DaWorstPlaya

iTrader: (31)

Hi Guys,
I've been following these 2 threads with great interest. This one is about the new feature in the Xede that enables "active tuning" ...
https://www.evolutionm.net/forums/sh...hreadid=142221
This other one is about the 05 ECUs and possibly newer 04 ECUs...
https://www.evolutionm.net/forums/sh...d.php?t=143847

Its seems to me that the first thread talks about a new program on the Xede that agressively and actively manages AFR, constantly making changes to boost and fuel to meet a certain AFR at a certain given RPM. Simulatenously the Xede manages timing constantly, advancing it to the maximum available amount to give the maximum available power.
So everything is trying to meet a certain AFR constantly, thus MAF and MAP signals become secondary, sort like a relative variable with a reference point on the fuel, boost, timing, etc maps that are loaded on the Xede. Thus BOVs and Intakes are not an issue anymore because the amount of fuel injected is not dependent mainly on the MAF signal anymore but the AFR. This is true for the drastic weather changes we experience in the US as well. Too rich and it will immediately start leaning it out ... to lean and it will richen it up ... clever little program ...
BUT, I've also stumbled accross the 2nd thread ... that goes on to say that people are not experiencing the same stumbling issues, they did with BOVs on 03s, on the newer 05s ... So that to me says that has changed parts of the code in the 05 ECUs, where by, now it is more dynamic and will try and adjust to varying readings from both the MAF and MAP.
It seems like originally the coding in the ECU was more dependant on the MAF signal but now uses both MAF and MAP to adjust fuel accordingly thus avoiding the stumbling issue. Now we all know that we have 2 O2 sensors on the car that monitor EGTs and thus extrapolate AFRs and catalytic converter efficiency and the ECU uses the first one to figure out what AFR it needs to be at, given the map that is loaded.
We also know the ECU is a very complexly programmed device that is constantly learning and adjusting to various conditions, using signals from various sensors. It seems like, while the 1st gen of ECUs were more strict in following the map loaded and depending only on the MAF signal, the current gen of ECUs are a little more flexible .... I hope some of you geeks are starting to catching my drift ... I know that the ECU advances timing but cautiously after it experiences knock activity and it relies heavily on the map loaded to balance out AFRs depending on boost, rpm, load, etc .... but now that the ECU is more effectively using the MAF and MAP to figure out fuel, one should be able to use aftermarket intakes and not run into lean issues (theorically)... also, and here is the kicker, with a more aggressive map loaded on your ECU and with the ECU able to correctly figure out the extact amount of air mass coming into the engine, it will be able to compensate with the correct amount of fuel to hit the target AFR loaded on the map.
So, wouldn't this mean that the new active tuning system on Xede for the newer 05 ECUs would be pointless? after all your ECU can actively tune for the map that is loaded? Not only that but it seems that one system is using AFR as the major reference point and the other is using air mass as the major reference point ...
So, to me that would seems that this active tune on the Xede is only good for people with 03-04 ECUs ....
But as always people are quick to jump on a bandwagon and are trying scoop up Xedes, trying to get the latest and greatest ...

Thoughts?

edited: for spelling mistakes ... also broken into more paragraph for easier comprehension ...

hooptee

iTrader: (4)

Thanks for posting this. I pretty much gathered the same conclusion after following the same threads with extreme interest.

Nez136

iTrader: (10)

very interesting point

BlackEVO

I am also subscribed to the Xede auto-tune thread. The fueling is nothing thing new...at least on a standalone system and is pretty much like the boost control on a closed-loop UTEC. You enter the AFR you want to be at and it does what needs to be done to get there.

It will help, but you will still need a map that is relatively close for your mods as it will only adjust so far. In other words, don't expect to be able to just keep adding mods and think it's going to tune for best performance no matter what you do, *especially* with an intake that can drastically alter MAS readings and thus, fueling. Presumably, if you have no clue what you are doing, you can simply look at the adjustments autotune is trying to make and keep tuning the base map in that direction...not unlike tuning part-throttle fueling by watching the short and long-term fuel trims that the ECU makes.

The idea is kick *** and if it works well, I may switch management systems once again. It will *fine tune* and save you from having to tune to the edge (that extra 1° advance) and then worry whether that will blow your engine on a colder/warmer day, change in fuel, etc. I know it's prelim and from what I read, may max at a +2-3° additional advance, but that most certainly could translate to a substantial TQ/HP gain. I've seen 20WTQ gain (peak) with a 2-3° advance in the midrange. Putting WOT driving aside, that on-the-fly extra advance could do wonders for throttle response too.

I would be a little more cautious with your assumptions. The "MAP" (MDP) sensors have never done anything other than monitor for the EGR system and they have been on 4G63s for a long time. While there may have been code changes or recalibrations, to assume after 10 years they simply started using it for part-time fuel management between '04 and '05 seems like a stretch to me. If they started using the MDP for anything at all management-wise, they woud have had to come up with alternate calculations and crossovers points to switch between MAF and MAP readings. That costs $$$ and is utterly redundant (2 methods to determine air).

O2 sensors do not monitor EGTs, they monitor the oxygen content.

As for the ECU and knock, it's not "that" cautious. It can retard pretty heavily on severe knock, but to drop to a lower table and stay there, it would have to knock repeatedly and often. In other words, it is a learned behavoir by the ECU. If you knock once, it's not like the ECU automatically drops to the lower timing table and stays there until you prove to it over time that you are no longer knocking. I could have knock at peak boost with a decent retard and still see 21-23° by redline in the same gear/pull.

What I am waiting to see is what the autotune does with timing between 5000-5500 because that has always been my biggest sore spot.

The MAF incorrectly determines the mass then? MAS/MAF = *Mass* Airflow Sensor...LOL. MAF is inherently more accurate than MAP or Speed Density because it actually measures the incoming air. The rest of the sentence is pretty much what an ECU in any car does as it's full-time job (closed loop), so what is the kicker?

No, it wouldn't be pointless. Yes, all ECUs "actively tune" any time the car is running....they are constantly monitoring and adjusting. I think the point you are missing is that Xede w/auto-tune will instead tune to user defined AFRs instead of looking up fueling tables in the stock ECU and it will add extra timing over the ECU tables whenever possible. Also, one is not mass-based and the other AFR-based. Air flow measurements are always the basis whether it's MAS/MAF, MAP, VAF, Speed Density, etc. Figure the air, determine the fuel needed based on load, IAT, etc.

I guess in short, the autotune is doing exactly what the stock ECU does on a day-to-day basis, it's just 1) targeting fueling that the user specifies instead of the ECU and 2) adding some extra timing above what the ECU would give when possible. It only works up to a point at that, so it's not a plug-it-in-and-mod-your-****-off, it-will-tune-everything-for-max-performance deal....you'll need a base map that's close to correct and it will fine tune. Don't get me wrong, that "fine tuning" is the difference between good performance and great performance. It could be worth it's weight in gold, which is about what it will cost for the full setup. It's easily in the full-standalone price range after you get a flash to go with it to remove the fuel cut and rev limiter...they (Xede w/autotune and standalones) are just two totally different animals...

SuperHatch

iTrader: (23)

Excellent responce BlackEvo, I couldn't have said any of it better myself.

- Steve

gnulooks

iTrader: (1)

I started reading the first post, but got lost somewhere in the middle. It would be nice if you could break up your long post in paragraphs to make it easier on the reader. Off topic, but I thought I should address this issue...

DaWorstPlaya

iTrader: (31)

I agree with you BlackEvo, I don't believe you can continue to mod away and expect the tune to keep compensating for all the mods. Obviously there is a limit ....
1° advance producing 15-20 Torque is a lot but then again that is only where the ECU has retarded timing because of knock activity within that rpm range. If you were to look at a dynosheet, it would be where there was a small sudden dip in power, correct? I mean if you are accelerating through gears quickly the impact that would make would be very little ... I mean you would pass through that rpm range within a blink of any eye, especially in the lower gears .... This would probably have a greater impact in the higher gears?
So, from what I have learned its better to run slightly richer and run max timing advance (that gives you most power) vs to run slightly leaner and not run the same amount of advance. The second, senario gives you less power? Vishnu is known to design their maps to be leaner than other tuners, correct? from what I have seen in the past, his dyno-graphs are not as smooth as some other tuners ... is that a problem?
Yes, O2 sensors don't measure EGT. Sorry I meant to say that they function after the EGT warms the sensor up to a certain operating range ....
Well if the ECU is not using both the MAF and MAP and thus not causing the car to stall, then what is? Its definately not magic ... there is a perfectly rational explanation for this that can be scientifically quantified, correct? What's wrong with the explanation that the ECU code was tweaked. If anything that could probably be the easier thing to change and most of it is all code work, that doesn't require the manufacture and fabrication of new physical parts. I mean they did add a rev-limiter to the 05 models when the car is not in motion and has the clutch depressed, correct?
Not only that but supposedly the code and alogrithm of ACD for the EVO 9 has been revise to provide better preformance. Every version of the EVO is in a constant state of improvement. They even added MIVEC to the 4G63 .....
We all learned that when the 05s were coming, they were going to have a "tweaked" ECU. I sure there are a lot of little things that were done to the car that, we the general public, don't know about .... so I don't see why my assumption would be so far fetched. I mean they could have an extra piece of code that says under a certain circumstance, if there was a drastic difference between the MAF and MAP signals, that are used to extrapolate the mass of air coming into the engine, use the MAP values to calculate the amount of fuel needed to be added, correct? Thus the vent to atm BOV aren't a problem.
I'm not an engine management expert by any means but would like to learn. So if the Xede allows the user to enter the target fuel and auto-tune optimizes based on that, which as you, yourself have stated is what the ECU does on a daily basis, how is that different from a custom flash and letting the ECU do the rest? That's my point! I mean you could have a reflash, put some good quality fuel, reset your ECU and voila, your ECU starts from a blank slate and tries to use the most aggressive map/setting ...
I mean I userstand that if cuts down on the # of trips that Vishnu makes and people can have a package shipped to them. Also, it will autotune for any map that is loaded on the car and try to extract the maximum power. But at what cost, the system is by no means cheap ... for the same/lower price you could get a custom tune that would provide the same power level. Am I missing something?
I believe on auto-x and track courses, driver is the biggest factor and what really determines fast times. So under the strenuous conditions of tracking isn't it better to go with a safe tune. Thoughts?

Wrench

The auto tune feature will help your car deal with extreme conditions. What if your safe tune doesn't take a set of circumstances into consideration, like very cold or very hot ambient temperatures?

A tune is just a record of what the tuner expects of the conditions your engine will see. It is subject to human error. I once blew a piston because my tune hadn't taken into consideration a spike in barometric pressure which was at it's highest in 2 years on the day it went. Who would have thought?

To me the extra $500 is worth it for the improved drivability. If you don't want to spend the money, then go with the safe tune.

It seems to me though if you are racing autox then you will be burning through brakes and tires, so in the scheme of things, the piece of mind might be worth it, not to mention the improved engine response. You could try it and if you don't like, I'm sure you could get most of your $ back by selling it.

Just my 2c

Erik@MIL.SPEC

iTrader: (94)

Just for the record, the autotune always works from a basemap. So, if you add a few mods, you'll need a new tune.

DaWorstPlaya

iTrader: (31)

Doesn't the stock ECU do the same? I'm sure its set from the factory to take into consideration the varying extreme conditions and has various types of maps for various conditions, correct? I understand what a tune is, its a look up chart that have values that are changed by the tuner to make the engine perform more aggressively from the factory. In laymans terms a spreadsheet with altered values.
I mean if one just changes the aggressive map on the ECU to a more aggresive one and were to leave the rest unchanged, if there is a dangerous situation arising, doesn't the stock ECU first try to retard timing to cease knock activity and then switch to a different safer (richer) map?
So, while the ECU did try to extract the most power from the engine and couldn't due to environmental conditions, it switches to a safer map and if the conditions aren't that great to be pushing hard, maybe one shouldn't be pushing so hard to begin with. How is what the Xede doing any different and ground breaking then?
Also what's to say the newer 05 ECUs aren't using a different algorithm that is more proactive and can compensate better for varing conditions as shown by the BOV not stumbling issue. It isn't uncommon for this to happens, eg: the FD series of Rx-7s had ECUs that were progressively better and produced more power .... and thats within one single model ...

Wrench: when you blew a piston what sort of engine management were you using?
Reflashed ECU with just the aggressive map changed? or all maps changed?
Stand alone unit eg AEM?
Xede which is a intercept device that works relative to signals from the ECU (meaning it adds or subtracts to the values the ECU is receiving)?
Utec which is more of an intercept device that in not relative but more like a standalone and its values are absolute?

BlackEVO

Actually, it was a 2-3 degree advance, but a noticeable amount of torque nonetheless. It wasn't retarded due to knock, it was retarded due to increased boost. The higher the MAS reading (more boost), the more careful the ECU is about handing out timing at peak boost. If it was a retard due to knock, I assure you that I wouldn't have advanced it

Well, if a 15-20WHP/WTQ dip is "little". If my car is pulling timing on a pull, I know it without any logging. I can feel the loss in power and that's just a few degrees. The upper gears are longer, EGTs have more of a chance to climb and will be more knock prone because of it, if that's what you mean.

That has always been my tuning philosphy. Richer (but no overly so) is safer of course, especially on pump gas. So, richer with an ideal amount of timing for it is the way to go IMHO. The ECU can always pull timing if knock is detected, but if you're lean and knock, the margin of error is somewhat smaller. The second scenario will not necessarily produce less power. Keep in mind that the amount of timing necessary for a complete burn is relative to the AFR. All things being equal, if the AFR was "constant" and just the timing was altered, it would change the AFR.

I am very sure there were code changes, but again, adding code/calculations/conditions for the MDP would still probably be the least likely change. Coding the ECU is not like writing a prog in Visual Studio I can think of any number of things that could've changed to stabilize it that would be much "easier": clamp the minimum freq on the MAF readings, remove the voltage cut to the fuel pump with the clutch out, slow/smooth the transient response on big TPS changes (kinda like a EIDS), etc.

Don't confuse what the ECU does simply to run the engine to a performance tune. For example, at closed loop, the ECU will always be making adjustments to fuel trims to keep as close to 14.7:1 as possible (why narrow band O2 gauges "bounce"). At WOT (open loop), it doesn't do anything like that (why O2 gauges read more stable). Timing tables have a MAX limit. It's not like it starts from 0 and just keeps adding timing until it hears knock. It adds timing until it hears knock OR reaches it's limit. The most aggressive map in the ECU is still a joke on a stock Evo (especially fuel), let alone a modded one. It's going to be mostly or all preconditioned table look-ups and knock monitoring.

Timing will be at it's lowest at peak boost. The ECU retards the timing as boost is coming on, reaches the minimum at peak boost and should climb rather steadily through the RPM range after full boost is hit.

Going back to my 2-3 degree example above, at 19psi, my ECU would give about 8-9 degrees peak around 4000-4500. Turning the boost up to 21psi and that dropped to 4-5 degrees max. Not because of knock, just for safety. So let's say under my conditions, the 5 degrees is the max advance the ECU will ever give there as conditioned by the code. I added 3 degrees back and voila, another 20 WTQ and still no knock. Maybe I could've added 4 or 5 degrees and picked up even more TQ there, but I stopped and said that is good enough. Now wouldn't it be nice if I kept the 3 degrees and TQ that I safely added and if even more timing was available knock-free, to let the autotune do it on the fly? That's what I am talking about when I say the ECU does it's thing (gives 5 degrees max, but will range lower depending on conditions), that a base map should be used (my +3 degrees) and the autotune would fine tune it (maybe add even more timing and thus, more TQ if there's room).

That might have been a confusing explanation....I'm tired If I confused you, I will try to reword it. Just know that the timing tables have a max and that max is for a particular point and depends on engine conditions at that point. A flash or piggyback will add "x" amount to what the ECU would give at that particular point and that's it. The autotune would go one step further and make further tweaks (like the ECU would do) if it's safe to do so.

DaWorstPlaya

iTrader: (31)

No your explanation was not confusing at all, it made perfect sense to me. Tuning is not "rocket science" by any means its actually very easy once you figure out the basics ... only thing is there are multiple variables to keep track of ... I can understand why you would want to advance timing even more the higher your rpms went, given all else constant. If the pistons are moving at a higher rate and if the rate of combustion in the engine cylinder is constant you would want the fuel to ignite earlier to sorta finish its burn cycle as the piston hits TDC (on its way up, correct?). Since the pistons are traveling fasters at higher rpms you have to start the ignition earlier to finish at the same time.
Then it also makes sense not to advance timing as much when you are at lower rpms. Max boost on a stock EVO occurs somewhere around 3000-3500 rpms ... If you have more boost (more air in the cylinder and I'm not talking about the rate of flow of air) you will also have more fuel to compensate. More of each in the same volume means that the pressure is going to be higher in the chamber and combustion is more explosive .... so it also means you don't need as much advance in timing ...
Ofcourse you don't want to advance ignition when you hear knock ...
I agree is 15 WHP/WTrq at lot but again, I have to question how useful it is as you are probably not going to notice it through the entire rpm range but more like a 100th of an rpm. As Shiv himself has said that it is own to smooth out the dips. Is small gain in power worth the price you pay? I guess that is up to the buyer
Ofcourse I'm not suggesting they use Visual Basic ... but I'm pretty sure they use sort of high level language that compiles into Assembly or Machine Code or what ever they run ... not only that but the system is probably using embeded code that goes through a bunch of tests to check on stability ... You'ld be surprise how quickly some people can write code ... test it ... and release it ...
So what you are saying is, with an aftermarket solution you can advance the timing more than the stock ECU can? even if the ECU has been reflashed? ... But how do you or I, know that about the newer 05 ECUs? I'm not questioning what you know about tuning because its obvious you know what you are talking about ... but I'm more concerned what we know about the workings of 05 ECUs ...
I do know of the folley of the O2 sensor and closed loop mode ... what I didn't know was, it switched to open loop mode under WOT, as I am still learning. I was under the impression open loop was only used during start up. Good to know ...
Anyways the proof as they say will be in the pudding ... we'll have to see what impact it will make in the 1/4 mile and what dyno #s one gets ...
I'ld rather know what is going on and how things work for myself than just blindly believe claims ....
Thanks for taking the time to explain your thoughts to me BlackEvo

BlackEVO

Exactly. The thing is, depending on variables the AFR will change day-to-day (minute-by-minute really) and thus so would the ideal amount of timing. Piggybacks will allow you to tweak the timing to offset what the ECU does to better match the AFRs for making more power.

It doesn't quite work like that. Adding 3° at 4000RPM doesn't mean 3999 and 4001 RPM points are 3° less. The timing change is usually extrapolated (usually over a couple hundred RPM) to spread out changes made by piggybacks. You do of course need to tune the entire RPM range anyway, but my additions are such that they taper in->peak->taper out to make for a nice smooth pull from 2000-7500RPM.

It is beneficial from a performance standpoint to have higher TQ as early in the powerband as possible and it does make a difference at later RPM if you are starting from a higher point. A sizeable area under my TQ curve was in increased. Peak numbers are nice, but it's the entire area under the curves that determine performance, so moving a good chunk of it up is always a good thing.

Exactly and yes, even on top of a flashed ECU. Many people with flashes still use SAFCs to fine tune their fueling. ECUs need inputs from sensors to do their thing. Piggybacks basically intercept these signals and output altered ones that include any tuning adjustments. Piggybacks that have their own injector/ignition drivers like the UTEC have total control and allow setting the exact ouput (not +/- adjustments to the ECU) since they drive the injectors/coils themselves.

As long as ECUs require sensor input to base their decisions on, piggybacks will be around. Pins and code may change on ECUs which will require different taps, firmware/software/hardware updates, but that's up to the piggyback manufacturer to take care of. If you want to know exactly what has changed, you'll need to find someone with experience specifically on the '05 ECUs.

That's why it's easy to tune at WOT. Tuning part-throttle (like for larger injectors) can be a pain because the ECU is making it's own adjustments at the same time...at WOT, not so much (mostly just watching for knock). Coolant/IAT temp, etc. may play some small part in AFR changes at WOT, but for the most part it's just keeping an eye out for knock.

FYI, WOT happens somewhere over 75% throttle OR >4500RPM, so it's possible to hit open loop without it being floored.

No problem I don't spend nearly enough time posting on the forums, so I am glad when I can help someone out!

DaWorstPlaya

iTrader: (31)

I figured as much, that WOT didn't necesarily happen when the gas was pushed all the way down as I get performance similar to WOT under near WOT throttle positions .... So the SAFC is a piggy back device that controls AFR and timing advance, but doesn't control boost like the Xede, correct? So what happens when you advance timing too much on the SAFC after a tune and hit knock? Is the SAFC static, meaning you set advance to a fixed value you want on top of what the ECU decides to set its advance and its always that fixed amount more than the ECU? ...
Another thing I want to know is, auto-tune is supposed to be proactive vs reactive, reactive to me means the ECU takes action to reduce timing once it detects knock activity ... what factors can to measure/keep track of to be proactive? to be able to push timing as much as possible but not hit knock? or predict how much timing one can run with out hitting knock?
I mean do tuners know this because they have been tuning for a while and through experience know that they can run this much timing advance, under the best conditions (colder weather)? or are there calculations that you can use, that incorporate rpm, boost, etc ... as we have already dicussed that timing seems to be based on rpm and boost ...
Is it also true that wideband 02 go bad quickly?
BlackEvo: will you be getting an auto-tune kit?

BlackEVO

It controls AFR. Technically it controls how much of the MAS reading the ECU sees and the larger or smaller reading (depending whether fattening or leaning) causes the ECU to add more or less fuel and less or more timing.

Since the MAS signal also effects the timing (like I was saying more boost = more MAS Hz = less timing at peak boost), timing is altered too. You can't really tune timing specifically with the SAFC since it's a by-product of adjusting the MAS. Adjusting the MAS % directly changes fueling only. If you lean out the SAFC enough to hit knock, it's simply too lean in that area and/or just before that area and will need to be richened up.

It's fixed in the sense that the FUEL will be a static +/- % than the ECU would do on its own. TIMING, as mentioned, is not directly controllable with a SAFC. To directly tune timing, you would need something else like a UTEC, ECU+, Xede that has separate support and tables for it.

Supposedly it is proactive, meaning if it's going to knock with the extra autotune timing, it will not add that extra timing. The stock ECU has (from what I understand) a pretty complex knock control setup and will still do it's own thing where knock is concerned, but the Xede autotune apparently has a proactive knock control for the on-the-fly timing it gives and will only add it when safe to do so. Most piggybacks and standalones are reactive in that they can take measures only once knock is heard.

The autotune must have some kind of formula to determine how much extra advance can be given without knock because it is not a set addition, but rather something between 1-3°. While it may not predict just how much more advance could be given at a max, you can always watch what it does for adjustments and tune the base map in that direction accordingly to zero in on it where 3° would be more than you need from the autotune.

No tuner knows what the absolute max each and every Evo will take for timing under every condition. The ECU is obviously a base and is preprogrammed with tables that have max values under all conditions, so that's a starting point. Experience and logging/dynoing will get you closer to ideal when tuning, but a safety margin still needs to be left because conditions do change. That's the beauty of the autotune...it will use more of what's left on the table when safe to do so.

Timing is based on much more than simply RPM and boost. IAT, coolant temp, TPS, load, baro, etc. are all taken into account. Because the ECU has to account for so many different variables, it has a safety margin built in which is why the max aggressive tables are often still not enough for max performance....especially if you are running race gas where you can take advantage of much higher timing without knock than the ECU would give.

Since conditions change often (constantly), without the autotune, if temps dropped say 20° between day and night and you could run 3° more timing without knock and pick up an extra 15-20WHP/WTQ, you'd need to physically adjust it on a map for a piggyback. The autotune will do that automatically 24/7.

Wideband O2s have replaced narrowbands as standard equipment on some cars now as OEM....what does that tell you? 1) They should last the same or close to the same time as narrowbands and 2) as they become more and more widely used, pricing on replacements will drop even more. They really aren't much more expensive than replacing a stock O2 now. The controller is the expensive part, the sensor itself is only ~30% of the cost of a kit ($130ish). Mine has been in for quite a while now (> year) and is still working just fine.

If it works like Shiv says it does, I am going to really have a hard time not at least getting one to try since I am one of those people who does feel the need to tune just running to the store for milk Unless Shiv wants to send me one free/cheap for "testing", I will probably just have to wait and see. My only hold up on getting one would be 1) if I need to swap to a different wideband controller and 2) the price. If I can use my AEM UEGO with it, it will be more feasible. Short of that, a full setup with autotune, wideband and limit removal flash puts it dangerously close to me keeping my current management and upgrading the turbo instead