NefMoto

I was curious why you can't just adjust KFLDIMX over increasing LDRXN and KFMIRL for increasing boost via load.  Is it that request boost will be way off from actual boost?  Just curious because I have franken and they make boost very easy, seems i could reach true stage 3 levels easily off just KFLDIMX..

Are you trolling?

If not, please read about how boost control works in this ECU. KFLDIMX is I limit for the PID, and it should always contain the WGDC to reach the boost defined in the column.

I'd read up on PIDs first, THEN read through the FR section describing the boost PID :)

IMO it is impossible to tune proper boost control on ME7 w/o a decent understanding of PIDs.

I've been messing around with the PID for a while and I can change the amount of boost my car is producing just by tweaking KFLDIMX that is why I asked.  My turbos easily make 17-18 psi at with load set at 172 peak and 1000 mbar column set around 52.. seems like alot of head room just tweaking the waste gate duty cycle.. Sorry not trolling just asking questions to get a better understanding.  I get that you are setting how much duty cycle is needed to make 1000 mbar, so I guess it is stupid to fool the computer into making more boost then it thinks it's making.

That's just not how it works.  Using DIMX with the intention of altering boost basically means that you're not in control.  Use the proper map to set desired boost (LDRXN, capped as appropriate by KFLDHBN etc.) and then tune the PID (including DIMX as necessary) so actual matches desired.  DIMX is used to tune behavior, not to drive it as a primary element.  From your comments it does seem like maybe the change you made to DIMX was coincidentally a necessary one, but it's really a very good thing to understand why.

Who set the 1,000 MBar point to <95 (or whatever it is stock) to begin with?

Really, stop doing this experimentally; you REALLY should learn how PIDs work before proceeding.

I do have a basic understanding, I'm not changing values radically I'm making slight adjustments to my tune over the past year and I feel I have the car dialed in pretty well.  I mean you can only learn so much from reading the 3 sentence descriptions in the wiki, the many threads in the forums and comparing different files have given me some understanding.  Your base file for the china turbos has been a great start for my PID nyet.

That's not the only information on PIDs. You know PIDs are not specific to ME7, correct?

man, you gotta be real careful...  seriously, this is a sophisticated concept and its very precise.  nobody here is trying to talk down to you, but you really gotta read and understand this:

http://en.wikipedia.org/wiki/PID_controller

nobody wants anyone to break shit...  we're all on the same page here.

I'm not trying to break anything, if I would of left my car on the stage 1 giac tune that it had when I first installed my frankens I would of broken something already considering the wgdc's  are set way to high and the thing would just overboost all the time.  I have it controlled at a peak of 16 psi on stock injectors, and conservative timing, I'm really happy with what knowledge I have gained and understanding I have gained from this forum and the wiki.  I understand it was a stupid question and I have a lot to learn but I do have a general idea for what most the maps in the PID do.  The reason I asked was it seems load numbers, are just numbers, so my logic was instead of having to change all the load axis's, rescale kfmirl etc it would be easier to just leave them as is and increase the amount of boost generated at the peak load.

Here's another nub question, why does the KFLDIMX map only go up to 1000 mbar, do we not have control of the duty cycles beyond this?

I believe that axis is lde...

The patent is really good at explaining everything.  http://www.google.com/patents/US6918250?printsec=claims#v=onepage&q&f=false

Thanks I'll start reading, just ordered EV14's so the more knowledge I can get the better before those get here :)  I'm sure I'll start another annoying thread with logs and trying to figure out my new tune once installed.  Thanks for the help as usual everyone.  Oh and nyet sorry what does lde mean? Are you referring to a table to change that axis?

the input to that map's axis is lde (ld error)

So is it wise to increase the axis to support the higher levels over 1 bar of boost or usually unnecessary?

It's error (I think that's correct).  So 1,000 mbar of error, not 1,000 mbar boost.  So no, no need to change the axis.

Please, read up on PIDs :(

All of those complicated maps are a part of ME7's gain scheduling of the PID

http://en.wikipedia.org/wiki/Gain_scheduling

http://en.wikipedia.org/wiki/PID_controller

... since single values of P I and D aren't really sufficient..

Which map?



How does it only go up to 1000mbar?

Its the error of the Measured Target....  So if you wanted 23psi but you were getting 18psi that'd be a 5-psi error.  1000 mbar is 14.7 psi ERROR.

Maybe ME7.1 is different, but I doubt it.
At least on ME7.5 and MED9 the pressure input to KFLDIMX is relative target pressure (PLSOLR_W)

Why would it be "boost pressure error"?

EDIT:
Ok, quickly looked at a ME7.1.1 .dam File.
Axis is boost pressure, unit hPa, 0-2559.9609375
Not pressure error.

I could be wrong about KFLDIMX, but the rest of the PID (and gain schedule) is driven by lde.

I'd have to look at the FR again.

If you are right, the 5120 hack would be one solution.

I have looked at the ME7.5 and MED9 Funktionsrahmen.
Trust me, it is plsol_r.

Yes. You are right, I was wrong. It is plsol_r in ME7.1 as well.

Which is plsol (requested boost) - plgrus (basically ambient pressure). I don't know what CWPLGU.0 is in ME7.1 but I assume it is 1 since it has an ambient pressure sensor?

I have no idea how BGPLGU works so hopefully others can chime in here.

plgrus_w is not as described in the FR.

plgrus_w is plgta output from KFVPLGU multiplied by pssol_w, then plsolr_w is plsol_w minus plgrus_w. This is all done in LDRPID.

phila: have you ever logged plgrus? How does it differ from pu_w? Is CWPLGU.0 1 or 0?

They are nearly identical.

CWPLGU is not considered.

:/

Yea...you can disregard that completely.

This is one of the few examples where the FR isn't even close in our application.

I don't get this. KFVPLGU seems to be all ones.

This would make plgrus_w basically the modeled pressure drop across the throttle plate (difference between plsol and pssol), not pu. (plsol = pssol / vpsspls)

What am i missing?

So it is as prj says?  Determines how much duty cycle to make 1000 mbar?  Or it is it boost error?  Confused now...

It's how I said.


Have not checked on 2.7TT, but good luck with that approach on 1.8T.  ::)

Fixed. I wrote it wrong, pssol_w is the axis variable for KFVPLGU.  The output is multied by pu_w.

Good luck with what approach?

I'm just telling it like it is.

Look at page 865.
Normally CWPLGU should be set to 1.

If you set it to 0, plsolr_w will change a huge amount, at least on 1.8T.
And some 1.8T variants have this not set, and KFLDIMX calibrated completely differently, and "grundladedruck" or "base pressure" has a completely different meaning there.

It is definitely used, just not on 2.7TT it seems.

I don't need to look at any page, it's not used. My comments were not in general, they were straight from the M-box code.  ::)

Edit: I guess I could have been more clear that I was speaking about a specific application.

Yes, that is my point. This thread is pretty general, so it would be good to clarify that you are speaking specifically about the 2.7TT application where it is indeed not used.
In fact the DAMOS does not even have this constant defined. I checked ME7.1.1 for 2.7TT too, and it's not in there either. Not in the DAMOS nor in the code.

I was not trying to take a dig at you, sorry if it came out like that.

Which, if KFVPLGU is all ones, is functionally identical to CWPLGU=1 (plgrus = pu)?

Yes, that is how it appears to be.

Does information from the 2005 patent come into play, which if I'm reading correctly, and ignoring misc. corrections, says that plsolr is subtracted from plsol, making the pressure axis values for KFLDIMX relative to the charge available at 0 WG duty, i.e. WG cracking pressure?

That would make sense, but in the 2.7t, clearly not the case, where the pressure axis is plsol-pu (or relative requested pressure over ambient).

I checked the log and plgrus_w == pu_w and manually confirmed that plsolr_w = plsol_w - plgrus_w.

This obviously should be the case, but I just wanted to confirm it and correct my previous statement.

I agree...I was just being grumpy.

Exactly. This is how it works in some other applications.