Pages: 1 ... 4 5 [6] 7 8 ... 21
Author Topic: Bosch EV14 Injector Migration  (Read 361482 times)
s5fourdoor
Hero Member
*****

Karma: +33/-3
Offline Offline

Posts: 617


« Reply #75 on: August 09, 2012, 04:57:59 PM »

stock airbox, stock maf, stock intake setup, AWE intercoolers, RS4 K04 up-pipes, BW K04's, 3" ASP DP setup into a milltek single, stock fpr, rs4 fuel pump

currently testing the following:

krkte:
  .0535

tvub:
  2.23
  1.49
  1.10
  0.80
  0.69


fkkvs:  (transpose and repeat for all rows in the table)  [in other words:  all RPMs get the same PW non-linearity treatment.]
  1.26
  1.18
  1.13
  1.08
  1.06
  1.05
  1.04
  1.03
  1.02
  1.02
  1.02
  1.01
  1.01
  1.01
  1.00
  1.00

will report back, but i believe this is my final setup calibrated to the stock airbox.
so serious and have yet to get the gas stink.  stft's and ltft's don't show much seek-and-find either...
will confirm this, but i'm convinced i've precisely cracked the code on these injectors.
« Last Edit: August 09, 2012, 05:18:00 PM by nehalem » Logged
s5fourdoor
Hero Member
*****

Karma: +33/-3
Offline Offline

Posts: 617


« Reply #76 on: August 09, 2012, 05:04:27 PM »

additionally.  for the mathematically inclined out there.  if the fkkvs i'm conjecturing is not correct, i believe the correct one to be in the affine hull of this monotonic plane.
if you didn't understand what i just said:  if you for any reason find the FKKVS is wrong, i believe it's at most wrong only in the modulo sense.

let's take the estimator FKKVS_hat:= (FKKVS-1) * (1.00+K) + 1.00, there exists a K such that minsup(FKKVS_hat,FKKVS) can be labelled optimal.

again if you didn't get what i'm saying.        subtract 1 and look at the positive values as values which need to be multipled by a scaling fraction to arrive at the optimal FKKVS_hat.
adjust K accordingly.    seriously i'm out of ways to explain myself now... lmfao!!!

lemme try one more description:  1.00 is no adjustment.  so if you find this to be wrong, just uniformly multiply the residual (correction factor above 1.00) by some percentage, either up or down (most likely down)
« Last Edit: August 09, 2012, 05:08:46 PM by nehalem » Logged
s5fourdoor
Hero Member
*****

Karma: +33/-3
Offline Offline

Posts: 617


« Reply #77 on: August 25, 2012, 11:27:56 AM »

Fact?  Audi updates their FKKVS Table Definition in the RS4 K/Q-box.
It appears they solved the exact issue Nye / others mentioned:  Large Injector, Low-Latency pulse non-linearity.

They shift the Pulse-width axis to have lower ms table values.  They also use the table, not just setting it to Constant 1.00 in both the K-box and Q-box.  Perhaps I'll check the F-box later as well just to see if it was an overall RS4-issue, or just from the K/Q which appear to be M-box derived.  Seeing the same values in the F-box would imply they knew bigger injectors required active / massive compensation.  Some young engineer at Cosworth Racing - probably named David Brown or [Aston Martin] - solved the issue.  Fictional stories aside:

From [M-box]:  {1.50 2.25 3.00 4.50 6.00 7.50 9.00 10.5 12.0 13.5 15.0 16.5 18.0 19.5 21.0 24.0}
From [Q-box]:  {0.85 0.95 0.98 1.00 1.50 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.0 11.0 12.0}

More to follow?  Stay glued to your seats dude.  [pours a white russian]

The Q-box's table from WinOLS below:
Logged
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Offline Offline

Posts: 12232


WWW
« Reply #78 on: August 25, 2012, 11:46:53 AM »

My question is, given an FKKVS of 1, how does a constant (for a given bus voltage) but non-zero TVUB affect this:



A fixed offset to ontime does WHAT to that error?

I'm not smart enough to do the math, but i bet you are 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.
s5fourdoor
Hero Member
*****

Karma: +33/-3
Offline Offline

Posts: 617


« Reply #79 on: August 25, 2012, 01:09:10 PM »

Great questions.  I'm not totally sure on the answers tbh, but here's what I do know.  The pulsewidth seen on that graph must include the VBAT offset already.  That's why it's starting at 0 until like .75 ms.  Therefore, assuming vbat is right, FKKVS simply corrects the pulsewidth modification as a whole.  If vbat-operating-voltage offset is say .80 ms, then the pulsewidths would start from there.  so if the car needed 1.5ms calculated, that's only .70ms of which is actual fuel, .80ms of which is recharge to basically sputter (idle is like a dribbling straw).  For this reason, at lower pw's regardless of rpm, we have to correct the expected fuel portion of the PW to account for the fuel "reload friction" by increasing the requested PW.  This isn't the case on a smaller set of injectors such as with the stock s4.  Also--- the S4 has settings on the KFLF map which may be accounting for this too.  Recall PRJ's hack which multiplies kflf * kfkhfm.

I'm not certain yet on whether to change the airbox map kfkhfm, kflf, or the combination but the attached spreadsheet show's the FKKVS axis change and the KFLF 1.00'd out.
My idle is fantastic now.  I'm going to try kflf and kfkhfm from the rs4 next.  Logs are looking good, i'll try to clean them up and post.

Attached are my current settings of things I believe to be relevant for changing to these injectors.  Hoping to bring this thread to some form of closure.
My current updated XDF is also attached.


« Last Edit: August 25, 2012, 01:12:04 PM by nehalem » Logged
s5fourdoor
Hero Member
*****

Karma: +33/-3
Offline Offline

Posts: 617


« Reply #80 on: August 26, 2012, 04:45:04 PM »

My question is, given an FKKVS of 1, how does a constant (for a given bus voltage) but non-zero TVUB affect this:



A fixed offset to ontime does WHAT to that error?

I'm not smart enough to do the math, but i bet you are Smiley


Here's my concern.  Where can this exact graph for varying voltages obtained for our actual 52# ev14 injectors which are 628cc @ 4-bar.
Should I send one to ID and get them to benchmark it?  What does that cost?
If not, then we are sort of guessing based on the specs of that 1000cc model, perhaps assuming a sqrt(4/3) to convert that graph to 4-bar.
I could try digitizing that graph and making some basic modelling assumptions.  I'm happy with the stock RS4 axis/values so far.
Logged
ironital
Jr. Member
**

Karma: +1/-3
Offline Offline

Posts: 39

Love what you do tell you do what you want


« Reply #81 on: August 26, 2012, 06:36:01 PM »

I'm going to be buying ID 1000cc injectors... I wonder how well ME7 will cope  Grin
am waiting to know your news because i have the same me7
I want to know how to tunne it my self
Logged

Mohamed fouad
jibberjive
Hero Member
*****

Karma: +23/-2
Offline Offline

Posts: 536


« Reply #82 on: August 26, 2012, 06:49:45 PM »

There are probably quite a few of us with 1000cc EV14's on here, just fyi.  You just tune them like any other injector.
Logged
Bische
Sr. Member
****

Karma: +25/-4
Offline Offline

Posts: 396



WWW
« Reply #83 on: August 26, 2012, 07:21:05 PM »

This is a great thread, im also on 1000cc EV14's currently dialing my fueling. I have been having some trouble getting uber rich on hard transitions, down in low .7x's at some scenarios.

Im on returnless and I believe I have tracked down my issue to lay in the FRFLSDP -fuel rail pressue compensation, in the stock form this almost doubled my KRKTE over 300mBar boost/fuel rail pressure drop. Just wanted to put this here if there are more people attempting to dial in returnless fuel systems.

I have recalculated the whole map/axis now to suit up to 1.5bar boost, and I hope this will get me in the right direction.
Logged
Bische
Sr. Member
****

Karma: +25/-4
Offline Offline

Posts: 396



WWW
« Reply #84 on: August 27, 2012, 12:06:53 PM »

I also want to add one thing regarding the FRFLSDP map, the factor shows up in Winols like 0.000031, the factor is really 0.00003052 but rounds off as it is entered/copied.

This is obviously going to make a big difference, I found this out the hard way and wanted to put that here for future searches regarding FRLFSDP.

Logged
jibberjive
Hero Member
*****

Karma: +23/-2
Offline Offline

Posts: 536


« Reply #85 on: August 27, 2012, 06:15:12 PM »

From [M-box]:  {1.50 2.25 3.00 4.50 6.00 7.50 9.00 10.5 12.0 13.5 15.0 16.5 18.0 19.5 21.0 24.0}
From [Q-box]:  {0.85 0.95 0.98 1.00 1.50 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.0 11.0 12.0}
Interesting that the Q-box is almost off by close to just a scalar (ie. similar values if you multiply Q-box by 2).  After multiplying the Q-box by 2, the shape of those (just by eyeballing) isn't too crazy far off.
Logged
Bische
Sr. Member
****

Karma: +25/-4
Offline Offline

Posts: 396



WWW
« Reply #86 on: September 04, 2012, 11:34:08 PM »

Question: Should one correct KFKHFM in the lowest load region(ie when the IPW is too low to be linear) before correcting the FKKVS?

Since the average 02 correction can exceed +15% and more in those regions in my case, I dont think it would be wise no? I was thinking of correcting the KFKHFM down to the lowest load where the IPW's are still steady, then extrapolate down in the lowest load columns.

After this is done I correct the FKKVS?
Logged
GaroBlu
Newbie
*

Karma: +0/-0
Offline Offline

Posts: 14


« Reply #87 on: October 09, 2012, 06:14:54 PM »

stock airbox, stock maf, stock intake setup, AWE intercoolers, RS4 K04 up-pipes, BW K04's, 3" ASP DP setup into a milltek single, stock fpr, rs4 fuel pump

currently testing the following:

...

will report back, but i believe this is my final setup calibrated to the stock airbox.
so serious and have yet to get the gas stink.  stft's and ltft's don't show much seek-and-find either...
will confirm this, but i'm convinced i've precisely cracked the code on these injectors.

Any news on this or logs? I don't have the stock airbox or maf and I am trying to get fueling dialed in. I would assume that if this was correct, similar post-maf setups would have little effect. This would mean that those of us with different intake systems would only need to bother with KFKHFM, MLHFM etc. I know there is a lot of debate on which maps to change, but I believe in getting a good baseline and changing the appropriately labeled map.
Logged
s5fourdoor
Hero Member
*****

Karma: +33/-3
Offline Offline

Posts: 617


« Reply #88 on: November 27, 2012, 12:33:02 PM »

Sup players!  Nehalem back again with your rock-star EV14 injector update.  No doubt, Santa arrived early this year!
Lol, anyways.  For the standard injectors, here's the deal.  EV14 52# @ 3-bar / 60# @ 4-bar.  Run the following:

KRKTE:  0.05495
TVUB:   2.5203   1.2695   1.0081   0.8881   0.7654

The elite trick:
KFFWL_0_A / KFFWL_1_A:  Multiply the entire table by 1.10, or a 10% increase.  Could be more or less, but I've found 10% to work quite well so far.  Obv, I'll update as I continue to tune.  Car starts right up and doesn't hesitate very much during cold operations.  You'd obviously be an idiot if you go testing the car's power while the engine is cold.  What I do know is that this trick restores most of the car's cold-temperature smoothness and eliminates the nasty bucking we all have grown to know and hate.  My theory, in fact many people's theory, is that the characteristic bucking occurs because the car's electronic throttle interrupts and retards timing and throttle-plate angle, which to the driver feels super-non-linear because there isn't a physical cable connecting your foot and the throttle.  Thus it's really an issue of "expected" versus "perceived" acceleration, for you HCI (human computer interaction) experts.

Therefore, you can't just "press the gas" more like you would on a cabled throttle [or an old / antique / carbed] car.  Basically, without this trick, the incorrectly calibrated cold start of the engine (read: overly lean AFR) makes the response feel like the car is straight up terrible to drive until the operating temperature raises a bit.  The worst part is that the closed-loop lambda operations turn the leanness into compensated richness, and thus the fueling variance of cold-starts without this trick straight up sucks.  It sucks, got that everyone, it blows huge dongage.  Use this trick.  [or whatever "trick" you like to use, just sayin.]


One last bit:
KFZWWLRL:  If you run advanced timing, like the baller we all know you are, then you probably want to *double* the warmup timing retardation chart.  But hey, you know your setup, and it's your engine.

KFLF, FKKVS, KFKHFM are still stock M-box.

OK home stars.  Knowledge is power.  Fly high young jedis.  Dr Brown out.

p.s.  I run the Stock MAF / Stock Airbox / Stock Bosch MAF sensor from the early model A-box. M.Y. 2000 car.
« Last Edit: November 27, 2012, 01:01:21 PM by nehalem » Logged
rnagy86
Sr. Member
****

Karma: +6/-11
Offline Offline

Posts: 367


« Reply #89 on: November 28, 2012, 01:18:24 PM »

Hey,

So I've migrated to EV14s a couple of months ago and I must say I am really happy with them. berTTos has also done the migration before me and he was so happy with the new EV14s not to mention that they were super cheap to get, I also did it.

So I am about to share our findings and experience about running with these injectors. Not that almost all of this information is comint from berTTos himself he just
does not have the time at the moment to post about it so I am just relaying this information with his approval and I have no intention to take any credit for it.

First of all we got the Ford Racing EV14s. I ordered them from summit racing for a really awesome price $285 for 8 injectors, plus $35 for the adapters.
The part numbers at summit racing are FMS-M9593-G302 (injectors set of Cool and FMS-M-14464-A8 (adapter kit). Since these injectors are a wee bit shorter
than the stock ones some people just install an injector adapter to make them as tall as the Siemens Dekas, but instead we just used the bottom o-rings from the
dekas to make them seal better at the intake manifold and I did not have any leaks there after doing about 4000 km.

I am running with KRKTE set to 0.05583 using 98 RON in Europe (which is still a little bit too much for me), but berTTos has to use a way higher value than this (0.06105) due to the fact that he is using US fuel which it seems has way more ethanol in it.
Also he had a chance to talk with some Ford tuners and they mentioned that in Ford applications it is common to see fuel pressure slightly exceed the FPR rating
when using a high flow fuel pump (which I am (Aeromotive)) and they also noted that it is usually a consistent deviation so it can be accounted for by slightly lowering KRKTE.
This might not be the case for us but I think it's nice to keep in mind in case someone can actually do a real test.

Currently the weather is kinda cold here, around 4-5 degC during the night and it's going to get colder next week and I haven't seen any cold start issues and the car is totally smooth even after starting to drive right away after a cold start.

berTTos found out that the stock TEMIN, TEMINVA are too low for these injectors so we have been running with the following values:
TEMIN: 0.7788
TEMINVA: 0.7788

Against the low temp wall firm development we have been using increasing the values in the acceleration enrichment factor tables (KFBAKL_0_A, KFBAKL_1_A) for low temperature ranges and the results have been pretty smooth.
Additionally KFVAKL_0_A and KFVAKL_1_A has been 0'd out completely and lambda regulation has been also altered.

ps: we have been using 85mm MAF housings with a Hitachi MAF. One car is having an old-style IM and the other one is having a new-style IM.
Logged
Pages: 1 ... 4 5 [6] 7 8 ... 21
  Print  
 
Jump to:  

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Page created in 0.026 seconds with 17 queries. (Pretty URLs adds 0s, 0q)