Pages: 1 [2] 3 4 5
Author Topic: KFVPDKSD/E and WDKUGDN throttle tuning near wg cracking pressure  (Read 99036 times)
nyet
Administrator
Hero Member
*****

Karma: +607/-168
Offline Offline

Posts: 12268


WWW
« Reply #15 on: October 26, 2015, 07:45:00 PM »

Nice work man..

Upped WDKUGDN just a little to allow for the 12psi gates. I just took a guess here in the areas that the f21s will make more than the k03s.

I'm still confused as to what the WDKUGDN changes should represent... :/
Logged

ME7.1 tuning guide
ECUx Plot
ME7Sum checksum
Trim heatmap tool

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 ex
zillarob
Full Member
***

Karma: +13/-1
Offline Offline

Posts: 152


« Reply #16 on: October 26, 2015, 09:11:29 PM »

WDKUGDN is the throttle angle at a certain rpm where the throttle plate is no longer a restriction to air flowing into the intake mani.

I'm not sure if I understand it correctly, but this is why I did what I did.
Somebody please correct me if I am wrong, 'cause there is a good chance of that  Grin

I added a few degrees to open it up a little because 12psi gates are going to be pushing more airflow through there than the stock 6psi gates.

The throttle is only opening to WDKUGDN in areas on the KFVPDKS maps that are <.95.
When you cross over into an area that is >=.95, the throttle blade sorta changes strategies and opens up more.

If you leave WDKUGDN stock with gates set higher, there will be a bit of air backing up behind the throttle blade that the ecu is not expecting.
When you cross over from a .94 area to a .95 area, the throttle is going to open and this extra air is going rush in unexpectedly.
You may feel this while slowly rolling into the pedal. When it crosses over, you might feel 1 quick surge as that air rushes in.

edit to reflect fr numbers instead of 551m
« Last Edit: November 02, 2015, 05:30:30 PM by zillarob » Logged
Lost
Hero Member
*****

Karma: +21/-14
Offline Offline

Posts: 556


« Reply #17 on: October 26, 2015, 10:11:28 PM »

Does not exist in RS4, it's hardcoded.
It's already fixed in your basefile.

Ok. Thanx.
Base file i use is org RS. Nothing is fixed here.
But i think i got my head around this now.
Logged
SB_GLI
Hero Member
*****

Karma: +116/-10
Offline Offline

Posts: 1022


« Reply #18 on: October 27, 2015, 05:26:50 AM »

On my Jetta 1.8t F21, I've only tuned the KFVPDKS axis, and I have been very happy with those changes as it were.  After reading your description on what WDKUGDN does, I will go in and play with that a bit too.
Logged
MIL_on
Full Member
***

Karma: +12/-2
Offline Offline

Posts: 119


« Reply #19 on: October 27, 2015, 09:37:58 AM »

My understanding (which cant be that wrong regarding my calcs):
The term "ungedrosselt" (="unthrottled") refers to the point where the opening area of the Throttleblade does not limit the gas-exchange of the engine anymore. But from a pure "mechanical" point of view ( = 100% rl) and has nothing to do with the boost you are aiming at.

The calculation for a 1.8T Engine for example:
Engine Displacement = 1781cc
nCyl = Number of Cyl = 4
Vcyl = Volume per Cyl: 445cc
rhoL = air density: 1.2 kg/m³   = 0.0000012 kg/cm³


using the equation:
mAir[kg/h] = Vcyl[cc]*rpm[1/min]*rhoL[kg/cc]*60*nCyl delivers this theoretical 100%rl. All you have to do now is multiply it by 0.95 taking PSPVDKUG into account.



then you have to calculate the corresponding throttle value to these mass air flows using KFWDKMSN. I also used excel for this.
The calculated Values vs. Values out of the original File of WDKUGDN:



As you can see the fault is around 1% between 2000 and 5000 rpm. The error @ 1000 rpm is 3 degree Throttle Value and @ 6000 rpm its 2 degree which seems to be ok for my definition....
« Last Edit: October 27, 2015, 09:42:02 AM by MIL_on » Logged
nyet
Administrator
Hero Member
*****

Karma: +607/-168
Offline Offline

Posts: 12268


WWW
« Reply #20 on: October 27, 2015, 09:40:54 AM »

But why use ambient for air density? Shouldn't the density be at wastegate cracking pressure?
Logged

ME7.1 tuning guide
ECUx Plot
ME7Sum checksum
Trim heatmap tool

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 ex
zillarob
Full Member
***

Karma: +13/-1
Offline Offline

Posts: 152


« Reply #21 on: October 30, 2015, 07:48:25 PM »

I would hate to see the math needed to calculate that throttle angle  Shocked

Im pretty sure I wont understand it, but might be interesting to see if you post it up.
Logged
nyet
Administrator
Hero Member
*****

Karma: +607/-168
Offline Offline

Posts: 12268


WWW
« Reply #22 on: October 31, 2015, 08:34:48 AM »

The throttle is only opening to WDKUGDN in areas on the KFVPDKS maps that are <.97.
When you cross over into an area that is >=.97, the throttle blade sorta changes strategies and opens up more.

I don't think that's right. PSPVKDUG = .95 so we're talking about a range of (.96-.95)/(1-.95) through 1, or 0.2  through 1.0, so i think it is more properly stated that it is proportional to where the map output lies between 0.95 and 1.0
Logged

ME7.1 tuning guide
ECUx Plot
ME7Sum checksum
Trim heatmap tool

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 ex
zillarob
Full Member
***

Karma: +13/-1
Offline Offline

Posts: 152


« Reply #23 on: October 31, 2015, 11:00:20 AM »

I did see that proportional part in there, but hasnt clicked for me yet   Huh

Is your description based solely on whats in the fr and 1.8 numbers?
Or are you trying to apply the fr to the (different) 2.7 numbers?
« Last Edit: October 31, 2015, 07:43:21 PM by zillarob » Logged
zillarob
Full Member
***

Karma: +13/-1
Offline Offline

Posts: 152


« Reply #24 on: October 31, 2015, 09:00:14 PM »

OK, I think your description was enough of a kick in the butt to jar some brain cells loose and get this to click for me  Grin

I'll use the values straight from the fr here to keep things consistent.
Will also make some of the other numbers up to keep the math easy.



Say we are at a pedal pos that req's a pr that puts us on a cell in KFVPDKS that is .95, and at 60% throttle (WDKUGDN at that rpm?).

The .95 in KFVPDKS being <= to PSPVDKUG(.95) means that the turbos can make the req pr on wg psi without any help from the n75.

I think the throttle will be at WDKUGDN here with no reason to open it anymore because it is not a restriction at this angle. It is allowing it to give everything it can on wg psi. (.95 is actually right where the throttle changes strategies, but the formula for the new strategy cancels itself out at this point)



Now, say we give a little more pedal, move to a higher req pr, and land on a .98 cell in KFVPDKS.

To make this new req pr (that is now above wg psi), the ecu needs to start feeding wgdc.

The throttle also needs to open a little more to stay out of the way of the increased amount of incoming air. That formula is how it does this smoothly on its way to 100%.
This is how it works, or at least my best guess anyway  Grin
We were at 60% throttle. That leaves 40% remaining (on its way to 100%) as the Residual Throttle Value.
With the formula (.98-.95)/(1-.95), we get .6 (or 60%).
60% of our Residual Throttle Value (40%), is 24%.
Now we add that 24% to our original throttle value of 60% (WDKUGDN?), for a new target throttle of 84%

« Last Edit: October 31, 2015, 10:50:19 PM by zillarob » Logged
nyet
Administrator
Hero Member
*****

Karma: +607/-168
Offline Offline

Posts: 12268


WWW
« Reply #25 on: November 01, 2015, 04:16:37 PM »

From what I can tell from logging, for higher PRs, a reduced WDKUGDN might be needed; it doesn't seem that at higher boost, the throttle is doing enough to prevent boost flutter when req load is near wg cracking. Can anybody else confirm this? With the WGDC at 10%, and stock WDKUGDN, boost is still building with vpsspls_w at minimum (.95) and throttle plate at 40%

Anybody else see this?

Also, the active range of vpsspls_w is definitely .95 to 1.0, where 1.0 is entirely wg controlled ingested air and .95 is entirely throttle plate controlled ingested air.
« Last Edit: November 01, 2015, 04:18:54 PM by nyet » Logged

ME7.1 tuning guide
ECUx Plot
ME7Sum checksum
Trim heatmap tool

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 ex
zillarob
Full Member
***

Karma: +13/-1
Offline Offline

Posts: 152


« Reply #26 on: November 01, 2015, 06:10:22 PM »

Sweet, this cleans it up a bit and should make it easier to find in a search.

If you still havent noticed, the 2.7 is .97, not .95 as in the fr.
Prob a good idea to use the fr numbers when we can to keep things consistent with the document, but something we need to keep in mind.

What I am trying to make go away is chirping/hooting at ~5psi when you are at a constant load that will hold the motor at a fairly steady rpm. Like going up a hill or something.

Playing with this stuff def seems to make the transition into boost much smoother and more controllable. I think I have it close enough that the throttle plate is steady and not what is causing it. The car drives and responds to pedal input almost as smoothly as stock anyway.

I am currently leaning toward what snowtrooper said about not asking more of the turbos than they can reasonably supply. We may just be asking too much of the turbos down low and pushing them into surge. They rip through just fine on a hammer down pull, but seems like they are just not able to sustain that amount of load in a steady state. But, as with pretty much everything I post here, Im not really sure if that is how it actually works  Grin
Seems like its leaving a little on the table, but will pull ldrxn down a little and see if anything changes.

Here is a old pic.
I dont remember much about it, but the throttle blade was steady at first, and joined the party where you see it change toward the right side.

Logged
zillarob
Full Member
***

Karma: +13/-1
Offline Offline

Posts: 152


« Reply #27 on: November 01, 2015, 07:50:19 PM »

Here is an interesting one.
I spotted the file name on that graph so I hunted it down and took another look at it.

Looks like boost reg goes active, but seems like something is holding the throttle down?
Would think that the throttle blade should be heading toward 100% in a hurry once boost reg comes online.



I should be seeing this car in the not so distant future so will have to try the things I have done to mine to see what it feels like.


PS: Hey Nye, any way to make ecuxplot hold the black background? Easier for my old eyes to see the lighter color lines  Grin


« Last Edit: November 01, 2015, 07:51:52 PM by zillarob » Logged
ddillenger
Hero Member
*****

Karma: +641/-21
Offline Offline

Posts: 5640


« Reply #28 on: November 01, 2015, 08:44:55 PM »

If nyet makes ecuxplot look like that I will kill someone.

Logged

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!

Email/Google chat:
DDillenger84(at)gmail(dot)com

Email>PM
zillarob
Full Member
***

Karma: +13/-1
Offline Offline

Posts: 152


« Reply #29 on: November 01, 2015, 09:12:23 PM »

Damn, you know where I live   Shocked

Not change it, just wondering if there is a cfg or some other way to save it.
Usually only looking at a few squiggly lines at a time, but the lighter color lines are tough for me to see against the gray background.

« Last Edit: November 01, 2015, 09:15:32 PM by zillarob » Logged
Pages: 1 [2] 3 4 5
  Print  
 
Jump to:  

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Page created in 0.028 seconds with 17 queries. (Pretty URLs adds 0s, 0q)