NefMoto

Can anyone educate me as to the relationship between KVB (cm^3/min) and the mpg display of fuel consumption?

I'd like to re-calibrate and although I can get a decent figure for mpg I'm not sure how to adjust KVB to scale it properly.

Cheers,
C.

Hi Carlossus.

This is an interesting question. I'd like to know how to do it as an academic exercise so I might look into it further if I have time. I recall seeing something in the funktionsrahmen about the displayed fuel consumption. However, if you just want to correct your displayed fuel consumption to more accurately reflect the actual gas mileage you are getting, this can be done easily with VAG-COM/VCDS. Details are in the attached text file because I'm still experiencing truncation of posts.

TTQS

Thanks TTQS. That's probably an easier way to go although I'd really like to understand how the constant is used. Unfortunately my engine mount just let go so fuel consumption is low priority now :(

O.k., I did a quick translation of pages 1593 & 1594 of the funktionsrahmen for the KVA 41.40 module with equations and constants. See attached Word 97-2004 format document.

TTQS

Thanks for taking the time to do that. I'll eventually learn to search the funktionsrahmen for myself <insert face-palm here>.

So, that's pretty clear (apart from RKTE_W = Relative proportion tank mixture VENTILATE?). Seems if KRKTE is calculated correctly we shouldn't need to mess with KVB.

So, using the VAGCOM method presumably alters adaptation channels... does anyone know if these get wiped when we Re-Flash?

Ta.

Yeah, I don't think there should be any need to change any of these parameters except KRKTE if, for example, larger injectors and/or a higher pressure fuel pump is used.

On point 1, that translation is rough, I'll have another look at some point and tidy it up and add the diagrams for completeness. RKTE_W should be "relative proportion of fuel-air mixture in the fuel tank breather" I think. My laptop is dead at the moment (NTLDR files got accidentally deleted and I don't have a boot CD-ROM because both operating system and applications are on a recovery disk) so I was using my wife's Mac which I'm not used to.

On point 2, I would think this is almost certain. I believe these will also all clear if power is interrupted to the ECU, e.g. if you have a flat battery or disconnect it.

TTQS

the adaption channels wont necessarily reset when power is removed.  i took out my ecu and removed the eprom then put it back in and the adaption channels were still they way i had them.  but i have also done it and had to reset the adaption channels.

My understanding is that adaptation is non volatile, stored on EE. it's just that Tony's software warns of losing adaptation info when you flash. I flash a fair bit so it would be a pain to have to keep using VCom afterwards.

Flashing over OBD sets the adaptation channels to defaults.

The way I scaled my KVB, was I got my KRKTE scaled correctly, then I scaled KVB by the same amount.

So, KVB is the  average flow rate of the injectors given the conditions used to calculate KRKTE?

It should then be scaled if FPR pressure is raised?

It should be scaled according to Bernoulli's eqn, hence in direct proportion to KRKTE?

Phew, aaaand breath.



Damn, cant get this to display correctly...
                         TMAIN
                         ---
                         \
TKVAG =            /     (rk_w + rkte_w) * KRKTE [ms] * KVB [cm

(http://oi54.tinypic.com/ht7oef.jpg)

sorry nyet, I genuinely tried to use wiki format but lost the will to live.
Have TTQS' version instead.

Post it in mediawiki format and i'll put it on the s4 wiki, and it will generate an image you can post here ;)

http://www.mediawiki.org/wiki/Manual:Math

Heh. I feel your pain.

FYI something like:

So, to close off the thread: -

KVB is injector flow rate.
If you up the fuel pressure, the new flow rate is calculated by: -

FN = F0 x SQRT(PN/P0),

FN = New flow rate
F0 = Flow rate at old pressure
PN = New fuel pressure
P0 = Old fuel pressure

...BUT if you also up the boost then is the flow rate is affected by the differential pressure across the injector or is this negligible? I have seen differing rates quoted for the EV6 injector at 3 bar.

The rising rate fuel pressure regulator handles keeping the fuel pressure the same relative to the intake pressure.

Perfect, thanks.

As an aside it is common to put boost gauges on the FPR pressure line... which is where the FPR gets the "reference" manifold pressure.. and which might be different from MAP readings, since the MAP is pre-throttle plate.

Hmm. Well I just calculated my offset from the six refuels since my remap and the average was 111.1% and there was good cluster around the mean.

So I plugged in my laptop and VCDS and followed the procedure I uploaded only to find that I couldn't test the new value. Instead I got an ERROR message. I checked back with the original remappers to see if they locked out any access to change adaptation channels and the answer was no.

I'll try logging in to the ECU see if that makes a difference, but I'm not hopeful. Maybe an e-mail to Ross-Tech will shed some light on it...

TTQS

Do you need to log on with your Security Key code (SKC) first? Just a punt.

Well I eventually got it to accept 110 so I suspect that despite the pop-up windows telling me the value could go up to 115, I suspect that (on my ECU at least), the limit is 110.

Some hours later...

Another go reveals that it will accept 115%, so perhaps it can only deal with 5% increments which isn't very useful because that leaves a choice of only 105%, 110% or 115%. I've e-mailed Ross Tech support to see what they say.

TTQS

Did you try using Lemmiwinks? Lemmiwinks bypasses the limitations imposed on the adaptation channels and writes directly to memory.

No, I haven't progressed beyond VCDS just yet.

Andy at Ross Tech replied requesting an Auto-Scan log to see which controllers the car has got and whether there are any fault codes which I've just supplied. There were no unexpected fault codes, just historic stuff relating to an alarm activation and a faulty brake pedal switch which I replaced several months ago.

I don't think this will shed any light on the situation. I also had a problem with installation of the VCDS drivers but deactivating my anti-virus software sorted that. Full of hope, I tried again with the fuel consumption correction, but still the same result. :( 10% is good enough, but it would be nice to get to the bottom of this.

TTQS

FWIW i didn't bother to "derive" my KVB from first principles. I just filled up to a full tank, reset the MPG trip, drove till it was empty, refueled, calculated my actual MPG, compared it with trip MPG, and changed KVB by the percentage error.

Only had to do it twice, and it converged fine after two tweaks.

Andy at Ross Tech confirmed that the factory repair manual states that the increments are indeed 5%, not 1% (for the TT 8N at least) as allowed by VCDS, so it would appear that you would be better off adjusting KVB to get the displayed fuel consumption to read correctly.

TTQS

Well I corrected KVB after running a few tanks for MPG calculation.

I was quite suprised that the error was 24%. I hadn't previously checked, but looks like my tune had not  corrected KVB for 3bar -> 4bar fpr conversion.

So the reality is that I'm averaging mid 20's for a stage1 1.8t. It got me thinking a bit about the big name tuners. There is such a wide range of mpg's banded about on the cupranet forum (20s to 40s) that is largely populated by APR, REVO and Custom code tunes that I wonder if these guys make any correction at all from stock?

I'll add to this thread.

Standard KVB value can go only up to 776cc injector size due to reaching the maximum value it accepts.

To accommodate injectors larger than that you have to also change FKVA "Constant conversion factor for consumption display", S4 Mbox 8bit at 0x1A4D5, conversion 0.062500*x.

This is a multiplier for KVB. I have it at 2.00 in my tune and KVB around 500 to accommodate my 1000cc injectors. It works :)

I can't tell you much about the constants and formulas as you guys are way out of my league but in the vag-com method since the adaptation is done through the clusters adaptation channels I'm pretty sure the base calculation is in the ecm but the correction adaptation is done in the cluster.  

If I'm right flashing the ecm won't affect your correction factor even if adaptation channels get reset and it means even if a tuner locked out any channels in the ecm you'd still be able to adjust it.  

In addition I don't think lemmiwinks would allow greater than normal tolerances as I believe it only works on kwp2000 me7 ecms, most if not all of the clusters I have seen are kwp1281.  

However, I'm not sure if all clusters support this function as the documentation I've seen doesn't always mention it depending on the car.  you may be able to make adjustments manually to the clusters eeprom and write them back though as I suspect they are visible somewhere in a binary file if you played around a bit to locate them.  This method may allow greater than normal tolerances if needed 

If there is a login required to access the adaptions channels in the cluster 13861 is the only login I know of, other than the skc to reach channel 21 for key programming in an immobilizer car. 

No pun intended but please, use paragraphs if you want people to actually read what you're saying.

And yes, each cluster has a small adjustment range to accomodate off-fueling system that might come out of factory. But this is peanuts, I think around 10% either eay or so.

Most high HP cars like run of the mill Stage 3 is already 2x over factory power with fueling system delivering around 200% (in case of 660 cc injectors) or slightly less than that of fuel. Cluster doesn't hae enough range. Binaries for ECU need to be changed for cluster to show MPG accurately....

or you can leave it as is, enjoy virtual 100% :)

Sorry about my crappy grammar.  You were right.

I wasn't implying that the vag-com method was the end all fix.  I realize you guys get way outside normal tolerances.  I just wanted to correct a couple misconceptions I read in the thread.

Arg! Thank you! lol... I was all confused why it wasn't going higher... then I figured there was a limitation on the value itself.

Thanks for the tip.. I'm going to try that now.

Thanks Julex. Added to the s4wiki. Also, it should be in my next map pack release.

LOL I just signed up to the wiki to add it, and then I noticed it was there and was like WTF?  Was it there the whole time?  Good to know I'm not going crazy (yet) :P

FWIW I haven't been successful in using the modifier to get my KVB settings to give me something useful as far as fuel consumption goes...

Anyone??

I set FKVA to 10 and lowered my KVB setting to around 120.... my display show that I'm using 2L /100km...

I'd really like to get this dialed in

I *increased* KVB from 319 (stock) to 712

60lb dekas, 3bar fpr

stock FKVA

With my KVA maxed (around 712 or so) it's still not enough... I have 1000cc running @ 4BAR so almost 1200cc.

My understanding was that FKVA was a multiplier.... but it doesn't seem to work.

pic for giggle:

(https://lh4.googleusercontent.com/-HKFxa_OS4ps/TqHAM6yMU7I/AAAAAAAAA5o/uC96cQQ8zNY/s1024/IMG_20111021_141905.jpg)

Ah I see.

Looks like we'll have to do some FR digging.

It seems to work for julex, don't know what it doesn't for me :(

http://nefariousmotorsports.com/forum/index.php/topic,496.msg5693.html#msg5693

I'm in the same boat as Notorious. I've maxed the maps and I'm still about 40% short of actual MPG?

1000cc and 4bar pressure.

Did anyone figure the differences between working and non working setups?

Everybody has their own theory on how to set this.  It's actually quite easy if you take a numerical approach.  Divide your injector size at the operating-pressure you use by the stock size at this operating pressure.  OK:  ev14's @ 4-bar ~= 630cc/min   Stock b5 s4 ~= 349 cc/min      (630/349) ~= 1.75.

Set FKVA to 1.75.  Write your tune.  Reset your tripometers.  Fill up gas and start with a full tank.  Don't overfill.  Use same pump / same gas station for both fill-up and re-fill.  Calculate your actual MPG based on a precise procedure such as I've described.

Since your FKVA is large enough for huge injectors, KVB now has enough dynamic range to be adjusted precisely.  Rinse/repeat your procedure using static FKVA, estimated like I talked about above.  After a few trials your dash mpg will be within +/- .25 mpg.  Actually using this, I've got my mpg to within +/-.1 of the actual.

The trick is that you have to a) follow a precise refill procedure, b) understand what you are doing, c) start by making a static large-adjustment to FKVA from it's 1.00 stock value to .25 step-size rounding and then leave it alone.

The problem most people have is that they don't actually understand the equation governing this stuff.  Look its simple algebra.  Get your FKVA roughly adjusted, because the precision of KVB isn't sufficient for large injectors without FKVA bumped up.  Reread what I've written here, it's very easy.  Nyet's the one that tipped me off on how to properly do this.

Ok, now I feel really stupid!

What are the theoritical starting values (FKVA and KVB) for 1000cc@4bar ~1200cc of E85?

Can the FKVA be too big and still cluster would show too little?

For me: 1200cc/349cc=3.43 for FKVA (haven't tried that yet!) BUT if even 15@FKVA AND 770cc@KVB is not enough (about 40% short) and IF it's a multiplier I don't get it! So I will try 349cc and 3.43 but I have low hopes for it to be anywhere in the range of real MPG but as I said I'm sometimes stupid and stubborn.

For Julex 500cc KVB  and 2 at FKVA works and IIRC he runs 3bar fuel pressure and that goes with the theory.

Man, trust me, it's way easier.  You are thinking too hard.  I have 60# / 630 @ 4-bar bosch-ev14.  I am using 1.5 * 570 ~= 855.
Julex has 97 # / 1015cc @ 3-bar InjDyn.  He uses 2.0 * 500 ~= 1000.
You are using 1200cc @ 4-bar, arguably 1000cc @4-bar.
(i'm not sure which and it doesn't matter which.  that was my point with the "numerical technique approach".)
Set your FKVA to 3.0 and set your initial KVB to 400.  3.0 * 400 = 1200.

Start a static FKVA, do the procedure I've listed.  Report back the following.  I will help you make the adjustments.
We require:  Miles Driven, Fuel Consumed, and Tripometer MPG from your LCD display.

Seriously, its this easy, I promise everyone this.  Nyet advocated this basic idea on the wiki starting two-plus years ago.

Thanks, I will try it later today or tomorrow (left my laptop home) and report.

20.6 mpg  with 15 and 770cc

and now with the theoretical correct values:

28.3 mpg  with 3.4 and 350cc

Real MPG is about 15.7

So it went just like I thought. It's not THAT simple always it seems?

Updating this thread with some relevant info.

FKVA is used only in case of canbus clusters.
If the cluster takes the fuel consumption from a dedicated wire, then FKVA is unused, as FKVA is only used to scale canbus output.

Thus you run into an issue where you can not accomodate 800cc and larger injectors.
The solution is a direct modification to the code calculating TKVAG.

Since KRKTE is shifted right 4 times before it is multiplied in, you can reduce the amount of times it is shifted right, thus multiplying the effective value by 2 each time.
So changing 7C 42 to 7C 32 will double the displayed quantity, 7C 22 will quadruple it and so on.

Attached illustration (551K RS4 bin)

Hi all.
I have now 550cc injectors and i put KVB to 550 from 386 who was from original BAM ecu file.
FKVA in original file is 1.0625 for 386cc , somebody know why? I think must be 1.0000!
Must i modify FKVA to 1 for my new 550cc injectors?
Thanks in advance

Five should be 1
Bad def probably

Thanks

I checked again for 2 different file for
8N0906018SJ and 8N0906018SJ and 8N0906018H  0004 and the definition are ok at 0x1A477 with this conversion eeprom*0.625 and result 1.0625 value in tuner pro and winols too
for 8N0906018BP_0002 value in winols is 1.00
Something is strange
in the left is 8N0906018BP_0002

Reply after some time but maybe tell something more :)
Yes FKVA is 1,0625 not 1 because KVB is 371,72 not 386 but really injectors are mor like 394
and for 1,0625 x 371,72 we have ~`395
And for me if I take 395 as size when go from 3 bar to 4 bar fpr my KRKTE is 0,2% wrong so for me thats why constant are 1,0625 not 1

For injectors that flow more than 776cc why not just multiply the conversion factor for KVB by 3, put in your normal injector flow rate, then set FKVA to 3. This way you wont need to recalibrate FKVA vs actual mpg much and dont have to guess.