. Yours are all over the place.. Maybe try swapping O2s and EGT sensors ... |
|
|
|
|
Logged
|
|
|
|
|
judeisnotobscure
|
|
« Reply #53 on: May 27, 2011, 05:23:18 PM »
|
|
|
I was just looking at some logs from a stage 2 s4 that had very similar o2 readings to yours on two different files, giac and custom.
ended up being a vaccum leak on the turbo inlet pipes... silicone around the joints seemed to fix the issue.
|
|
|
|
|
Logged
|
I have a b5 s4
but i just want to dance.
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #54 on: May 28, 2011, 12:01:59 PM »
|
|
|
Sorry man. I looked at your logs, and i can't see any patterns to your STFTs In my case, it was pretty obvious, there were rich/lean spots that were obvious at several load/rpm points, but it was mostly deterministic. Yours are all over the place.. Maybe try swapping O2s and EGT sensors ... I know they're all over the place.. that's what worries me... Anyway, O2 sensors and EGT's were never faulty before the swap (heck even when I was driving the Tial setup on my Stg 2+ tune and stock fueling) and I don't believe in them being bad either if you look at my ECUx logs. They're not triggered at any point in time and the O2's also read nicely. So saying to spend ~$400 on EGT's and another $~150 on O2 sensors on a whim is not going to happen in the slightest. I have not however yet done a pressure test, so that is on my list of to-do's this weekend sometime, along with swapping to the other M-Box tune that makes the car seem to drive butter smooth and take some STFT/MAP logs then to compare with this current tune. I'm still a believer that it is a software/tune issue rather then a hardware issue. A leak that would cause such random lean/rich area's under VAC seems odd to me, since a) the inlets are one piece and there are no leaks points visible to me. B) If there was a boost leak I would think making the HP I did @ the boost level I am at would prove to be difficult and also cause my AFR's to be a little wacky for the loss of metered air, no? It just doesn't add up to me. Thanks for looking over the logs guys... I will update as soon as I start to tinker again. Cheers! |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #55 on: June 07, 2011, 08:00:08 AM »
|
|
|
Well, after some TLC yesterday evening: http://www.audizine.com/forum/showthread.php/421890-NOTORIOUS-VR-s-Stock-2.7T-TiAL-S605-Build?p=6577854#post6577854The car seems to run much better... I decided it was time for a new revision as well.. TVUB +0.005 from 7 to 12V KRKTE is now at 0.048 (from 0.051) slightly increased the MAF compensation table under higher loads from 3k on up to redline (I think to 0.99 tapering to 0.96). Slightly increased component protection (load fueling) from 5500 on up to prevent it from going too lean by redline. Car seems to run quite nice now. Nye, not sure if you can have a look at the quick STFT log from this morning. IT's not very long and still might have weird gaps from the logging randomly stopping (BTW I think I figured out what is causing that! It's the G sensor in the Thinkpad that turns off the HDD! I will verify later today after work by turning it off) I've also attached a VCDS log of the LTFT's and STFT's (which from what I can see look pretty good). Thought's and suggestions when you guys have them would be great! |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #56 on: June 08, 2011, 05:06:39 PM »
|
|
|
Another small update... revised the MAF compensation table a little bit, reduced KRKTE to 0.046 (was still too rich @ .048), slightly increased TVUB across the board (by 0.035). Also zero'ed out DSLOFS this time around, but I'm still just running 0% DC on the N75.
Car is driving nice and pulls well and AFR's are starting to look really nice to me, maybe just a touch too right from 10 PSI and higher but nothing to worry about (~0.87)... The only issue I seem to be having is at 21 in Hg and less throttle it goes stupid rich (0.78-0.85 lambda) anywhere from 1k to as far as 5k rpm... Would this be something to correct in the MAF tables as well?
I haven't checked the FT's yet with this revision, and no 3rd gear logs either because of the stupid hot weather (107 w/ humidity) today.
|
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #57 on: June 09, 2011, 05:19:11 AM »
|
|
|
So this is where my LTFT's are currently sitting (screenshot from carputer)...  If it weren't for 2 things I would be happy with the car right now. 1. Very light load (21 in HG and less) it's stupid rich. 2. between 2k and say 3.8k the car stumbles sometimes. 2b. I think the car is injecting too much fuel during starting now, gonna pull back from sold/hot start enrichment see if it gets better. Today it's supposed to thunder and rain, but tomorrow it's supposed to be nice and not too hot so I'll see if I can get some nice complete ECUx logs then. |
|
|
|
« Last Edit: June 09, 2011, 05:22:56 AM by NOTORIOUS VR »
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #58 on: June 09, 2011, 10:20:16 AM »
|
|
|
Also zero'ed out DSLOFS this time around, but I'm still just running 0% DC on the N75.
If you are planning to run boost past the MAP limit, you may want to keep DLSOFS stock and use a locked WGDC file. Also, i might pull a bit more fuel from KRKTE.. |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #59 on: June 09, 2011, 11:18:38 AM »
|
|
|
If you are planning to run boost past the MAP limit, you may want to keep DLSOFS stock and use a locked WGDC file.
Also, i might pull a bit more fuel from KRKTE..
Ahh I understand, to give a little extra resolution with PID control you zero out the sensor? I'll put that back then since I see 24psi on the WG alone. Yes you're right I'm going to pull a little more from KRKTE still as well. Any ideas on the super light cruise being very rich? |
|
|
|
|
Logged
|
|
|
|
|
