NefMoto

OK, I'm crying 'Uncle'.

What I want to do is reduce lag off the line under heavy throttle with my frankenturbos (tiptronic).  Reducing timing pretty drastically helps a great deal, as others suggested.  But what I don't want to do is have the timing that far reduced under lighter loads at low RPMs.  Problem is, the X axis of KFZW is actual load, not requested load.  So just modifying existing KFZW isn't a great answer.

What I think would work is by switching to a different KFZW, similar to what rajivc666 has done with map switching using the cc switch (http://nefariousmotorsports.com/forum/index.php?topic=1556.0), only triggering by WPED (or is it WPED_W?) over "X" percent.  If I were using an M-box, I think I could manage, since most everything has been worked out.  I'm using an R-Box (BEL), and I'm totally balled up finding the right routine to hook into.

Anybody willing to port KFZW map switching to the R-box, using WPED and speed as I described?  Just the code is fine; I've got patching down ok.  (And once this is working, I'll probably have enough info to change fueling as well, giving more benefit yet.)

Attached is the .bin I've been working with.  I added a 2nd KFZW at 0xC6B01, which was, I know, the easy part.  (The tune is fairly close to finished and is pretty smooth.  Stock injectors at this point.  No fueling changes since I do have some EV14s to drop in.  Cats and precats still in place. It uses the RS4 KFMIRL and KFMIOP to improve part-load driveability, and has modified Q1, Q2, KFLDRL and and KFLDIMX to suit my set-up.  And LDRXN and KFLDHBN, of course.)  XDF is also attached, with the new KFZW_antilag defined.

Thanks in advance.  Plenty of process control programming experience, but none at the assembly level.

Ignition routine is at 0xC22A6

I'm don't think this is the solution to your problem.

This sounds like torque intervention to me. Is it just slow to meet target boost?

Another issue that has been cropping up with Frankenturbos is mechanical wastegate issues causing slow spool.

Is your transmission built? Will it be able to handle the additional torque from this?

Why use wped and not rlsol?

Pretty sure it is simply slow spool due to lack of exhaust energy off-line, and not torque intervention.  I'm at 6,200'/1900 m, so essentially I have a low CR 2.7L down on power by 20% until there's enough energy to make things happen.  I've adjusted KFWDKSMX to 100% by 2,000 rpms and actual tracks.  That helped quite a bit.  By the fact that I'm able to drop spool (defined as 2 psig just to be consistent) to 2400 rpms or so from 2,750 just by pulling 15 degrees of timing from off-idle to 2,500 rpms, it certainly should help and having the separate map tracking off of a third parameter will give a little more control.

I had heard about the WG problem and inspected welds etc. before installing, and the quality was actually excellent.  Not to say that something hasn't happened in the meantime, but it doesn't seem to have changed over time.

Driveline drag has a bunch to do with it, as a rolling start from 5 mph/8 kmh and things are much better.  TC stall speed is per spec.

No, trans is not built.  I am running 20% non friction-modified fluid as an experiment that is working so far (based on the research of others).  Eventually I'm going to swap to a 6MT, so that's where the $$$ are going (more expensive than finding a different car, but I know this one too well and it is in too terrific of condition to want to switch.)

rlsol would work, too.  Advantage(s)?

Thanks for the input (matchew, too).  Always willing to learn from the experts.

I completely disagree with this. Reducing the timing will make your car slower, rather than faster, and this will be reflected in FATS times as well.
Reducing timing will *not* help you off the line, and pressure in an IEC is just a measure of resistance.
By running MBT the car will accelerate faster through the rev range and build more boost quicker.

Your request is for a problem, which does not exist, based on an assumption, that is incorrect.

Are you looking for a 2 map throttle based switch? Ie below say 50% throttle one map and above 50% another?

or simply compare nmot and rlsol to some constants of your choosing, and sub 15 from output of KFZW until nmot is above your constant or rlsol is below your other constant.

It's not an assumption and I'm not doing this willy-nilly.  I'm an ME with 20 years of rotating machinery experience, specifically with combustion turbines (read: thermodynamics).  Troubleshooting is pretty much what I do for a living, and that calls for taking a disciplined approach.  I've experimented quite a bit with this (but not to an end-point yet, admittedly) and, point for point, MAF is higher and acceleration is better when I pull the timing and get boost more quickly from *off the line*  I'm not talking about doing this under other circumstances, only for across-the-intersection type of thing.  Now, might there be a greater affect yet from *adding* timing, as in don't worry about boost and treat it as a NA engine for that first few seconds?  Maybe, and that is one of the things I want to try as well.  Being able to map-switch (or subtract or add based on conditions, as suggested by elRey) will allow me greater flexibility.  Plus, extending my programming experience to assembly is a good thing.  I've learned tons already.


Yes, that would work.  I will definitely consider that approach.

 

Yes, exactly, similar in concept to map switching based on cruise switch.

How about this
If acc pedal > 90 then
     if actual load < 110(say) then alternate KFZW
what do you think
 

^^^^ Exactly!

This is not how it works, plain and simple. If you want antilag off the line, then allow brake boosting and hold down the brake and gas at the same time, thus allowing your engine to build boost...
Your timing is probably way out, acceleration times will *not* improve from doing what you are doing. You are basically saying that everyone in the automotive world is stupid and you just came up with some mega antilag solution - this is a pretty bold statement.

The problem is, if he logs RPM vs boost, he may see what looks like a faster spool... but if he logs time vs boost...

BTW ecuxplot has both pressure per sec and pressure per rpm in the calc section..

Exactly. He claims that his acceleration improves, but I am quite sure no real world tests have been made, as acceleration will be the best at MBT.

The reason I don't like wped is because it is driver's wish at the beginning of the torque model. Alot can happen before the torque request is realized.

Forget rlsol though, it makes no sense. I would use milsol. This is the torque request for air charge.

I have to agree with nyet and prj on this though.

What exactly are you looking for? Someone to write the function?

post some logs. I'm sure there's a timestamp column, then we can all see what the OP is feeling and argue about the same data.

I'm not saying that the entire automotive world is stupid, and I'm not saying that *I* came up with anything.  That's a leap you made, not me, prj.  I'm far too humble a person to ever go down that road.  This is a hobby for me, while a vocation for others, and I understand the difference.  I asked for some help and gave some reasons why I wanted to do what I wanted to do, which is generally how such things are done.  I have indeed done 'real world tests.'  Quite a few, actually.  I see an effect that appears to hold with respect to *my* specific setup, under *my* specific circumstances, and I want to do some more experimentation to see if I can improve on it.  Nothing more and nothing less.  As I said, I haven't found the end-point yet, and *adding* timing may prove to be the most beneficial of all.  I'm not counting that out.

I looked at it both ways, boost vs. RPM and boost vs. time.  Things are not always what they first appear to be; that's a basic tenet of scientific approach.  I'll post logs -- I'm on a different computer at the moment and don't have them with me.

I understand true anti-lag.  My wife isn't going to be doing that when she drives, and I don't have much interest in it, either.  At least, not as a general practice.

Matchew was a big help, as was printing the entire disassembly to aid in searching.  I won't say I definitely don't need any help at all any longer, but I'm a lot farther than I was, at least.

Boost vs time and boost vs rpm is not important, acceleration is.

I totally agree with prj.

reducing timing gives you better spool, but cuts torque and efficiency a lot more.

What about your afr? a leaner mixture in low rpms should give some better spool without killing torque and efficiency. I dont know ow far you enrich at what pedl position at the moment. Might this solve the probs a bit?