funstick

i wanted a chance to speak freely so if the thread get locked hopefully it wont create to much of a stir.

Wb02's can be useless.
egts can be useless
reading plugs can be useless.
datalogs can be completely useless.

just ponder these 4 statements for a minute. then continue on

ive been wiating for somebody to really chim in a for a while to spark a conversation. but it has yet to happen. so here i am doing it.

you narrow band isnt really all that wrong
your wb02 can lead you in circles
plugs are very helpfull at times
datalogs clue you in

now WTF am i saying here you ask. im asking you to chellenge your convetion of thinking.

when does all this become usefull ???

when its looked at in a bigg broad way.

im gonna type in caps here

YOU HAVE TO ADD IT ALL TOGETHER !!!!!!!!!!!!!!!!!!!!!!!!!!!!

im sick of hearing about closed loop. forget about it. its about worth 2 piles of dog crap. its not the holy grail of tunning and often it'll drive you down the wrong road. once that cars runs good then find a way to make the computer control the fuel trim to the engines satisfaction not trying to let the comptuer trim it to its preprgrammed pid satisfaction.

so whats the deal you ask.

here it is.

dont focus all your effort on your wideband. while it can be a usefull tool rememeber its only reading whats comming out of the cylinder nots whats going in.

definition of A/F

homgenious mixture of fuel and air mixed in proper quanititys to create stiochemitry. well stiochemitry can vary a great deal depending on the fuel. typically gasoline is 14.7:1 depending on iso octane content and sevral other factors.

well this is where a MAF system has a broad advantage. its measures air directly. this means the fuel will always be delivered in a very proportional manner in regards to fuel. that doens however mean its 100% accurate. acctually the current deviation for even the newest vehicles is 2%.

this is where im going. no on particular item of the job of tunning can be ignored. the wd02 is a tool not the holy grial. if you forcusing on just the a/f as ( OBSERVED) by the wb02 you could be running really rich and or really lean and not catch it becuase you can affect a/f with timming.

plug reading.

very useful tool again but when viewed by itself can be misleading. not nearly as much as the wb02 but just as bad.
you have to add the results of all things to create a sum total


i dont want to dive much further becuase i hope by now ive made a point. you have to look at the big picture. you have to take all the info in. a datalog might show you where the ecm see's the engine operating so you know what parameters are being used to determine fueling and spark but it wont tell you how wel the fuel is combusting or how much is really required.

you should always make use of every critical peice of data you can aquire. it will only aid you in your cuase. if you become narrowly focused then you might mis something incredebile that could have saved you engine. i have this agrument with all sort of ( professional tunners) quiet frequently. way the evidence and make logical decisions. take in all information at you disposall. if you retard timing the engine will appear len if you advance it will appear richer and all withouth actually chnging the actual fuel table and it will show up in a wideband with no problem.

my tunning toolset becuase im spoiled

5gas
wb02
3 plug glass's of varyng power
4 types of dataloggers ( not just gm obd1)
multiple types of tunnign software
air measuring sensor ( not a MAF )
egt probes and surface temp probes.
ignition scope

the idea im trying to convey is not to narrow yourself to unidirectional thinking. thats what im suggesting. once you hit you IDEAL AFR then pull plugs have a look. do more then just wb02 fuel and spark. open your eyes. borrow some equipment. also i find secondary ignition patterns extremly useful btw.


i will say this much. unless you plan on doing heavy coding. stop worrying about ECM benches. dont worry about understanding the ecm. its logic is fiarly clear ,retrive input values use preprogrammed outputs for given inputs. it doesnt think. its just does math.

anybody like to add to this ?

1981TTA

First, the major point where I agree.... Using all the information available is always a good idea. Depending on only one source of information can definitely lead you down a wrong road. The times when all the available information points in the same direction usually indicates the correction/modification/analysis is good. Critically reviewing all the information from *all* the sensors, many time including the SOTP sensor, is very important

Now, the areas I take a slightly (to largely) different view....
Agreed it's not a holy grail. But, it can have the ability to compensate for conditions not included in a given software mask. And, it can attempt to compensate, over time, for component variations. For someone who datalogs and adjusts calibrations *every* second of engine run time, the compensation for variation may not be a big deal. These people would be constantly making calibration changes for any system changes. For those of us that might run for periods of time (days to months to years) without reviewing datalogs in great detail, it makes me feel better to know something is going to try to compensate for the injector that's starting to get clogged and flowing a little less rather than just going lean. I would agree the closed loop algorithms shouldn't be used as a cover up for lack of attention in other areas of the calibration. I'd also agree using only O2 information can be very misleading. Look at the number of people with considerable engine modifications that have to "ignore" the O2 sensor to get things running halfway decent. Something must be happening to make the O2 sensor "confused" in these cases. Again, this is a good example of an area that can benefit from the use of the other tools available to determine what's going on.

I've heard 10% deviation is closer to the OEM spec. Mind you, that's from sensor-to-sensor. Don't know what the variation of a single sensor is spec'd to be. (Maybe this is the 2% you mention?) I do know the same sensor will tend to give different readings when placed in different induction hardware (air cleaner, piping, etc.) or placed in a different area of the same induction system. There's a lot more variability in MAFs than most people think. Given this, how would the average individual determine where they are after bolting on the MAF? Oxygen sensor feedback, maybe?

Nothing personal here. However, since you would seem to understand the "fairly clear" algorithms, how about a straightforward writeup for the rest of us? It seems a large portion of posts on this board are concerned with engine behavior caused by ECM "math" that's not understood or calibrated correctly for the engine application. For example, a recent post describes a problem with spark moving around at idle which is resulting in stalls. Recreating this condition on an ECM bench to see if anything pops up could be very revealing. Maybe things would be better put this way... An ECM bench is a *requirement* when doing code modifications. An ECM bench *can be* another helpful tuning tool even when not doing code modifications. I'd view an ECM bench in a similar light to WBO2 sensors. As you mention, they don't solve all the worlds problems. But, they have the potential to be much more helpful than just a NB. Given your viewpoint of "adding it all together" and your list of impressive hardware/software tuning tools, I'm somewhat surprised you don't include an ECM bench to your toolbox.... The physics of an engine are pretty clear, too. But, there are a number of people and companies that learn "something new" or a new way to do things all the time. Can a person get by without understanding ECM logic? Yes. Can a person get by without understanding engine physics? Yes. Can a person walk into an auto parts shop (or go online), randomly pick up a cam, heads and induction system, bolt them together and have an engine that runs flawlessly? Yes. However, I think we'd all agree the chances of the latter are incredibly low. Only after there's a reasonable level of understanding of each of the above will things come together. Expecting the ECM to cover up a poorly matched hardware combination is unreasonable. Expecting hardware to cover up poor ECM calibration is equally unreasonable. Only by understanding each to some level will any problems have a hope of being resolved.

Again, I'm not trying to make any personal attacks on your views. And, in re-reading this post, I notice quite a few "I agrees" sprinkled around. Hopefully, we can keep this civil and have enough input from a number of people to learn something new ourselves!

funstick

not taking it as a personal attack im trying to get people thinking. i dont like closed loop. do a good cal and the code will work exactly as its intended to.

JPrevost

First off, the only personal attack I'm going to make is that your spelling sucks, spell check . Now for the non personal part of reply; How can you said you don't like closed loop and generalize like this? If you were tuning a car that had a cat and HAD to have a cat, would you also run it in open loop? I wouldn't, maybe open loop idle and WOT but the part throttle stuff would be closed loop for the sake of preventing cat damage. Another reason to keep closed loop is sensor failure. A sensor goes bad and it starts dumping a ton of fuel into the engine, or the other way, and closed loop will get you home without spitting molten ***** of cat guts out the exhaust. So I think it's YOU that needs to look at the big picture. I don't tune every vehicle with MY preference, I tune the vehicles to the vehicles preference. This keeps me on my toes. I only check plugs when I'm WAY off on the tune (like a stall or no start or ruff running) or when I'm so close to perfect and am now getting the proper heat range selected. The big picture is being able to stand back, get some requirements, come up with some goals and then determine a set course of action. Engineers do this all the time and it's why they are usually the best at getting a good tune in an engine FAST... not to mention they should have access to a lot of useful tuning tools.
I use one method every time I walk upto a tune and that's "give the engine what it wants." I could care less what I've got to work with, all I know is that I want to make the best of it and if there's no extra o2 bung and the cat is there then I don't bother with a wideband. Sometimes, with this one tune I've been fighting with, you need to use these tools to confirm your progess with the code. I personally believe in spending LOTS of time coming up with a safe starting point calibration to prevent engine wear and tear. With 14.8:1 compressiong, 16000rpm, and only TPS to sence load I wouldn't DARE subject the engine to a generic starting point. I spend more time on the bench and doing calculations than I do tuning a running motor... and this is why the best tuners (in my view) don't bother with real-time tuning. They sit down and review everything with a hawk eye. Now of course those same tuners know when real-time tuning is a benifit (idle, cruising no movement TPS, and WOT dyno pulls).
So myth or not I can't agree with your generalizations but I can agree that there isn't one tool or one path to perfection.

funstick

ok heres an odd but true fact. the code fuel correction logic is course at best. and if you had a sensor fail and it threw a code it would immediately go openloop. not to mention after it goes open loop depending on the malfucntion it may or may not dump fuel.and yes ive tunned cars open loop with cats. if you take your time and plod through a good cal before inserting said cat then you dont have to worry about catalyst failure. while people like to beive cats only function at x afr thats not entirely true. cats function at a chemical combination of exhuast gas that cant be achived just using afr.also factory engineers do al there in house tunning OL and the fuel trim is only there to correct for engine wear. but also thats why almost all OEM's have now gone to mass air systems. there not worried about observed afr there worried about actuall afr.

JPrevost

The cats work best when you go rich lean from stoich, not held at a steady afr. Also, factory engineers at Honda do not do all their in house tunning let alone OL. I can confirm that Cadillac doesn't do what you say they do.
As for engine wear, nope, that's not the only reason why fuel trim is there. Not all computers go into open loop with a sensor failure (thank ***). Oh, and that isn't the reason why OEM's have gone to mass air. Again, you're not looking at the BIG picture. I don't expect people to know the details and this is why I try my hardest to simplify and teach the people I help. By giving them the real reasons and not just the quick answers I might overwhelm them with information but after they ask me a question about a lengthy reply things start to click. It's like math, teaching somebody how to integrate is fine and dandy but show them how to use integration and their eyes get bigger .

dimented24x7

iTrader: (2)

The 10% deviation would be a large varience in fueling with respects to drivability. Youd definatly be able to notice it. With a basic MAF algorithm, the fuel varies in proportion to the MAF sensor output so a 10% change in output would be about a 10% change in fueling.

Any time you change the intake ducting or sensors, though, the tables should be recalibrated to the sensor. Ask the LS1 guys, some of them will chew your ear off witht he problems they encounter with different intakes. There was even one guy who had a meltdown on a message board after his LS1 firebird ran like complete *** at idle because he cut out the screen.

I found out in a real hard way that the calibration table is way off when you chagne the ducting. I parked a massive truck maf with nothing more then a k&n in front of it and put the stock values into my MAF table. The engine was way lean.

funstick

dont know which place your getting your source info from but all the tunes begin openloop. i hang out at quiet a few places where the oems rents dyno time and develope there calibrations. its not till the engine gets in a car and the cal has been run significantly on a a/c engine dynofor 300-400 hrs that the engine even gets in a car. then once its in the car they enable closed and beging the road validation running closed loop. also the shed and standing test plus the EPA drive cycle are all done just minutes from my home. i spend time down at the EPA test site occasioanly.i know a few calibrators from the big 3.

as for catalyists there operandi does not come from switching from a rich lean threshold either. cats dont operate on the principals you think they do espcially this one. also new 3 way obd2 compliant cats also have a much greater need for more hc and c0 to work properly. regardless of afr. the only reason the o2 sensor pid toggles afr is becuase the 02 sensor cant read anything but above and below stiochemetry. its cant even read stiochemetry.

funstick

a screened meter with the engineered and vlidated ducting system including filters is acurate to within 2-2.5%. if everything is in top notch condition. i know all about MAF calibration and the effect ducting has on it. spent some time working for pro-m racing a few years back doing development work on the gm meters. and a few others.

Grumpy

Maybe when you understand all that can be done with one, you'll grasp how important they are.

What's your point other then bragging about have a 5 gas analyser?. Not many folks do.

Even with a 5 gas, unless you can datalog with it in car, at WOT in the various gears, you're even, justl guessing.

Does a tune need to be considered in it's entirety?, well, dugh, no kidding. Goes right along with giving an engine what it wants, and that goes backs untold years.

Plug reading is a lost art, lots of posers nowadays, but few really get past the obvious. What works on Champions, should not be assumed to be true for other brands. Lots of folks continually get confused on that issue.

BTW, how ya doing on the from scratch code project?. Been months since any reports.

dimented24x7

iTrader: (2)

I dont have any trouble believing that kind of varience, especially with a physical system. I was looking through some stuff on various sensors and some have around a 1% deviation across the operating range from temperature alone.

funstick

Gumpy i think your taking this the wrong way

got some algorythms together im working on code structure. trying to make use of the ford EDIS 8cylinder system since there so much easier to implement in code and physical reality and there readily avaiable. is proving to be challenging though as there isnt a 100% understanding of the hardware but were 70% of the way there.

. i am using an ecm bench,logic analyzers and some other nifty tools for development on the code project but i dont think the average joe really needs that type of acess. this code may go onto to other things but its learning experience for me and my crew i work with.

My point is im trying to kick some folks here including myself out of ruts. we all get in them. some of the agruments im making even i dont truly belive. but im making them in order to hear dissenting opinions. sometimes getting on the opposing team helps you better understand the one your on.

why not air out some other thoery and conjecture. chellenging the established norm is called progress. and yes i agree a 5gas can be a usefull and useless tool. but youd be suprised to know that most of the memebers here are auto techs and you can rent a 5gas from snap-on its a good tool to eval your tune. but again by itself can be misleading.

as i siad at the start. this is a thread were we can let the ideas fly and debate them on merit etc.lets hash over some ideas without getting emotional and try to pick each others idea thoerys and methods apart. then towards the end we can all get together and come up a a synapsys of the whole thread and make it a sticky.

D's89IROCZ

I'd like to see all this stuff debated . I am a noob..... and trying not to get stuck in a rut

dimented24x7

iTrader: (2)

Ive always wondered about the statement that timing can change the apparent AFR on the exaust side. Obviously if your running way too much or too little timing it can since it adversely affects combustion efficiency, but how much just for a varience of +/- one degree of where the engine ideally should be? It seems to me that it should be relatively small since combustion efficiency shouldnt be effected very much. The only other thing that I can think of that would cause the apparent AFR to vary with minor changes in timing would be secondary oxygen consuming reactions. Things like maybe a larger percentage of CO2 formation or the formation of NOx if the mixture is burning hotter.

Probably the only way to actually verify if the apparent AFR based on O2 really changes significantly with timing would be to use a milti-gas analyzer and other tools to see what percentages of each gas are present after the combustion process is complete. funstick, have you ever done any sort of analysis like this?

Grumpy

I'm a newbie myself, only been doing it since about '66, and I don't find myself in a rut.

Ruts are for those that don't ever bother to really think about what they're doing.

Grumpy

In what I've been doing, it takes about 3d to make a noticeble change (with a WB).

There's been considerable controversy at the CARB about what an oem O2 sensor actually measures, and responds too.

An O2 sensor is but one devise, adding EGT, plus plug reading gets you about as close as close can be. Each one's a tuning aid, and the more aids you have the faster you can develope the tune. Getting familiar with what the tuning variables are, and what they do are the other side of the equation.

Grumpy

I think you're so far off base, as to about be laughable, how's that for a reply.

You open up with myths about tuning, and then go on to express opinion as being fact. Then when some doesn't agree with you, you go with *I think you maybe taking this wrong*.


The technics that are in play today, come from decades of hands on tuning, so far you've only gone on to say they're nonsense, and pose nothing of real merit, or fact to support your allegations.

I think you ought to start a thread about one specific item, if you really want to get any in depth discussion rather then tossing out half a dozen, some of which really don't merit a reply.

Triviazing what an ecm bench can do, just shows your lack of understanding what it's really capable of. Unless you just like running down the strip all day, wearing the engine out uselessly.


BTW, I noticed you dodged replying about how the custom code work was going.

JPrevost

Since I've been fortunate to tune on an electric dyno I can tell you that it takes only a couple degrees to move the AFR readings from the wideband. I don't have my papers infront of me but from memory I believe it was 2 degrees retarded from lowest peak torque yielded a .2 afr rich, 4 degrees was .5 rich, and 10 degrees (traction control) was 1 whole afr. Keep in mind this is with a motorcycle engine running 14.5:1 with gt-100. Oh yeah, and 10 degrees retarded took out 20% of our torque, even when we leaned it back out to 13.6afr.
EGT's were very interesting. Retard the timing and the EGT's would go through the roof (as should be expected). Great for turbo's, useless for n/a engines that have the headers less than 2 inchs from the drivers back .
Grumpy is being harsh but I agree with him for the most part. You can't say the average Joe wouldn't benifit from using these tuning aids.... I AM an average Joe or at least I was and I wish I had these toys to play with 4 years ago! It sure beats using a narrowband for WOT tuning so I'd say that tool is needed if you plan on spending the least amount of time and saving money on gas/engine wear.

Grumpy

Not to mention folks seem to continually forget that they peak at stoich.. Either way from Stoich, and they drop.

funstick

i never siad these tunning aids werent helpful. there all helpful. you missed the point i was making. without alot of these tunning aids working together your going to beat yourself up. you cant do that much work to the calibration unless your working with extremly sophisticated modeling software. best you can do is punch it through software like desktop dyno and then plug the ve curve in and make assumptions. and thats not such a bad way to start.

i didnt dodge your question on the code the answer is right there. the big algos like the fuel and spark calcs are working. where we are struggling is controling ignition modules. theres alot more to control the outputs then we had figued that was our ignorance. but we are getting there. so where is your code ? its a great achivement to get $58 to the $60 source. no doubt there. but there is still tons more that we figured it would take less time to code up similar strategy from scratch. and so far 65days in weve made good progress.

other projects are slowing the consortium down. when we get to beta testing ill post it here.

again the point im trying to make debating some of this stuff is that why not challeneg convetion. get folks heading in different directions. maybe the change in afr vs optimum timming are critical there is a great tunning aid !!!

as for documentation on timming affecting observed afr. seems to depend on bore size and cam timming. small bore engines with very mild cam timming seem to be the most resposive to this while large bore engine seem less so. is thee a hard fast rule ?? no not that ive seen but 1-3 degrees can be fiarly citical to
Obseved AFR

as far as i can tell nothing works better for calibration then stepping back and looking at the big pictue.

Grumpy

OK, try explaining what your *point* is. You named the thread *Myths about tuning*, and so far you've not exposed any myths, and seem to be forming your own. See items 2+3

The algos are working, but the ignition isn't. So that would seem to mean, you have some lights flickering on a bench, rather then it actually running in a car. There's a huge leap from the bench to car.

Giving the engine what it wants covers all you've had to say.
Looking at the big picture, is learning what the code actually does, and how it effects the car. Once you combine that with some basic understandings of how an engine works, it's covered.

So what myths are you talking about?......

dimented24x7

iTrader: (2)

Interesting... So IOW the farther you retard from the ideal timing, the more the burn efficiency drops. Was this with MAF, or with MAP based fueling? Ever experiment with going a degree or two over what you normally run?

JPrevost

Good question, I'll see if I've got any of the log files where we were tuning the advance. I believe the afr also dropped when we went too much timing but not by much if any. Oh, and another thing I'll check is the MAP. The traction control tests where we retarded the timing was done with alpha-n with individual runners so that might be a better indicator since it won't move nearly as much as the MAP. Now you've got me curious again . It makes sence though... not at peak timing and the efficiency of the motor drops which is the same way of saying less fuel is being turned into work. We only got 2 evenings worth of data with those tests so it might be inconclusive... maybe not.

funstick

actually the bigger point of the post you keep missing is the narrowness of the way people tune/calibrate. relying on one sensor for all you tunning work is like working in a tunnel.

there not a huge leap from bench to car. we can make output devices ( IAC ) etc work on command. we are having trouble writing the dual purpose code for controlling the est lines. we have acess to an old simucar setup. should alow us to test the hardware and code in the loop first try.

3.8TransAM

Hmmmmmmmmmmmmmm

Some neat points were thrown out here and a lot of common sense style stuff that the avg guy can do and afford, like the timing affecting AFR. Nice flat road and a Romulator with a wideband could tell you a lot. Increase/decrease, what happened and how much.

I started when most of this stuff wasnt available and still find myself not doing things on the fly. It seems to help me more if I evaluate things one at a time and then proceed from there. The on the fly stuff tends to get used more when I am experimenting with one thing at a time or when I am having trouble finding the cause of a driveability issue that doesnt show up in logs.

I love working on the tune of my vehicles and friends, I'm pretty sure I'm not the best one here, or the fastest at it, but I do enjoy it and not try to keep my eyes closed. I was amazed when I got my first wideband(Innovate) and immediately began working on it and trying new things with it from cruising to AFR I played with everything. Now its just another tool. I dont proclaim faith in it, although I'm not running anything really on the edge of crazy/insane combos.

I think the biggest tuning myths are:

#1 Its always the computers fault
(mechanical and electrical must be sound for you waste any time attempting to tune, period, ask anyone here myself included)

#2 Someone can tune your car over the computer/phone
(It is possible if you want to take just short of forever)

#3 More is always better syndrome
(seems really popular with #2 users, where do they get all the fuuel and spark from?)

The other biggest problem I see is anybody taking on facet of tuning and using it and only it to establish a tune. Like only using the stock 02 and not using a wideband to verify or not pulling out a few plugs at a minimum to take a peek.

The more tools you can use and/or have access to the better I think, you can establish your tune and have a reasonable idea that it is indded doing what you intended. Not being a computer junkie per se, everythign I have done in relation to tuning has taken me twice as long as the more established folks on the board.

Even with that I try and keep my mind open and if I find out about something new or something affordable to me that I can use as as aid I investigate the possibilities.

My number one goal? I want a dynojet in my backyard lol I know its not the end all of for driveability, but geez, think about it. Computer controlled load simulation where u can stand there and optimize every point, before the car goes anywhere*drools*
(this is purely a fantasy check and while a real goal is only a dream at this point)then work out the driveability quirks.

Stuff I use/have:

wideband 02
narrowband 02 :-)
TC and TunerPro(limited use)
Diacom and moates
Romulator
my brain
prayer
guys/gals of the prom board

later
Jeremy

RednGold86Z

iTrader: (1)

The BIGGEST problem of "Giving it what it wants" is FIGURING OUT what it wants, hehe. Tools help you do that. Experience (trial and error, a good ear, a good logical head, lots of studying, learning from mistakes) helps you use the tools correctly. Experience also lets you use fewer tools faster, and also tells you to check for problems occasionally, and take a step back occasionally, but full steam it when all is smooth. Experience also tells you to plan your goals, then plan your method of attack. Efficiency becomes key when doing this on a daily basis.

My tools (since we're all laying it out):
2 Mustang Chassis dynos (eddy current only)
1 water brake engine dyno
3 5 gas EGAs with software to convert the 5 emissions to A/F
2 VMAS systems (to measure actual mass emissions)
4 Innovate Widebands or more
Thermocouples all over the place
3 or 4 ECU test benches (slick little buggers I must add)
and my best friends the Tektronixs Oscilloscope
(nevermind access to our source code, DAC (digital to analog converters on the SPI, and high speed datalogging))

Things I like the best:
The FULL ELECTRIC Chassis dyno at Beijing University (you can go from engine OFF at 2000 RPM in any gear to ON to FULL throttle without the RPMs changing more than 100 RPM and this is on a chassis dyno, so no rediculous setup time.
And of course the Innovate. The display allows almost crystal clear understanding of what is getting to the engine in *most* situations that we run accross.
When things go funny, then break out the stethoscope and rubber gloves.

5 gas = check for misfires, or check real emissions. Not much else, other then general troubleshooting if there's problems.

VMAS = emissions on real drive cycles, booh.

Water brake engine dynos = steady state VE and spark, eh, so-so.

Thermocouples = not on anything wild, so it's mostly catalyst testing

Chassis Dynos = now here's where we start putting everything together, if we could only see the A/F...

Innovates = I know, I know. There's more expensive ones out there, but we've compared them to EGA calculated A/F's and it's close enough.

the **** = Oscilloscopes and DAC's. You can WATCH calculations in real time. You can see any memory location updated every... who cares, it's fast (like 5 microseconds).

DAVECS1

iTrader: (2)

Well I am by no means an expert, and I please ask that you take my comments with a grain of salt. I would have to say I have tuned some very interesting combos. I would also say that I have used a couple of different ECMs for the tuning of thes combos. I think the comment stated by RedNGold is right on with the exception you cannot leave the driver out of the equation. I would say that is your most valuable tool. I have tuned engines using a dyno jet and wide band, and have gotten things looking very good with alot of area under the curve, plugs, and temps all looking well, and then we go out for a test drive and the driver will complain that it feels rough when you are loping it around through the drive through or at the car show. I also have complaints about off idle power or revving it up. I went to the trouble of setting up a bench for the 730, but to tell you the truth i do not use it much, am i a bad tuner maybe, but in the end Ihave found it much easier and much quicker to tune the car so it feels satisfactory to the driver, Then I check plugs temps and AFR if they are not way off or in an area that will cause problems for the engine, I do not change it much. I think in the end you will find this to be the case with manufacturers, When driving a mule or a prove design, and something feels wrong, I have rarely driven back to change code. Usually we find a parameter or parameters that have the most effect and change them until the machine feels satisfactory.

My main tool the Butt dyno!
I have others but who cares.

Grumpy

I keep missing?.
Read your own opening post,
YOU SAID:

Wb02's can be useless.
egts can be useless
reading plugs can be useless.
datalogs can be completely useless.

Now, your saying
**relying on one sensor for all you tunning work is like working in a tunnel**

Do you see where YOU'RE just talking in circles?.

Not to mention you totally bogus comment about ecm benches.

L8R

funstick

yes each item can be useless if Use by itself. the only one talking in circles here is you.

no my comment about ecm benchs isnt totaly bogus. theres not really notmuch to understand about the algorythms for the average joe other then what tables after there behavior. this stuff is all extremly straight forward. if other folks here would stop clouding the issue it would be alot easier for everyone to communicate.

DAVECS1

iTrader: (2)

Well I am greatful to anybody who CONTRIBUTES to this board, and I would like to thank Grumpy because he has helped me in the past with tuning my first 730. Some of the tips and trips I gathered from grumpy's post I still use today. Maybe more guidlines than tips. One thing I have found is that there is a difference between book knowledge and practical, but a good tuner can mix it together and produce a good running car. One thing to remember(at least in my opinion) is in the end you are just controling a solenoid, and the resolution of that solenoid is only so much. You just need enough adjustments to deliver the right voltage at the right time, nomatter what it is called in the code or tuning software. Discovering what each parameter does after it is produced can be a challenge and that is were a bench can come in handy. If you change the code so a parameter works right that is were a bench is handy. If you know what parameters do what, than your need for a bench might be little to none. That is my take on the situation.

Grumpy

Bzzt.
Sorry, try again.

You subject was Myths aabout tuning. So far you've not mentioned any.

Yes, it is. If you'd think about it, there's more to a bench then working out code.

Clouding what issue?, you ability to go off on tangents?. You also might try rereading what you''ve said so far.

It still all boils down to making the engine, happy, and the process of doing that means getting as much data as possible, and understanding what you're looking at.

As for the bench just being able to run the code at WOT and see exactly what the PWs are, makes it cost effective, just in wear and tear saved on doing it on bench compared to on an actual car. With gas at $2+ a gal, plus what a secent set of tires cost, not to mention engine wear, it's a no brainer to think spending $75 is a waste of money.

Grumpy

Bingo, at least someone's getting it.
I'm only going to point the way, and give some direction, it's to the individual to get as involved as he wants to.

Yep, it's just about solenoids/ windings.

BTW, if it wasn't for an ecm bench, and the DIY-EFI community, folks would be years further behind then what they are now. It was hammering on 747, that got things out in Public Domain, that got people involved with DIY Prom burning. That and the work of Scot Sealander/ Mike Pitts on the 148 code. They did things from the code understanding end of things. While the 148 was rather specific, it and the popularity of the 747, that got things going.

DAVECS1

iTrader: (2)

Well there is no doubt, that the fact this board exisits to any degree of practicality is work that is far beyond what most people on this board today take for granted. To tell the truth I would not be tuning my car with a 730 and all its bells and whistles if someone had not gone to the trouble to pic apart the code and produce a definition file to look at the binary that is being modified. What needs to be realized here is the resource that is available to everyone, This is not trivial, major OEM's work years to develope the same knowledge network we have on our little forum and at the DYI site. I have been an engineer for years now and the stuff I use to tune my car and modify 730's I probably wouldn't or couldn't of figured out in such a way that makes it as practical as itis today. There are points to this discussion, but it needs to be put on track in order for it to contribute. I understand the frustrations. If observations are to be made let make them positive and straight forward. At the same time we must respect and lend a little leeway to those that have granted us this access to all this information. Companies have gone to great lenghts to protect a lot less knowledge base than what is presented on this board. We must use common sense to keep a great thing.

funstick

glad to see someone else talking about using 5gas, thermo couples, wideband,narrow badn, dynos and road sims to get it right. i agree with 3.8's statement. facet tunning is the problem. you hear people say. well just tune for an afr of 12.8 to one of 14.7:1 or whatever. well what if your running a huge camshaft with serious lift and duration, the overlap below 2500rpm clouds the o2 readings no matter how accurate the sensor from oxygen being spit out during overlap. this is where plug cuts/5gas start to help. im not making the state ment that none of the tools are effective. they can be. when you group them up. grumpy keeps missing the point. ecm benchs are handy do dads. most of the folks here are looking for sound tunning advice. the best advice we can give them is get familiar with the tables in your ecm and how to adjust them and here's how to determine the conditions form a multiude of facets. im suggesting using tools together.

91L98Z28

Random comments in reply to random comments, directed to nobody in particular:

1) Why I am glad I just finished soldering up my ECM bench:

* when I finally convert the 406 in my '71 RS to 730/8D control, I can at least rough in a tune on the bench, verify pulsewidths and reasonable fuel delivery at WOT, timing, etc., and that everything still works with all the bells and whistles turned off (EGR/AIR, no knock sensor, etc, the stuff my 71 doesn't have) instead of turning my 406 into a trail of 406 little pieces behind the car after it "blowed up" because of something stupid that an ECM bench run would have caught. If you know the pulsewidth, and you know the fuel pressure, and you know the injector lb/hr rating, then you know the amount of fuel delivered. that's HUGE. Then, if you're in a situation like me, where you can throw the car on a chassis dyno with a carb and calculate HP and BSFC, well then you can compare that to the fuel delivered at WOT in the first chip, before you even crank the engine over on the '730 for the first time. certainly that's got to help because you will KNOW you are in the ballpark.

* I've been wanting to add shift light code to control a bright LED based shift light off an ECM output. The $100+ I'll save over some commerical unit goes towards a 383 in my '91.

* add code to run the reverse lockout solenoid on my T56

* it's a first step in learning how to reverse engineer the '7748 code in my 88 4cyl fiero which nobody in their right mind cares about enough to reverse engineer; but I want to anyways as both a learning experience and to solve the little crappy things GM did in that code.

Using the ECM bench I can test my code without driving the car. It is TOO EASY to make one little mistake, and end up with drivability issues - ones you may not discover right away. For example, I modified some code once and ended up crashing the ECM when the engine temp went above closed loop threshold. oops. ECM bench would have saved that drive home in limp mode.

2) why closed loop IS important

* gas is 2.60+ a gallon in the peoples republic of california

* catalytic converters are mandatory (i do like driving my car on the street)

* there is no such thing as a perfect tune, and on a daily driver where conditions vary and you have to keep a cat happy, the ONLY WAY is with an o2 sensor. Even the old early 80's cars with cats and carbs used a feedback loop to vary the fuel flow to keep the cats happy!

* for mild daily drivers where it works just fine, why NOT use it?

3) ECM Logic

* If you don't understand it, you can't possibly hope to improve, correct, or add to it. The "why does it do that?" question can't be answered. so you bump the fuel pressure another 3 pounds, crack a beer, and call it a day; because that's the only option you have.

* gm engineers were pretty smart. maybe theres something to learn in there?

4) MAF measuring airflow

* on a stock cam, sure. but it's no holy grail. throw a big cam in that blows some oxygen out the exhaust and causes an unsteady flow of air in the intake, and watch the accuracy go out the window. but now combine the MAF with MAP, and hmm now we got something, we can combine and filter both values and get a more accurate picture. Oh wait, GM did that. now if we understood the logic better, we could toss the opticrap and run 8051 SFI code on an old L98. That'd be nice.

5) timing effect on observed AFR

* this is definately big. especially if you are in getting knock counts. watch what happens when you bounce up against the timing retard, and then lose a few degrees. AFR all over the place. this alone is a huge case for not going ultra agressive with the timing...back off a degree or two.

JPrevost

I'm going in reverse.
Funstick, you're way off base on this one. Closed loop is very important for a catalytic converter. One reason that's obvious is the fact that every automotive manufacturer uses them. If they could they would run the engines in open loop but for emissions reasons they don't. Now if you need more proof well I've got the science. First off, the typical 3-way cat has 2 bricks. A reduction and oxidation catalyst. I believe the reduction lowers NOx and somehow produces O2. The oxidation is after the reduction and it reduces HC and CO by burning... which requires oxygen. This is where closed loop is used as the control. It adjusts the afr to help build up and then break down the crap on the bricks. If it didn't build up and release it wouldn't work as efficiently. The most important part is that the cat gets a lean condition (extra oxygen in the exhaust) to allow the oxidation brick to burn the HC and CO. There, that's what I remember from class and if I'm wrong well by darn-it so are a LOT of SAE members. Another method used to heat up the cat is injecting fresh air post reduction and pre oxidation to aid in the burning but the idea is to keep the engine swinging around stoich. I'll get you the SAE paper I was forced to read... I hated that paper.

As for the tunes done by the auto manufacturers, they do not start open loop. The general routine for an engine that's getting ready for production (all components are made and installed on the dyno) is to use the simulation data that uses BSFC. They then put it through a closed loop routine using very expensive exhaust gas analyzers to minimize the emissions during all conditions. It's amazing how much testing and tuning they go through before production. They no longer take the cars to various parts of the world; instead everything is done in controlled environment test cells. They also do NOT use the production computers during the initial tuning. Instead they have dedicated computers that do the controlling and algorithms which then do self tunes. The only open loop tuning that gets done is when the o2 sensor fails OR they're tuning an engine that will never have to meet emissions requirements (race engines in the C5R for example).

For a happy engine open loop, for a happy environment, closed loop with a 3 way cat using a narrowband o2 sensor or a way to switch rich lean with a wideband (which is better).

dimented24x7

iTrader: (2)

The only problem I hear with that, though, is that the MAF and SD components can really be at odds with eachother when extensive mods to the intake tract and engine. When I was hanging out on those boards while I was doing my MAF project it seemed like there where alot of guys that just wanted to go straight speed density.

That sort of setup makes good sense from a factory standpoint, more control over the engine and redundancy, but from a tuners standpoint it adds extra complexity and variables.

I just went straight MAF for the fueling with some control of the MAF being done based on the MAF. The timing and all other things referenced off of pressure are done by the MAP. Not as precise, but its alot easier to handle overall. The real problem is in the details, especially at very low flowrates. The motor gets kinda flakey below 10-12 g/s with a large maf and only having an 8 bit ADC.

Grumpy

I'm missing the point?, try looking in a mirrror.

As far a do-dads, you've now just labeled yourself a jerk, IMO.

That's the absolute dumbest thing you've ever said, and that's saying a bunch.

You're backpedaling, and mistatements are just laughable.

NO ONE HERE, is suggesting to tune to a specific AFR, (other then in closed loop), other then YOU.

If you want to repeat idiocy, fine, I just hope people remember that's about all your suggesting in this thread.

The only myth being exposed in this thread is your having a grasp on things.

All the tuning aids you can use are a benefit, period. If someone has trouble understanding what they're telling them, then they need to brush up on things.

Again, an ecm bench is a practical asset, wasting fuel, unless you like just wearing out engines, needlessly. If you want to do ANY code work then there necessary.

It's profeciency that makes a tuner a tuner. Knowing how to tell what the engine wants, and how to give what it wants is all it's about. It can be done with just plugs cuts, and a DVM. The more tuning aids you have, the better and faster the job will be, having the tools, also means knowing how to use them.

If you want to continually stick with the idiotic quote of me *missing the point*, I will reply, with blunter, and blunter messages.

Grumpy

Only when done improperly.
If one was to really work at getting things to line up, teh would match, physics is physics, SD or MAF the engine uses x amount of air. If the load calculations are correct, for the two systems, they'll match.

I'd almost bet in the not too distant future even the Afermarkets will start using joint systems, ie both MAP, and MAF....

Grumpy

Whatta shocker.

dimented24x7

iTrader: (2)

For me the problem comes when you have to match it all to a new engine combo. Anyone can tell you here that either system can be a pain to tune on their own. Both systems could be a nightmare.

It just seems like there would be just too many pitfalls to justify mixing the two systems together. For the factory its probably a real boon. But for an amature tuner it just seems like it would be too hard to pick them apart and tune each.

91L98Z28

It doesn't always have to work in one way. For example, let's say some engineering type came up with a way to adapt a modern MAF (something with a greater than 255g/s limit) to input into a '730, and you could log it in the datastream (just like a WB02). All it is, is data that you didn't have before. Any additional data you have, can only contribute to helping you tune. And if you don't like it, you can just delete the column in excel and it's like it never existed.

The more streams of data you have, the more sensors, the more accurate picture you will have of what's really going on.

Example: let's say you are using a WBO2 to to some prelim WOT AFR tuning. you notice that above 5000rpm, to tune for a steady AFR, requires a flat VE curve. Having a MAF sensor in the circuit (to show flow flatlined at some value) will help to confirm what's going on.

like anything, each system has it's good and bad points, and i can see value to taking the positive aspects of both and combining them (like GM did in the 8051/LT1 stuff). If I remember correctly, GM uses one of the sensors as the primary sensor, and the other as a backup/confirmation type sensor to the first, to make sure the whole system is working and being honest.

JPrevost

The Lt1 stuff is (like everything that's new) a step forward. The Ls1 pcm code is really cool, soft rev limiter for neutral... wish I had that. Infact I be a lot of aftermarket ecm's wish they had that feature. Really though, the Lt1 MAF/SD pcm is a great piece of work. I've recently started getting more involved with them in our car club and they are wonderful to use except for the whole flashing part. The data, the large MAF, the SA based on MAP (I love this one), and the redundancy SD maps make it a real treat to look through. I've spend DAYS studing and comparing bins to each other and this is what I enjoy doing because once you see how GM did things and you know the physical differences between each calibration you get smart. I even have spreadsheets that compare the torque curve with VE of the L98 vette TPI, L98 f-body TPI, LT1 vette, LT1 f-body, LT1 imp SS to each other. It's rather amazing once you do all this work to sit back and see the trends.

dimented24x7

iTrader: (2)

I have that It also controls my fueling, just for yucks. The LS1 MAF and its code I wrote to use it replaced the SD fueling algorithms in my 8063. The MAP now handles timing, transient fueling (AE), and other pressure referenced things.

Theres no question that the new PCM code is great stuff. Get in an LS1 and you know as soon as you turn the key and drive it. For me, though, I find too much control to be overwhelming. I just like to have what I need to get the motor to run the way I want. The tach Ive taken now is to start out with the bare basics and add stuff as needed. Just my opinion. Some people like more, some like less...

Lucid

iTrader: (4)

Hiya Grumpy
I know this is a little of topic, but i was wondering if you could clue me in on some plug reading. I started off using those standard AC Delco plugs and had to change them very often because my engine eats them up, then started using the Bosch Platinums. I've seen plugs with worn electrodes, black oily plugs, black sooty plugs but the newer Platinums had deposits that i had never seen before. These things had white thick deposits like fetta or parmesian cheese on them. A few had such deposit buildup between the electrode and the insulator that there was no gap, just filler. I don't know how my engine was running at that point, maybe it was arching around it or something. Could you give me a rundown of what I problems I should be addressing when i see deposits and such. I'm planning on getting a 50,000 volt coil and spiro core wires to help burn the AF mixture better. I'm also looking to get a PROM burner too, because i have modified the exhaust, camshaft and a couple other things and i need some better driveability. You the man grumpy

Thanks

funstick

dude ive always tried to be cordial to you but you've just gone to far here. you wanna talk about calibrating engines lets talk shall we. How many cars you done ? how many customers you got ? you do calibration work for the oem's ? I do some consultation work for one tier1 on there aftermarket program on occasion. i think you need to reel yourself in. Its your assumptions that are the problem.This constant blathering about widebands. ive read the stuff you posted on water injection and its BS. water injection doesnt drop IAT's. your seing water wicking on the sensor im sure with a long duration of water injection use for water to accumulate youll actually see the temp go back up once you reach thermal entropy between the incoming air charge and the water accumlating on the sensor.

No catalytic converters dont give a crap weather the ecm is openloop or closed loop. they only care about CO,CO2,HC, and NOX in the combination of gasses that arrive there for catalyzation into H20 and Co and Co2. the assumption that running closed loop maintians an AFR of 14.7:1 is also BS.

As for myths there pretty clear. dont rely on one input for all you data. dont think more is less and less is more think whats achieavable with what you have.and go for it.

As for the combination MAf/Sd systems. youll find they arent so crazy as people might think. they work exceeding well. but the sd portion isnt there for alot of the reasons people think. its there for 3 reasons that we can air out.

1. its easy to spot a vacum leak.
2. its a great fall back for a MAF failure
3. It does well at transents unless you have large surge volume.

heres a few unoffical reasons.

1 you can have the ecm detect a larger camshaft
2 you can detect a warranty vioding nitrous backfire


it was done more for warranty and emission failure performance more then for AE control. that just happened to be a side benefit.

Along with myths heres a great one. are EGT's are super critical.

we could look at some evidence.

advanced ignition timming lower EGT
retarded igntion raises egt.

but what effect does this have on the cylinder tempature itself ?

we also know that retarded ignition timming makes things appear lean and advanced timming makes things appear rich.

are these to things correlated ? maybe

is it becuase zirconium reacts to heat ?

hmm maybe.

this comes back down to total air mass and what is it. we can didle timming all day long and fuel as well and wind up with substianttly reducded output. but have excelent egt's and wideband 02. or we cna go the other side and be up on power and still have excelent egts and wb02. and we can go up in power not knock look lean run kewl and still make even more power. or we could lose power. all from just a wb02 and a egt

and lets not forget our good frend humididty who can actually demand changes in spark advance as much as 5degrees or more to maintian consistent performance from one day to the next and still not affect VE.

you wanna critisize me fine go ahead. i doubt grumpy has a boosted car with output pushing 3 times the designed output that can run from detriot to miami at 100mph plus climbing hills and spending 60% of the drive in 15psi or more of boost with no intercooler on 87 octane fuel and do it for 100k. fact is while your busy preaching again about stuff you dont practice but on your narrow apllication im busy working everydaying in the midst of calibration learning and trying to distill that knowledge back down to something everybody can use.and i share it freely when people ask.

For the most part this board not due to most of its users has become a complete waste of time. becuase everytime i can, not only prove a thoery but back it up with data somejackass has to come along and pronounce his tunnig glory.

this thread was started as a way to rile things up. shake us down. id like a nice long debate but some folks cant help but make things personal that shouldnt be. its a lack of maturity on there part. i cant do anything but offer my observations.

Grumpy

Got anything to support that claim?.

LOL, I do?, you're the one talking thur his hat. You might spend a few more decades working in the field before your go around bragging about what you think you know.

You're stating as fact WI doesn't drop MATs?. OK, thanks for making that statement, as it just further discredits you. Can there be *some wicking*, yep, it's possible, tho some things are alot less likely then others. Remember you've only seen some of the logs. You may have noticed, that I point the way, rather then make claims of the absolute. Might try thinking about that for a while.

Blathering about WB's?, you really need to get a grip on what you're talking about.

It's always fun watching others brag, you're really starting to sound like an ex-board member, he too was an expert with alot of customers.