Pages: [1] 2 3 ... 7
Author Topic: Big turbo IRL/IOP/LDRPLS/KFWDKMSN tuning  (Read 67085 times)
nyet
Administrator
Hero Member
*****

Karma: +417/-48
Offline Offline

Posts: 9254


WWW
« on: July 30, 2013, 12:53:53 PM »

my KFMIRL and KFMIOP and they have been super happy campers.

Can you summarize your approach to tuning those maps?

Quote
The only time I know I've seen N249 problems was when I flashed someone a tune with the N249 coded out and he still had it lol, that acted like you describe. But a change in settings and he was good to go again.

Got any details? Smiley
« Last Edit: July 31, 2013, 05:34:14 PM by nyet » Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
britishturbo
Sr. Member
****

Karma: +12/-1
Offline Offline

Posts: 306


« Reply #1 on: July 30, 2013, 12:56:00 PM »

Can you summarize your approach to tuning those maps?

Got any details? Smiley

Sure.
When I get time tonight I'll post some screen shots of what I run for them.
I'm sure they will be "wrong" though haha, someone once said they were Impossible :-)
Logged
nyet
Administrator
Hero Member
*****

Karma: +417/-48
Offline Offline

Posts: 9254


WWW
« Reply #2 on: July 30, 2013, 01:00:28 PM »

Sure.
When I get time tonight I'll post some screen shots of what I run for them.
I'm sure they will be "wrong" though haha, someone once said they were Impossible :-)

just in case I wasn't clear: i'd rather understand your thought process than just see the end results Smiley
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
phila_dot
Hero Member
*****

Karma: +157/-11
Offline Offline

Posts: 1704


« Reply #3 on: July 30, 2013, 02:53:21 PM »

All of the details of how this works is already posted numerous times.

The rest is just logic and preference.

What kind of answer is everyone looking for?
Logged
nyet
Administrator
Hero Member
*****

Karma: +417/-48
Offline Offline

Posts: 9254


WWW
« Reply #4 on: July 30, 2013, 03:03:27 PM »

Just wondering if he took a different approach... and IMO the surge problem is evident in big turbo tunes with tight wastegates regardless of IOP/IRL tweaking.

There is another thread which suggests tuning the throttle plate angle maps... (don't remember the map names off hand and I dont have an FR handy for reference).
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
prj
Hero Member
*****

Karma: +281/-27
Offline Offline

Posts: 3456


« Reply #5 on: July 30, 2013, 03:12:26 PM »

The throttle angle maps need to be set to reflect base boost as silentbob said.
So that FUEREG is active when it should be.
Logged
phila_dot
Hero Member
*****

Karma: +157/-11
Offline Offline

Posts: 1704


« Reply #6 on: July 30, 2013, 03:26:25 PM »

The issues are unrelated to the torque model.

The torque model is almost a seperate entity. Fast path intervention is on/off and controls ignition angle with zwsol. Slow path outputs a torque request for conversion to desired load via IRL.

If you don't have desired load problems below rlmax_w, then look beyond the torque model.
Logged
Snow Trooper
Hero Member
*****

Karma: +62/-24
Offline Offline

Posts: 675


WWW
« Reply #7 on: July 31, 2013, 12:08:01 PM »

What issues do people seem to be having with IRL/IOP tuning on large frame turbos?  To generalize, these turbos exhibit two main characteristics.

First they lag compared to ko3/ko4 type turbos and or have slower on throttle boost response even in the spool zone.  So one could be safe to assume that you can really mellow out those ranges of these stock maps you are editing that are in full load on small frames.  Be realistic about your systems capability down low and don't ask for something it can't make.

Second, large frame turbos don't taper.  Some people taper them and that is just lame.  If you are going to do anything, taper that thing up.  I rise 5 full Psi on my pump files as the revs go up after the initial boost hit and spike so that the car just nevers stops pulling like a beast, holds good timing after peak tq, because that is what big turbos are supposed to feel like.  I see a lot of stuff that drops in boost for no reason.  Your turbo, if truely a large frame twin or single setup wants to do work on 2.7 liters and can handle it just fine.

Never neglect to have a LDRXN that makes sense too.  These are my thoughts on the matter, I think its really simple to have a huge turbo setup that still drives nice.  BTW, I have no n249, I have no issues because of this.

Logged
nyet
Administrator
Hero Member
*****

Karma: +417/-48
Offline Offline

Posts: 9254


WWW
« Reply #8 on: July 31, 2013, 12:13:44 PM »

What issues do people seem to be having with IRL/IOP tuning on large frame turbos?

I think the main point of contention seems to be surging during part throttle when the wg cracking pressure is well above req boost/load..
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
Snow Trooper
Hero Member
*****

Karma: +62/-24
Offline Offline

Posts: 675


WWW
« Reply #9 on: July 31, 2013, 12:22:10 PM »

I run a 21 Psi gate and a near zero wgdc in areas that make sense too.  But in theory I can never run less than 21 Psi unless I control the throttle.  So I control the throttle by having a nice and linear set of throttle maps.  I don't feel I have these issue nor do other cars I have tuned.  Can someone post their logs of the behavior?

I am familiar with the feeling, some ots tunes do it when you run cranked wastegate rods, obviously.  Logs would be telling on throttle angle and such, so would the bin.

Maybe people arent really editing their maps enough on the very different than stock hardware and are still just trying to sort of italian tune stuff like you can do when you go from ko3s to ko4s.
Logged
prj
Hero Member
*****

Karma: +281/-27
Offline Offline

Posts: 3456


« Reply #10 on: July 31, 2013, 02:41:19 PM »

I run a 21 Psi gate and a near zero wgdc in areas that make sense too.  But in theory I can never run less than 21 Psi unless I control the throttle.  So I control the throttle by having a nice and linear set of throttle maps.

You don't control the throttle, FUEREG does.
Logged
nyet
Administrator
Hero Member
*****

Karma: +417/-48
Offline Offline

Posts: 9254


WWW
« Reply #11 on: July 31, 2013, 02:45:06 PM »

Prj: any chance i can get you to discuss "base boost" vs pu?

my bad german is not serving me well right about now.
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
prj
Hero Member
*****

Karma: +281/-27
Offline Offline

Posts: 3456


« Reply #12 on: July 31, 2013, 02:52:35 PM »

First you need to find out whether plgrus_w is equal to pu_w or whether it's mapped from factory.
The RS4 K-box for example has plgrus_w equal to spring pressure, and then the entire boost control dances around that.

I find it a PITA to tune, so I hacked the binary to have plgrus_w = pu_w like it is done on the S4.

After that, you need to tune KFVPDKSD and KFVPDKSE to have values lower than PSPVDKUG (just a little lower is enough) in the areas that are below the spring boost of the turbos. Otherwise the ECU goes WOT when it shouldn't and you get massive throttle cut on part throttle and kangarooing.
Logged
prj
Hero Member
*****

Karma: +281/-27
Offline Offline

Posts: 3456


« Reply #13 on: July 31, 2013, 04:48:01 PM »

To elaborate why I said the driver does not control the throttle - it is because wped_w, mifa_w nor rlsol_w have much say in this.

rlsol is simply the input to the throttle setpoint algorithm FUEDK, which is based on KFWDKMSN.
After the initial setpoint, the throttle angle is controlled by FUEREG, which is closed loop feedback control.
FUEDK and FUEREG are heavily affected by LDRPLS - LDRPLS basically tells it whether it needs full throttle to reach target filling or not (req boost below spring boost or not).
Get this wrong and it overshoots it's throttle control all the time. Nothing to do with what is requested - the ECU is not able to deliver request correctly. Like it can not accurately deliver boost without recalibrating boost PID.

Even if rlsol_w is constant, the throttle will open up and then close down as acceleration of the engine is not instantaneous, and on a turbo engine you also have the inertia of the turbocharger(s).
KFMIOP, KFMIRL, KFPED and so on mainly affect how the engine reacts to "gas pedal" inputs, but have little effect on the movement of the throttle.

And that's exactly why I said the driver does not control the throttle. The driver only requests torque. The ECU then calculates throttle plate angle and other parameters to deliver said torque.
I would go in as far as to say that IOP, IRL and KFPED can be left mostly stock, if you want the car to feel stock (within limitations due to turbocharger spool characteristics) in the lower end of the scale. The main modification that needs to be done there is making sure that enough load is requested.

What needs to be recalibrated for the turbocharger are LDRPLS (for specifying base charge pressure) as well as KFWDKMSN (and it's inverse map), so that FUEDK has a rough idea of what it is doing before feedback control kicks in.
« Last Edit: July 31, 2013, 05:00:47 PM by prj » Logged
nyet
Administrator
Hero Member
*****

Karma: +417/-48
Offline Offline

Posts: 9254


WWW
« Reply #14 on: July 31, 2013, 05:42:04 PM »


FUEDK and FUEREG are heavily affected by LDRPLS - LDRPLS basically tells it whether it needs full throttle to reach target filling or not (req boost below spring boost or not).

Thank you. That is the most concise explanation of FUEDK/FUEREG i've seen so far.

Quote
I would go in as far as to say that IOP, IRL and KFPED can be left mostly stock, if you want the car to feel stock (within limitations due to turbocharger spool characteristics) in the lower end of the scale. The main modification that needs to be done there is making sure that enough load is requested.

Excellent. I've always felt this way intuitively. Glad to hear you confirm it.

Quote
What needs to be recalibrated for the turbocharger are LDRPLS (for specifying base charge pressure) as well as KFWDKMSN (and it's inverse map), so that FUEDK has a rough idea of what it is doing before feedback control kicks in.

Thanks. I have edited the thread topic accordingly. By inverse map, do you mean the cruise control map (KFWPFGR)?

Also, how do you feel about italian tuning KFWDKSMX, assuming IRL and LDRPLS are calibrated correctly?
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
Pages: [1] 2 3 ... 7
  Print  
 
Jump to:  

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Page created in 0.024 seconds with 17 queries. (Pretty URLs adds 0s, 0q)