Garfimp
Full Member
Karma: +1/-1
 Offline
Posts: 50
|
 |
« Reply #60 on: April 18, 2019, 05:30:56 AM »
|
|
|
You seem to be name dropping random maps. KFMIRL is a load request map, capped by LDRXN. KFLDRL is the final output from the PID to the N75. If you increase boost using KFLDRL, you'll likely trip load and boost limiters. You need to understand the relationship between load (cylinder filling) and the ECUs requested boost. i.e why does the ECU request x mbar pressure for y% load. There's no point randomly modifying load maps and boost control maps without knowing what happens in between. The FR is your friend.
Hi gman86 and thank you for the help ! I don't know indeed how exactly work KFLDRL and KFPLGUB maps you are right, but it wasn't random naming, just someone from another forum who said that I should have a look at them so I was asking your opinion before digging into those  For KFLDRL thanks to you and confirmation on the wiki, I understand now that it is the linearization of the PID so I should not modify it to adjust requested boost. Concerning KFPLGUB, I had a look at it and it seems it is a linearisation of boost request vs MAP sensor and rpm ? but really not sure.. and maybe it shouldn't be touched ? Concerning KFMIRL I know it is capped by LDRXN, what I meant is that maybe I didn't scale it properly and the output from my modified KFMIRL doesn't request enough load to hit the LDRXN load cap I have set. So I think I should try to raise KFMIRL values. I am correct ? I will go again to look for information on the FR yes, but I find the FR really "unfriendly"...  |
|
|
|
« Last Edit: April 18, 2019, 07:21:43 AM by Garfimp »
|
Logged
|
|
|
|
Garfimp
Full Member
Karma: +1/-1
Offline
Posts: 50
|
|
« Reply #61 on: April 18, 2019, 06:17:11 AM »
|
|
|
Can you please give me your opinion on my tune of KFMIRL/KFMIOP/LDRXN ? I just checked it to see what I did wrong but found nothing wrong ? and finally I think it was OK, req load shouldn't be capped before LDRXN.. so if nothing is wrong here, I will increase LDRXN in order to have more requested boost at high rpm as said before by Nyet. KFMIRL :  KFMIOP :  LDRXN :  Thanks |
|
|
|
« Last Edit: April 19, 2019, 01:48:57 AM by Garfimp »
|
Logged
|
|
|
|
Bitshifter
Full Member
Karma: +10/-7
Offline
Posts: 94
|
|
« Reply #62 on: April 18, 2019, 11:41:09 AM »
|
|
|
Remember: KFMIRL is inverse KFMIOP  |
|
|
|
|
Logged
|
|
|
|
Garfimp
Full Member
Karma: +1/-1
Offline
Posts: 50
|
|
« Reply #63 on: April 18, 2019, 11:42:51 AM »
|
|
|
Remember: KFMIRL is inverse KFMIOP Yup, that's the case, unless I'm missing something ? |
|
|
|
|
Logged
|
|
|
|
Bitshifter
Full Member
Karma: +10/-7
Offline
Posts: 94
|
|
« Reply #64 on: April 19, 2019, 01:37:50 AM »
|
|
|
Yup, that's the case, unless I'm missing something ?
@MIRL: I don`t like...looks like the cracked OEM shit. @MIOP: the same But it is my opinion, maybe other users say it is OK. And now ask yourself: is it plausible to change map-values (Z) and leave the Y-values untouched for this case? |
|
|
|
|
Logged
|
|
|
|
Garfimp
Full Member
Karma: +1/-1
Offline
Posts: 50
|
|
« Reply #65 on: April 19, 2019, 01:46:04 AM »
|
|
|
@MIRL: I don`t like...looks like the cracked OEM shit. @MIOP: the same But it is my opinion, maybe other users say it is OK. Ok that's interesting point! So, to explain, what I did is extrapolate but keeping the way it was done originally and then making IOP/IRL matching. But maybe it is wrong? Do you have an example of how you do it? And now ask yourself: is it plausible to change map-values (Z) and leave the Y-values untouched for this case?
Which map are you talking about? I modified/extended IOP load axis already so I don't understand? Thanks |
|
|
|
|
Logged
|
|
|
|
Bitshifter
Full Member
Karma: +10/-7
Offline
Posts: 94
|
|
« Reply #66 on: April 19, 2019, 02:30:49 AM »
|
|
|
Ok that's interesting point! So, to explain, what I did is extrapolate but keeping the way it was done originally and then making IOP/IRL matching. But maybe it is wrong? I see what you have done. Wrong or right....everybody have another opinion. What say the experienced users? Do you have an example of how you do it? linearize! Which map are you talking about? MIRL I modified/extended IOP load axis already so I don't understand? right. Thanks |
|
|
|
|
Logged
|
|
|
|
Garfimp
Full Member
Karma: +1/-1
Offline
Posts: 50
|
|
« Reply #67 on: April 24, 2019, 04:35:26 AM »
|
|
|
I digged around again, and may have found a lead. I think that I misunderstood the sentence "IOP has to be the inverse of IRL" and some wrong information is shared around on this point. If I understood correctly now, IOP is the inverse function of IRL, but it doesn't mean that every values in those maps have to be the exact inverse. So what I can try to do is to alter IRL in order to have enough load requested to match with the cap from LDRXN (this is already the case I think, but will check), but then in IOP, I think I should leave stock values or even lower the values and maybe also extend axis, so that if I have a little more actual load than requested load, then the IOP table translates actual load into an actual torque value which is under requested torque value, and so I would not have any torque intervention, making drop in requested load etc. In // I will work to regulate my boost so that actual boost is not too much higher than requested boost, mainly at spike around 3500 rpm. Can someone confirm if this is the good way to tune IRL/IOP? Thanks |
|
|
|
« Last Edit: April 24, 2019, 04:40:24 AM by Garfimp »
|
Logged
|
|
|
|
|
prj
|
|
« Reply #68 on: April 24, 2019, 12:40:01 PM »
|
|
|
There is nothing to tune. It's mathematical inverse. All you need to modify is last column in both.
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #69 on: April 24, 2019, 02:34:50 PM »
|
|
|
Can someone confirm if this is the good way to tune IRL/IOP?
I'm with prj on this. Just keep it stock where you didn't modify IRL (usually low load/rpm areas) taking axis changes into account. I.e. just move a few rows in their entirety to make room for the rest. Where you did modify irl, make it a straight inverse. If you do change the axis, make sure you fix kfzwop/2 accordingly, since they share axis. |
|
|
|
|
Logged
|
ME7.1 tuning guideECUx PlotME7Sum checksumTrim heatmap toolPlease 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 ex
|
|
|
Garfimp
Full Member
Karma: +1/-1
Offline
Posts: 50
|
|
« Reply #70 on: April 25, 2019, 02:18:07 AM »
|
|
|
There is nothing to tune. It's mathematical inverse. All you need to modify is last column in both.
I'm with prj on this. Just keep it stock where you didn't modify IRL (usually low load/rpm areas) taking axis changes into account. I.e. just move a few rows in their entirety to make room for the rest.
Where you did modify irl, make it a straight inverse.
If you do change the axis, make sure you fix kfzwop/2 accordingly, since they share axis.
Ok thanks to both of you, so I was wrong again Then if it is mathematical inverse my values should be already good in my last tune, can you please confirm ? Except that I did it for the 2 last rows :   Concerning KFWZOP 1 and 2, thanks Nyet, I didn't look at that so far, I will try to take axis modification into account and post the modification I intend to do here. |
|
|
|
« Last Edit: April 25, 2019, 02:20:38 AM by Garfimp »
|
Logged
|
|
|
|
Beaviz
Full Member
Karma: +8/-4
Offline
Posts: 190
|
|
« Reply #71 on: April 29, 2019, 01:57:16 AM »
|
|
|
If you do change the axis, make sure you fix kfzwop/2 accordingly, since they share axis.
It is not shared on MED9. ;-) |
|
|
|
|
Logged
|
|
|
|
|
gman86
|
|
« Reply #72 on: May 01, 2019, 03:58:29 PM »
|
|
|
It is not shared on MED9. ;-)
Yes it is. |
|
|
|
|
Logged
|
|
|
|
slowstarter
Newbie
Karma: +0/-0
Offline
Posts: 1
|
|
« Reply #73 on: August 27, 2020, 12:09:09 AM »
|
|
|
Yes it is.
Is it true? Based on the Med9.1 damos that i have on hands, it seems that KFMIOP is not sharing any axis with other maps. |
|
|
|
|
Logged
|
|
|
|
Beaviz
Full Member
Karma: +8/-4
Offline
Posts: 190
|
|
« Reply #74 on: September 01, 2020, 06:12:55 AM »
|
|
|
Is it true?
Based on the Med9.1 damos that i have on hands, it seems that KFMIOP is not sharing any axis with other maps.
I have not seen it in any of the fully defined files floating around. I have not looked at the code itself if that axis is shared or not. It is shared in e.g. ME7, but to my knowledge not MED9 |
|
|
|
|
Logged
|
|
|
|
|
