KevinD

iTrader: (56)

so we are working on evo 666s speed density tune and are having a odd problem. the car starts and idles decent (buschur bf272 cams), it cruises well, except at 3750 rpm. at 3750 rpm (and 0-3psi) the car begins to run lean and goes leaner and leaner until it begins bucking. it cannot go past this rpm point smoothly no matter what rpm VE or MAP VE settings i use. when it does go past, it runs smooth, builds boost fine, and and clean.

i've tried both the 9417 rom and the 9653 rom both have the same problem. i suspect it is a setting i am missing somewhere. we are using the kavlico 3.5bar map sensor (we have the 5bar waiting to get the car running right before switching over). i've changing the :

• Alternate Boost Desired Engine 'PSI' : PSIa8 JDM3barMAP -> PSIa8 OMNI4barMAP
• Boost Error Correction (PSI) : BoostErrorPsi JDM3barMAP -> BoostErrorPsi OMNI4barMAP
• Boost Desired Engine 'PSI' : PSIa8 JDM3barMAP -> PSIa8 OMNI4barMAP
• Boost Adder (PSI) : PSIa16 JDM3barMAP -> PSIa16 OMNI4barMAP

all to the appropriate kavlico settings.

any help would be appreciated!!

hediki12

those are omni 4bar settings....

KevinD

iTrader: (56)

i understand that, i changed them to the kavlico settings.

hediki12

o ok i thought you went jdm to omni . i hope you get help i havent gone s.d yet...

hediki12

umm do the ipw drop? i remember reading they had 2 ipw stutter fixes they should be in the xml just need to edit them.

RoadSpike

iTrader: (5)

Kevin,

Would you mind posting the map settings just for review, I'm pretty sure they are ok.

I have a suspicion that the issue may be in the Maf Comp (smoothing) table or Maf Scaling table if the problem is so directly tied to a specific rpm and boost level.

Edit:
I just read up on the thread the jitter fix is certainly interesting but I don't remember having this many issues when I tried it with the 9536 rom variant. Mine simply had ipw drop out at about the same boost level but wasn't progressive leaness and was easy to move past with a flick of the throttle.

KevinD

iTrader: (56)

the VE settings i have that seemed to work best are as follows:

RPM:
77.3438
78.9062
78.9062
79.6875
79.6875
80.4688
80.4688
81.25
81.25
83.5938
85.9375
88.2812
88.2812
91.4062
93.75
96.875
96.875

and MAP:
10.9887
21.0463
31.0106
40.975
55.0369
79.9944
100.016
499.988

(i know the last load point seems way out of line, but if i use a smaller value for the 300kpa range it only goes worse... with the 500 load point at least the car can get past the 3750 point.

i've tried everything from 200-500 in 30 increments. the 1byte load that the problem occurs at is around 160-180load for when using the huge 500 load calibration. if i try something smaller, say 350 load for 300kpa, the problem 1byte load becomes around the 120 range... this is a HTA86 turbo kit so i cant exactly built boost over that, and even trying to gently roll past the 3750rpm point it still causes a problem. every boost range from -4in/hg to the 3-4psi it does it, so i've concluded its something with the RPM point and possibly unrelated to the boost (although HUGE changes in the VE tables make it marginally better).


edit* and i've also tried RPM VE ranges from 100% down to 60's... doesn't make a difference at this problem point (but does everywhere else)

RoadSpike

iTrader: (5)

Thats pretty crazy so just to check the silly things.

You have changed the scaling on:
SD MAP Sensor VE and Calibration -> kPa Kavlico3.5barMAP

Valet mode and other things are all disabled for testing?

justboosted02

iTrader: (31)

Did you try increasing the RPM VE above 100% around 3500-4000? I run around 115% at some points in the table

scheides

iTrader: (10)

First, I would just put the 5-bar in and get it working properly on that straight away, otherwise you'll just be repeating your work.

I think the problem is your MAP VE table, second value from the top seems way low, and it is probably the load area you are having problems with. Log the car and see what the 1-/2-byte load is right as it is doing that. I bet the problem is occuring right around 120-160 load, and if the MAP VE value is too low in that area, it will be lean.

You might also consider rescaling some of the lower MAP VE values so you get a little more resolution in the problem area.

l2r99gst

iTrader: (2)

Why don't you post both sides of your VE tables. Hard to tell with only one side shown.

But, from what you have shown, I would say it's in your tune. Your RPM VE should be a curve with highest point in the midrange (depending on setup) and should taper to both ends.

MapVE should be similar, with lowest VE percentage (why we need to see both sides) at idle, and then possibly a 1:1 from there, if you want to use RPM VE only for tuning at WOT. But for mapVE, it's super important to have the correct scaling for your map sensor, which you say you have.

Did you do any maf logs to match the SD VE tune with? I'm assuming no, based on your values that you have right now.

KevinD

iTrader: (56)

the car was built with a forward facing turbo, so we are having to tune it from scratch on speed density (can't put a MAF on it).

the left sides of the VE tables are as follows (and again, i've tried many different things here as well):
MAP VE:
10.9944
21.1744
30.9472
41.1272
61.08
100.986
120.938
300.106

and RPM VE:
500
1000
1250
1500
2000
2500
3000
3500
4000
4687.5
5375
6062.5
6750
7437.5
8125
8812.5
9500



the weird thing is i took a rom from a working SD car i tuned as well, changed the injector scaling, and MAP sensor settings only, and it still did it.

l2r99gst, i would agree normally that this would be a tune problem, but it is so specific, and no tune settings seem to effect it at all, that i can only assume it is something to do with the kavlico sensor and setting it up. i've got the omnipower 4 bar, but because it is a AMS VSR intake manifold i need to make a bracket for it to run off a vacuum line.

the car is an '03 if that matters at all.

here is a log of it happening:


note that we were literally accelerating extremely slow (rpm line looks flat), we never even got into boost, just vacuum. if we try accelerating harder it just goes leaner at this one point

KevinD

iTrader: (56)

oh, and i checked a few logs, IPW doesn't appear to be dropping to 0 anywhere, but i shall try it logging nothing but IPW and RPM see if i can see something i'm missing.

scheides

iTrader: (10)

Can you post the whole table, possibly a screenshot so we can make sure everything seems right?

Wait, I think I see the problem. MAP 16bit. Did you rescale that for the new MAP sensor?

l2r99gst

iTrader: (2)

OK, let me post your tables again, so it's easier for everyone to read. Then, I can point out some things to change:

Map VE

10.99	10.99
21.17	21.05
30.95	31.01
41.13	40.98
61.08	55.04
100.99	79.99
120.94	100.02
300.11	499.99

RPM VE

500	77.34
1000	78.91
1250	78.91
1500	79.69
2000	79.69
2500	80.47
3000	80.47
3500	81.25
4000	81.25
4687.5	83.59
5375	85.94
6062.5	88.28
6750	88.28
7437.5	91.41
8125	93.75
8812.5	96.88
9500	96.88

If you multiply out these two maps, you will see that it's terribly lean and the areas where VE should be the highest.

From your log, you say that at around 3750 RPM and about -4 psi (roughly 70kpa) it's having this issue. Well, looking at your two maps, you are saying that map VE is about 60% and RPM VE is about 80%. Multiplied together, that is 48% VE. Way too low. That way my point about the tune and building the correct curves. If you make a 3D chart out of the two tables in Excel or something, you will see what I mean.

Now, assuming that you fix that up and still have the problem, there are two possible other issues I can see. (1) You don't have your map sensor setup properly...where did you get the mapVE settngs for it? (2) You may possibly be seeing the IPW jitter that I had seen in the past....mrfred posted a couple fixes to that.

But, I would first concentrate on properly tuning the VE tables. With that big turbo, I would suggest having a max VE somehwere around 6000RPM or so and taper both sides from there, moreso towards idle.