crush02342002

iTrader: (2)

now thats an intresting question

paul_dezod

The ECU is ALWAYS changing timing dynamically and seldom reverts back to the exact point from one pull to the next. It's always trying to test the det threshold.

One pull you may see a final advance of 18*, while the next, you may see 20*, then one of 19*......

AFRs seem consistent to you for various reasons....

1) the OE primary O2's refresh rate is about 20 times faster than any aftermarket wideband O2. There are micro-send flash-over points where AFRs dart quickly and changes are made at lightning speed by the stock ECU.

2) you do not have datalogging capability to see AFR fluxes in 250 bits/sec of data. When you can see 250 snap shots in a second, your perspective changes quite a bit on AFR consistency. Then the realization of how good that stock primary O2 really is and well it works sets in.

3) Average aftermarket wideband screen you are viewing refreshes at about 1/4-1/2 the rate of the stock ECU otherwise, your wideband would look cooky on the display because you might get a 14.7, 14.5, 15.1, 13.8, 14.6, 14.1 all within 1 second. Instead AEM, Innovate etc all slow down that sample rate and give a 14.7, 15.1, 14.6 read.

ScionFred

iTrader: (1)

^^^ Thank you Paul! That's great info. In addition to reading my AEM UEGO I do datalog the oem AFR sensor with Palmer ScanXL Pro and my ELM327 V1.3 OBDII interface module. Unfortunately I don't know the actual PID sample rate possible through the 38,400 baud rate ELM327 chip. I have set the AFR sensor sample frequency as high as possible and limited the number of logged PIDs but again, I have no idea how many times the AFR sensor is read per second.

This reminds me that I need to get a faster OBDII to USB2.0 interface than the ELM327. Any suggestions?

Edit: I've done a little searching and thus far it appears that OBDPros or OBDKey are among the fastest interfaces offering 1.25MBs throughput over USB 2.0 connections. It's hard to find good info on this.

thendawg

iTrader: (2)

Yeaaah - thats the one thing thats wanting to make me go to a standalone - the ecu's dynamic timing adjustment. Its really hard to get E85 to detonate, sooo the ECU sometimes goes a lil crazy adding timing. One day I did a pull where my timing approached 19deg at redline (about where I want it), than the next pull the damn thing advanced it to 22deg lol

crush02342002

iTrader: (2)

great info guys

ScionFred

iTrader: (1)

Soooo, back to my original question... why effectively disable the oem AFR sensor by entering 3.2v in the FIC's O2 map in low load vacuum areas that don't actually require any AFR enrichment or signal alteration?

This question was prompted by seeing a finished tune from a very well known and well respected west coast tuner. I can see doing this temporarily to make tuning the fuel map a little easier but after that the O2 map should be zeroed out below boost so the ECU and AFR sensor can do what they do very, very well. That is, maintain a perfect stoich AFR under any conditions.

If I'm wrong or missing something here, please let me know.

crush02342002

iTrader: (2)

one reason I can tell you is someone may want their idle fuel to run a tad rich to make the engine run a little more smoothly (overall feel).
Or perhaps a customer does not want his car to run rich when the a/c is turned on and at idle, so you would lean the car out a bit. Id lean it out to 15.4 at idle a/c off so that when the a/c is switched on it will richen the idle area to roughly 14.4 (all depending on how large the injectors are). This can also be helpful for those that have stalling issues with the a/c on.
The fic will not nullify the o2 sig in voltage mode, if it could then fuel trims would not change. after all its a tap and not a splice which leaves me to beleave that its impossible.

crush02342002

iTrader: (2)

Also if the fic could disable the primary o2 sensor then it would be reasonable to think that one could simply cut the o2 sig ( not the heater circuit ) and wire the fic o2 lead directly without a problem. I have not tried that but iv seen evidence that it wouldn't work, I may try it out tonight. Besides iv got a new fic harness on the way... Lol

ScionFred

iTrader: (1)

As I understand it, if you place 3.2v in the O2 map the FIC will modify whatever the O2 reads to 3.2v. The O2 sensor could be reading rich (2.5v) or lean (4v) but the FIC will attempt to modify the signal to 3.2v. That's why I said it effectively disables the O2 sensor. It doesn't actually disable or nullify it but by altering the true signal it is sending inaccurate info to the ECU which still affects fuel trims but perhaps not in the correct direction or amount.

Now if you use percent mode, you can specify a percentage to add or subtract from the signal. That seems better suited to accomplish what you were suggesting. Although my AFR is a constant 14.7 with AC on or off. I tried the offset mode but it's apparently setup only for NB sensors since the max offset is only ~0.6v.

In short what I'm suggesting is to let the ECU and AFR sensor do their jobs and tune the CL AFR until AFR enrichment is needed. Then and only then should we start skewing the O2 signal to trick the ECU into adding extra fuel. Nothing ground breaking but a minor epiphany I had recently when asked to look at fellow turbo XB bro's 'pro' tune that was causing lean, hunting idle, fluctuating AFRs, stumbling and frequent stalling on decel. Simply changing all the '3.2's in his voltage based O2 map to zeros has made "an extreme improvement" according to him.


BTW, the FIC can't replicate the AFR sensor signal but by tapping in it is able to alter it by adding or subtracting current.

http://www.autoshop101.com/forms/h37.pdf


...

crush02342002

iTrader: (2)

that will work fred to a point (letting the ecu take care of all the work), that is till you go even larger in fuel injectors. If you were to unplug your primary o2 sensor and take note of your afr with and without a/c on you will start to see the bigger picture. As of right now I have a full point difference in afr with and without a/c. 1 point can equal a fuel trim of -8 which isnt much by itself but coupled with other factors and your fuel trims skyrocket. I have viewed on 550cc injectors at 70psi fuel pressure an extra -8 ltft at idle with the a/c on ontop of a -3. so my ltft was sitting at -11, this is what the ecu has learned over the course of a few weeks. So lets just say your ltft is sitting at -8 to start with and having a similar setup as me, you turn the a/c on and now your sitting on -16. at -20 ltft (in my experiance) you will start to see problems. little extra throttle lag, stalling/stumbling, cel, and so forth. so you see in that situation you dont have alot of wiggle room.

unltimately its all on user preference, what you feel comfortable with. doing the above has served me well but some may not feel that its the best way. Find what works best for you Fred and stick to it, iv listened to others and tried their methods but felt those methods came up short.

btw i didnt get a chance to try out what i mentioned earlier but i figured it wouldnt work.

crush02342002

iTrader: (2)

iv tried the percentage mode only twice...didnt do it again...lol... but it is worth a revisit.

ScionFred

iTrader: (1)

Maybe I'm missing something but I still don't understand how skewing the AFR sensor signal helps with anything other than changing the AFR. All fuel tuning is still done with the fuel map. So unless you want something other than stoich, skewing the AFR signal is pointless. All fuel trims will remain the same as long as the AFR remains the same.

But as you say, whatever works for you. I stand by my theory but defer to your far greater experience tuning the FIC.

crush02342002

iTrader: (2)

thats exactly the point fred, changing the afr. or in more detail changing the fuel trim. use anymethod possible to correct fuel trims.

ScionFred

iTrader: (1)

Why do you say that fuel trims are more important than AFR? I'd say to use any method possible to acheive the desired AFR. A LTFT of -18% is nothing more than the ECU automatically subtracting 18% from the fuel table base value it would otherwise use. It's really the same as editing the fuel table as long as you don't exceed it's limit of +/- 20% LTFT.

crush02342002

iTrader: (2)

I didn't mean one is more important than the other. Just ment both should be looked at and kept in check for the best tune possible, even if your tuning consists of mechanical means.

crush02342002

iTrader: (2)

Also Fred if some one can keep their trims at a low and steady state it's easy to spot o2 or maf problems.
Btw iv thrown a cel at less than 18 ft. That's another aspect of the game, keeping the check engine light off. Plenty of reasons for that

ScionFred

iTrader: (1)

I still get an occasional P0101 for MAF out of range about once every 5k miles but that's it. FWIW I opened an old log file taken with 440cc inj's installed but no turbo, just a header, stock airbox and exhaust. The ECU had a few weeks to adjust trims before this log was taken and the car ran perfect at this point with no CELs. I'll add a new log file soon to illustrate how the LTFTs have normalized considerably over time. This relates to my previous thread question about longer term ECU memory.

I think the important thing to note here is how close the STFTs are despite up to -25% LTFTs. The STFTs indicate how close the ECU adapted LTFTs are and therefore there were no AFR or performance issues present.

crush02342002

iTrader: (2)

I gotta tell ya im amazed you had an ltft of -25 and didnt notice any problems. a trim like that always has indicated a problem and showed up performance wise with me, be it an o2, maf, or tuning issue. That is very intresting fred. though it looks like in idle you werent that high. was your timing adv at +6-+10 for idle?

ScionFred

iTrader: (1)

I think the biggest difference between your observations and my own is time. Since you're tuning a piggyback to match the existing ECU fuel trim condition, anything more than small variations will cause immediate problems. In the scenerio I just submitted, the ECU had weeks to adjust to the 440cc injs. The first day with them installed wasn't pretty. I experienced many of the problems you'd expect with large fuel trims; hunting idle, bogging, and even stalling on day 1 with the 440s. Day 2 was much better but I still got a rich CEL and some bogging when starting off. I wish I had kept a log but as best as I can recall 2 years later it took a hundred miles or so for the ECU to adapt to the 440's and after that the tune just kept getting better and better as the ECU continued to adjust to the 440s under different load conditions.

I think the key element in the data log conditions I submitted was that the LTFTs were accurate enough that the STFTs were near zero. Since the LTFT establishes a new starting point for fuel inj pulse duration, if it's accurate, there should be no driveability issues as you can see from the small STFTs. However if the LTFT is excessive and inaccurate, you'll have driveability issues that can be seen by large fluctuating STFTs.

The only relevance I see between this and piggyback tuning is that while overriding the AFR sensor in low load closed loop may produce better immediate results, I question whether this approach is best in the longer term. Perhaps it might be better to concentrate more on the fuel map and let the ECU fine-tune the trims with accurate AFR sensor data.

BTW, I'm still looking for some closely comparable data logs to compare my fuel trims from the August 2009 results I just posted to the present day. So far it appears that my max LTFT today is -18.8 vs -25 back then. However I can't definitively say whether the difference is ECU adjustment or other factors.

crush02342002

iTrader: (2)

I agree with you on that fred, after enough time and miles has passed the ecu will stop "freaking out" on you even though ltfts are off. It seems to be more "shock" to the ecu when fuel suddenly changes out of its normal range. Hence why i suggested o2 skewing and fuel corrections to meet the needs of the load thats placed with the a/c on. Idk about you but I switch my a/c on and off alot in the fall and spring times. As funny as it sounds this sudden change in fueling demand can be enough to "freak out" the ecu given the injectors are large enough to do so. even though this difference would show up first as a stft it will turn into ltft, so the ecu will learn what it needs to do to keep the system copasetic. However there isnt a set time on when the ecu will create a ltft, iv seen the ecu almost immediately create ltfts (normaly when things are pretty far out of wack) to a few weeks. In that time the ecu may not learn a correct trim for the added fuel with the a/c on. The ecu will automatically open the injectors a percentage longer which I beleave is a set value but I dont think the ecu will create a ltft that is activated (pulled from memory) that satisfies the specific situation on demand. Which would mean that the percentage may not change but the overall fuel output per cycle will grow as the fuel injectors are swapped for larger injectors. The ecu will know when the a/c is switched on by prolly amp draw/and or voltage and sudden change in rpms. This idea could be just a wash, it could just be the fact that there is no fuel added to the system with the a/c on but could just be the sheer load placed on the engine that changes the fuel demand and the o2 sensor compensates accordingly. Either way fuel demand does change suddenly. Imho its best to work with a duel map setup and use a relay to switch between the two thats activated by the a/c switch. As always there is more than one way to skin a cat.

This is just one example.