NefMoto

Hello everyone,

I've been tuning my injectors (Bosch EV14 52lb) to my car the past few days, and today did my first adjustment to TVUB since calculating it by interpolation.

I'm gonna use whole numbers in this example... these are not the values I'm actually using, just makes it easier to explain. I'm going by Tony's method described in his quote:

So let's say I started with TVUB values of:
5 4 3 2 1

And let's say my KRKTE is:
1.0

And after a good amount of driving at different conditions, my idle LTFTs were -1%. The way I interpret Tony's post is to multiply the idle LTFT (-0.01) by the KRKTE (1.0) which would give me -0.01. Then, if I understand his post correctly, I would add it to the existing TVUB values, across the board (although in this case it'd be subtraction), like this:
OLD: 5 4 3 2 1
NEW: 4.99, 3.99, 2.99, 1.99, 0.99

Those who've actually calculated TVUB values by injector specifications and interpolated theoretical values know it's not linear... however, if I interpreted Tony's post correctly, suggests to adjust it linearly. Is this correct, or did I misunderstand his post?

Thanks,
Brett

for that method you must go through couple of cycles of adjusting long term low range and high range LTFTs followed by TVUB to get idle in checks.

Then repeat it.

The problem is that TVUB affects idle the most but it also is a natural component of injection time for any load range albeit the longer the injector is open (higher load and maf flow), the less % wise TVUB affects it.

At idle, if the injector has to open for 0.5ms with TVUB of 0.7ms, the tvub is actually larger then required pulse width of injector to reac proper mixture. If TVUB is off by 10%, it will inbalance the mixture by 0.7/0.5 = 1.4, which gives 1.4 * 10% = 14% off the proper mixture.

At higher load when the injector needs to open for 10ms, the total open time wold be 10 + 0.7 = 10.7 which gives 0.7/10 = 0.07, which gives 0.07 * 10% = 0.7% off the proper mixture.

That's why you want to get the KRKTE by riding the car hard and once you have KRTE honed down to within couple of %, you switch you attention to idle and use tony's method. Eventually you will get it right however not the way you think.

You don't add just 0.01ms across the board since adding 0.01 to 0.01 means upping it 100% while adding 0.01 to 1.00 means upping it 1%.

Instead, you need to divide intended bump (0.01 in your case) by the TVUB value at the voltage you normally operate (14v or so) and get what part of the injection time it is in % at that voltage.

For example, you want to add 0.01 to say value 3 from your example.

0.01 / 3 = 0.00333

Now you want to multiply each TVUB value by 1.000333 and you will get TVUB scaled up by 0.000333 across all value EVENLY.

To properly interpolate TVUB you need to have reliable TVUB specs from injector manufacturer. And remember, TVUB values are always given for 3bar of fule pressure. Audis usually run 4bar FPR which means the injector needs to overcome more resistance and its TVUB needs to be multiplied by some value, usually ~16% for 3 to 4 bar bump but it might vary.

I hope I didn't bore you to death.

I had a feeling the way I did it was wrong... and that's why I asked. Thank you for the clear explanation!

As far as calculating the initial TVUB, I used the calibration summary that was linked on another thread for my injector, and interpolated the value for 4bar and multiplied it by to the interpolated values for voltages used in our TVUB map. That much I had down, it was just adjusting it after logging idle LTFTs that I was unsure of.

Are these the values you used

6 = 5.202
8= 2.184
10= 1.435
11= 1.210
12= 1.041
13= 0.907
14= 0.789
15= 0.699

Sort of. You have to interpolate the ms values for the volts values used in TVUB. BUT you have to multiply all those values by the multiplier for 4bar of pressure, which ALSO needs to be interpolated. But yes, you are on the right track: http://www.boschdealer.com/specsheets/0280158117cs.jpg

I will be posting a thread soon with the theoretical values for Bosch EV14 52lb injectors, and an excel sheet on how I came up with them.

I've got a question, looking for some verification on units from the example in the OP.  So KRKTE is in (ms/%), and you're multiplying it by the idle LTFT (in %) to come up with a value in ms.  In the example in the OP, KRKTE was 1.0 and LTFT idle was -1%. Is it correct to multiply 1.0 * (1), or 1.0 * (.01) to come up with the value?  Examples I've seen in other threads show the former, but the example here in the OP shows the latter, so I'm a little confused on the proper units (they're both technically the percentage).

Latin "per centum" = “by the hundred”  = 1/100th of a unit.

1% represented in mathematically digestible form is 0.01.

When using in a context "increase by 1%" it really means multiplying original number by 1.01.

I missed out on the latin lesson when my engineering professors talked about percentages ;)

I understand that in general, what threw me off, or rather made me question it, is the "dynamic_krkte_tvub" calculator excel spreadsheet in the other thread I just bumped.  If you look at cell E18, for example, it is multiplying the fuel trims in full percentage form (ie multiplying KRKTE by one percent by literally taking the KRKTE*(1)).  There's conflicting info, and I am just trying to figure out which is correct.  With load being units of percentage as well, I thought maybe there was some funky units thing going on for the calculations inside ME7.  Thanks for the clarification :)

 :-*

I don't have that spreadsheet but maybe there is some special notation in use that tells excel to multiply my % even though he is using whole numbers.

that spreadsheet i coded desperately needs repair.  no debate from me, i promise.
please give me any calculation corrections and i'll repair it.

I really think "tuning" TVUB based on log isn't really needed. You should start with the injector spec and use that. If your idle trims are significantly wrong, something is up with either your MAF or KRKTE, NOT with your TVUB

Thats my opinion, of course. Everybody tunes differently

Also, here is what I am using for my EV14s

http://nefariousmotorsports.com/forum/index.php?topic=320.msg18405#msg18405

Hi everyone,

have few doubts regarding LTFT & STFT.

Is LTFT = fra_w (multiplicative)
   STFT = rka_w (additive)

I am running bosch injectors at 4bar with 3bar tvub values. my long term fuel trims at idle are 1.3% off and at partial load thy are 0%. So my KRKTE seams to be perfect and my TVUB needs correction.

change in flow from 3 to 4bar fuel pressure is 15,5% or ~16 what julex said.
Is this the proper way to scale TVUB for a changed fuel pressure?

Leave TVUB alone. 1.3 percent is within the acceptable range of deviation.

Hmm thats what i heard before.
Just out of interest, whats the correct way if fuel pressure changes.

Lifting old thread

If idle is off by ie 2% at 14V. Why cant we just add 2% of TVUB value to that row?

If idle is off 2 percent don't touch it.

I know that, was just an example.
What about doing it this way?
I did some math both ways, the result is almost the same???

The only way you should be adjusting TVUB is if you don't have the data from the manufacturer of the injectors. Is this the case? Otherwise, input the values you have, and use FKKVS.

I have been working on my fueling a lot.
Idle is off by 6% as now.
I can not effect my idle with FKKVS??? Those tables starts at 1240 rpm. Or can I?
Or does it use first column values for rpm under 1240 as well??

My fuel trims are idle -0.1; partial on normal driving are -2.0 but if I give it a spirit driving fuel trims go to -6.0, does this mean that I should adjust  KRKTE? I need this right before go to RS4 MAF

I agree.

Remember. Axis can be rescaled. In addition, the for any values UNDER (or over) the axis, the first (or last) cell is used.

In any event, 6 percent isn't the end of the world. It's easily explained by the non-linear nature of injectors running low pulsewidths.

Also, O2 sensor accuracy drifts over time.

That said, a known good TVUB that isn't working right is also an indication that something else might be amiss in your hardware, or elsewhere in the tune.

Good to know!!
So, TVUB should never be touched if injectors come with right manufactor data.
I had the right data, but i understood the process wrong.
It is all good now.
Thanx guys!!

Well, also, don't forget there is more than one way to skin a cat :)

So everybody has their own way of doing things.. if you can't get anything to work right, but TVUB works for you, and the results are good, well..

These are just recommendations, please take them with a grain of salt!

Regardless of a datasheet, if the only consistency during fueling error is battery voltage, then TVUB should be adjusted.


This is the exact (and only IMO) time FKKVS should be used for correction.

As previously said, everyone has their own methods that will affect how you identify and correct fueling errors.

Alright, then how exactly would you recommend fixing fueling? KFKHFM? I've spent the time to dial in a MAF, dial in the injectors, and still needed to clean it all up under open loop.

After a lot of logging I can say that TVUB should not be touched if you have right data.
I was chasing my tale, fixing all over fueling, and alog with that I got worse idle.
Than I tweaked TVUB  10-12%, which resulted in leaning up top at WOT.
Changeed back to original TVUB and voala, I got all the fuel all the way and then some.
The rest I fixed with Fkkvs.
So TVUB does effect overall fueling not just idle.
That was my findings.
I also found out, that I make the most power running lean on e85, afr 12.5-13. I would not recommend it though.

TVUB is for correcting injector latency dependent on voltage

FKVVS is for injector non-linearity

KFKHFM is for MAF non-linearity

Now, you can ignore those guidelines and correct fueling however you want.

But randomly changing things until things are "correct" if you don't know what your underlying issue is probably not the best approach.

I was not all over the place, I fixed my krkte, and it affected my idle.
Than I made mistake and worked with tvub, and it messed up other things.
Trial and error :)
It is perfect now, lesson learned

You need to isolate the error, identify the root cause, and utilize the appropriate correction.

Applying the software as designed works perfectly, but your method has to do this by design.

I understand that, I was more asking YOUR thoughts on it. Especially when hitting the ps_w cap. I know we've hashed this out to death, and I know the proper way to do it, but honestly, does anybody do a perfect calibration?