we spend a lot of time talking about tuning. an idea i have thought about a lot for a fair amount of time is the automatic tuning fuel injector controller. first, lets assume that we have a decent aftermarket computer controlling everything as usually setup in a single conversion. it controls the fans, water injection, timing, etc...



now, we disconnect the fuel injector leads. we get a small homebuilt computer and a high quality wideband lambda and heated wideband o2 sensor and hook up the output from the wideband to the home made computer. (assume that the computer has a very very fast clock and is very quick at processing instructions.) the computer has outputs that are connected to the fuel injectors and other oputputs that are connected to the water injection system. further, it has inputs for rpm and oil and water temps.



now, the simplest version of the concept implementation. we write a very simple program that reads the input from the wideband and compares it to a value that we have set for the air fuel ratio. say its 11.5:1. if the wideband reading is 12:1, then the computer program will try to increase the richness by increasing the duty cycles of the injectors. it increases the duty cycle a bit and checks the wideband AF reading again. (the time involved would have to be played wioth a bit as to how often and by how much the duty cycle adjustments are made, since the injectors are mechanical and there will be a lag between the increase in duty cycles and the change in AF due to the increase.)



now, assuming that we can accomplish this basic, automatic adjustment of AF's, there is more that can be done. we assumed that we had one AF value set. in fact, we know that at the lowest rpm, well below the torque peak, we can have a higher AFR. we know that we should have the lowest AFR as we get to and go jhust over the torque peak and can then bring the AFR's up again as we go to redline. now, doing this with a standard map based ecu can be done, but its very hard and very time-consuming.



further, these ecu's have static maps in the first place. they don't adequately adjust for changes in barometric pressure or temperature at all well. rather, they have a map of duty cycle to boost by rpm. the duty cycles are set to result in a certain AFR result under the speciufic conditions they were tested in. if temps and baro change or anything else changes, the AFR's will not be they same for the same maps.



the solution i am suggestiong is nbot all that difficult. i suspect that we have guys on the forum who could program it and build the box. ideally, we would combine the wideband and the fuel control computer into one box. this is an idea worth pursuing,. because the result lets us play with targeted AFR mapos and not have to worry about tuning.



gordon

i thought i read the AEM can do this ? i bought my own wideband so still trying to figure what cool tricks it can do hooked up to the haltech https://www.nopistons.com/forums/pub...DIR#>/wink.png

thats sounds like a really good idea. Has no one done this before?

Alot of people use the Haltech datalogging to see what the A/F ratios are with a wideband. That can help out alot. The TEC II had a feature similar to what Gordon was talking about.

i think that the TECII had a feature for adjusting AFRs to maintain emmissions at 0-3000 rpm? i didn't know the AEM had such a thing. i looked at their software a while ago, but it wasn't there at that time. can somebody point me to the info on the AEM that does this? if i can find a computer that already does it, i won't need to build it and NO MAPS EVER AGAIN!!! -gordon

From the Wintec 2 manual, and observable when using the software:

no maps sounds like a good idea, all those tuners might get mad https://www.nopistons.com/forums/pub...#>/biggrin.png

that would be really cool. a friend of mine (www.16paws.com) has been looking at cracking the stock ecus. he is just doing it for fun, but even a stock 86na ecu has more features than any of the aftermarket ones, and it will start the car hot and cold.

on the other hand if we take gordons idea we will have one table for fuel, it would be afr vs load vs rpms and a table for timing, maybe using knock sensors? or the factory uses pressure sensors to detect knock, maybe one in each rotor housing? that way you have 2 tables, the fuel and the timing. for the timing you set the max cumbustion pressure.

hmmmmm



mike

what? Really? why the hell do standalones cost so damn much then?

Here is the way the AEM works from my understanding. You have to tune it one time so that it has a starting point. Then from that point on i THINK you can just tell it what a/f you want and it makes adjustments to get that.



I believe its based on the Speed Pro system which also auto tunes, AEM bought Speed Pro and couple years or so ago. We never hear about Speed Pro cause it didnt work with rotary but the AEM of course does.



STEPHEN

what? Really? why the hell do standalones cost so damn much then?[/quote]

They need to make money too. Try making your own ECU and see how much you'd have to sell it for to make a profit.

the electromotive kinda works like that too.

and yes the 86na ecu, has squential staged injection, it can compensate for differing loads, baro pressures etc etc.

the only 2 problems with the stock ecu are, it uses an afm (that isnt compatible with anything cool) and you cant adjust it for big mods/injectors



mike

i would have never thought it adjusted for baro pressures, that is pretty sophisticated. Stock ecus are always so frowned upon, i just assummed they were **** to begin with.

on the topic of the AEM, so that basically is what u would need right there right? Why don't the rest of these companies have that ability- to tune once and then set A/F ratio. too bad that aem uses the 15+ old wiring harness.

Hey Gordon.



My Autronic can do this, when you plug in one of their A/F meters : )



It's called AUTO TUNE, you simply punch in the A/F you want across the load/speed map from idel to max revs from vacuum to max boost and the computer does the rest !



Regards



Pete : )



P.S. Thnax for your e-mail, sorry I did not get back to you since then.

Welcome aboard Peter nice to see you here...

Welcome aboard Peter nice to see you here...[/quote]

yes, i second that



mike

I 3rd it!!!! lol



Great to see some more true gear heads coming over!!!!



STEPHEN

whats autronic? another standalone ems?

hey, peter! great to see you over here. far less bullshit (so far) and more acceptable sex... anyway, where can i find the info on the autronic? it actually tunes to a desired AFR across everything or how? the ideal system would let you sdo what i said. tune to richer afr's where there were problems and then lean it out as you went to higher rpm. also, i'm hearing that AEM and TECIII do this. i've got to check their sites. i'm working way too hard at this stuff, if there are boxes that do it... -gordon

i have a tec III sitting on my floor. i will not be reading the manual though

My version of the Microtech MTX 12 will do closed loop with a Mazda O2 sensor. It is (as on factory ECU's I believe) only for steady throttle conditions. It only works when your AR is not too far wrong to begin with, ie 12-15, otherwise it is outside the lookup table & does nothing. This is an advantage in one way in that I have a lean cruise mode set where I can roll along on light throttle and run at 16-17 AFR with big ign advance and get good mileage without it trying to trim the mixture. Same for WOT you want to be low 12 AFR so a certain MAP & TPS reading take the ECU out of closed loop.

You can watch it happen on the display when MAP & TPS are steady. The pulsewidths vary & AFR moves in 10ths.



Main downside is the lookup table can not be altered and so can not be calibrated. I got around this by calibrating the voltage from the O2 sensor so that the ECU readout was within a 10th of the reading from a big dollar sniffer

Robbomaz, the Microtech along with others will do closed loop for normal conditions, it will not have any input in WOT full load conditions, almost all ECU's resort to the pre programmed map if they are non air flow metered cars.



The Autronic with AUTO TUNE Wide band interface will program itself ! I think maybe MOTEC offer this as well ????



Gordon, type in Autronic ina search engine and you will find all the info under the official reseller, I will go look for it now (am at work, so no guarrantee )



But yeah I am with you, why re invent the wheel when it's already been done, Lots of pros in Aust use this Auto tune and swear by it, I will be using it on my PP turbo engine soon : )

Motec will make you a Cappucino on your drive to work........!

I didn't know thw Autronic could do that. Must have a 3D map that brings in load & throttle pos as well? Wouldn't want it trimming back to 14.7 as you blast past the apex at 15psi & the top of third gear......!

http://www.autronic.com/autotune.html



Reads up mighty impressively! Anyone used it IRL?

Welcome to Nopistons.com Rice Racing!

gordon this is already done alot. it's called closed loop tuning. this is not a very good idea as everything would rely on your O2 sensor. if you get oil on it, foul it up, electrical glitch, temp fluctuation, basically anything that could give it a wrong reading you're looking at a rebuild. now what you can do is a semi-closed loop mode, where you do some baseline tuning to get afr's and inj. duty cycles. then you just input the afr you want and the computer will spit back a complete fuel map to reach that goal. you just look over the fuel map to check if everything looks ok, hit apply and it'll automatically apply those duty cycle numbers. the dta ems i'll be using for my drag car has this function and it's much safer than a full closed loop mode.

closed loop tuning is common on virtually all stock ecu's and several aftermarket ecu's. however, their purpose is to maintain proper exhaust emmissions at lower rpm. they usually operate from 0-3000 rpm.



the system i'm talking about wqould of course require a wideband o2 sensor and would read that and target a preset a/f ratio, which wouyld require an onboard, integrated lambda gauge to instantly convert o2 into a/f and feed that back into the processor for correction. if i am correct, the closed loop systems do not do this at all, though someone can set me straight here if i'm wrong. the closed loop systems only are concerned with emmissions. i am pretty sure they are not very fast, either, since they are intended to check on emmissions and this is not a critical reading relative to detonation, but only to whether the car will pass emmissions or not. i know that peter farrell used to use the closed loop monitoring from the electromotive as a simple way of tuning for a/f broadly. he started using wideband lambda meters later on.



guys, if the technology and approach i am suggesting here were available, there would not be any tuners, since we would just hook up our aftermarket ecu's and set the a/f ratio we wanted and drive away...



-gordon

Electromotive and Autronic HAVE autotune, but it will still require knowing what A/F ratios you want, and you need to put in your own ignition values.

closed loop tuning or autotuning w/ a wideband is available on almost all new ems's. the problem is what happens when something causes your 02 sensor to have an inaccurate reading. if your fuel maps are completely based on what your 02 sensor is reading your gonna be in big trouble. even the smallest thing like the sensor not being heated up properly or oil getting on the sensor will cause it to give bad readings.

Most of the auto tuning ecu I believe just build a base map off your wideband, then after its complete your just running on that. I'm pretty sure thats how most of them are, I know the aem is like that. You hook up your wideband and the ecu uses that to create the map when ever you request it



STEPHEN

yeah that's what i was saying earlier, it's best to use sem-closed loop tuning to get a basemap to run off of. but what gordon's saying is he want's the 02 sensor to be constantly tuning a/f ratios at all times.