Kawasaki ZX-10R Forum banner

1 - 10 of 10 Posts

·
Registered
Joined
·
17 Posts
Discussion Starter #1
Read for many years every motor likes the AFR it likes. That may be true to a certain extent, but i suspect its more to do with how the cylinder head combo and valve overlap play a trick on the O2 sensor reading output. Raw fuel from the intake in the exhaust pipe makes for a rich condition, when infact the mixture was fine to begin with. Witnessed this on a build a bit back, when we kept adding fuel and MPH kept going up.

I know im new here, but im delighted to exchange some insight/ideas/experiences on this phenomenon with fellow members. This site is full of good info, figure i contribute.
 

·
Registered
Joined
·
975 Posts
A/F is and always has been a guide...not Biblical.....it's a tool to tune the machine.

5 gas, well, thats a whole kettle of fish i know little about, but tells a bigger and brighter picture.

As stated from the Factory Pro website, tune for best power....or Mph, etc.

Read your plugs, learn, learn and learn some more.
 

·
Registered
Joined
·
17 Posts
Discussion Starter #3
Datalogging for best mph and dyno tuning for best power dont equal the same E.T. in my experience. Have you have found that to be the same?

Additionally ramping up the AFR after peak torque has yielded good gains.
 

·
Registered
Joined
·
284 Posts
pretty new to dynotuning a few years back, had my 04 zx10 tuned for best hp on 1 dyno, best power came at 13.7 afr. +5 hp from 13.0. bike ran well but inconsistient at track depending on weather conditions. met my current tuner and we retuned at 13.0 full throttle position tune per him. bike ran better et/mph consistientley in the heat or cold temps. consistient at 8.70's/157, did run couple of 159's november races. he tunes all my builds exactley the same afr mapping at 10 20 40 60 80 and 100% throttle openings. busa's, zx14's,tens, 1k's. recentley tried another tuner on zx14 build with just head, original pulls showed power little low on his dyno, told him add some fuel, added .5 fuel bike jumped 8 hp. bike did not meet expectations on strip, pulled fuel back out, ran better et/mph. dont data log, dont change current tuners tune ever. i do what i do engine wise, all builds get fresh fuel pump and all injectors flowed before dyno and i leave the tune to him. couldnt be happier with results. stk piston gen 2 busas 220hp/121 trque, stk piston gen 1 zx14's over 200 hp/117 torque, mod for mod they run with anybodies builds real world at the strip. never see me or my customers on a laptop at the strip, we're there straight off his dyno. so ive experienced what ur saying, my experience- tuner/dyno error.
 

·
Registered
Joined
·
2,879 Posts
FYI: Raw fuel in the exhaust will not show as rich on a lambda sensor. If you want to verify this, richen it up until it misfires and see what the sensor outputs. ;) Lambda sensors are essentially telling you the composition of the gas... the pipe can be full of raw fuel and it will still read a lean gas if it wasn't in the combustion process.

With that being said, I've personally witnessed better power all the way down to 12.5:1 (ish) on N/A bikes. On my 07 I ran a dry shot that was tuned to 12:5, and spraying it was a steady 13:1. I ran out of bottle one day at the strip, and just left the 12:5:1 tune in it to see how it behaved, and low and behold it picked up consistent MPH over the 13:1 tune up on the back-up map. Even with forced induction you see more and more power up to that magical 12.5:1 mark, so long as your fuel can support it.

What will mess with your 02 reading are things like pulse reversion in short pipes, exhaust leaks, misfires, or if you have some really wacky cam timing/supercharging that is allowing you to flush "clean" air from the intake into the exhaust stroke event. Those sensors are also pressure sensitive, and the error under pressure grows the further from stoich the reading is. Changes in altitude can effect this, as well as restrictive exhaust designs. On top of that, sensors do age/wear... especially if you like leaded fuels. Sensor calibration is critical for repeat or comparative results.

Even if all engines were equal, every motor would likely have a slightly different 'happy afr' do to the myriad of variables in sensors, controllers, and installation. I feel like this accounts for most of the difference in what you see bike to bike, tuner to tuner. Jake is right about AFR only being one piece of the puzzle. EGT is a brilliant tool, and so is a 4 gas analyzer. At the end of the day, an accurate measurement of power (dyno, trap speed) VS destruction observation (knock sensor, plug reading) is the real measurement of what's best.... all the sensors are just there to help get you close and make it repeatable when you get to that last couple %. :) Just my opinion, of course.
 

·
Registered
Joined
·
17 Posts
Discussion Starter #6
Should have been more clear, air fuel mixture sweeps across from the intake to the exhaust during overlap. We've experimented with various valve overlap leaving the intake as a fixed LC. And we saw a defined change in air fuel. Additionally changing port texture on the same cylinder head using same 2 different LCs and overlap had a more pronounced change. We were head stratching thinking the LC's themselves was causing fuckery, but peak torque didn't change.

I'll say one thing dynos are great learning tools, just wish i had more time.
 

·
Registered
Joined
·
2,879 Posts
In that situation it could only really make it read leaner than it really would be, which usually isn't a destructive thing. What sort of different in AFR were you seeing change to change? I'd wager that +/- .5 of an AFR point from 12.5 would cover most of the power gains.

I've seen a lot of interesting stuff with individual cylinder tuning as of late too.... are you measuring average AFR or are you sampling all cylinders simultaneously? It could be real easy to mask gains one way or the other if you're tuning average AFR. Just a thought.

EDIT: Yes, dynos are so damn neat when they are used for education rather than phallic guestimation. Come hell or high water, I'll be buying one for personal use this winter. If you don't have data, you're just guessing, amiright? :)
 

·
Registered
Joined
·
17 Posts
Discussion Starter #8
It read leaner with rougher texture, and richer with more overlap. Could be almost .5 AFR. Funny thing is i have welded four bungs on a sidewinder for a Zx14, with 4 WBO2s, but my logger was awful and the sample rate was 10HZ. Not good for making any educated guesses. But i will say from cylinder to cylinder there was as much as 1.0 point AFR difference. Which is scary, but having low sample rate left me kind wondering. I haven't tried individual cylinder tuning on the dyno. But there is power to be had. Some people should start looking at the accleration chart on the dyno, offers more clues than one may think.
 

·
Registered
Joined
·
284 Posts
had about 70 injectors professionally cleaned/flowed on builds this year. mandatory on my builds. seldom see a set within 3-4 percent of each other before serviced. many in 8-10%lost flow range one or more injectors in a set, some clogged 30%+. do the math on about 13.1 afr, most common reason why individual cyl afr varies, why u think u have a good tune but actually do not, real deal, learned this hard way with dyno tuning. should not even have to deal with this,not a tuning issue, it's a builders issue. good build, balanced fueling per cyl, single bung cyl averaging tune works. time to power huge accleration factor, why u should throttle position tune.
 

·
Registered
Joined
·
2,879 Posts
Some people should start looking at the accleration chart on the dyno, offers more clues than one may think.
This is why I believe that final tuning should be done on the track. Not many dynos can accurately simulate load in sweep test (I only know of one that simulates wind resistance, and you would have to adjust the drum weight as well). I'd say the dyno -can- get you 99% of the way there, but that last 1% requires some performance based data.... and THAT makes me wish I was a whole lot better at chassis than I am. :badteeth:

10hz is pretty rough for datalogging, but it's good enough to show trends. At 200HZ you could technically distinguish individual cylinder pulses at 13K rpm with one sensor, but I'll be damned if I could figure out which cylinder to reference in that set-up. I've found that sometimes too high of a rate is just as bad as too slow... it's just noise at that point and hard to read with any accuracy.
 
1 - 10 of 10 Posts
Top