Part throttle boost (closed loop) tuning for FI

Discussion in 'Engine & Performance Modifications' started by Mr MoPar, Jun 7, 2009.

  1. Mr MoPar

    Mr MoPar New Member

    Messages:
    88
    Likes Received:
    0
    Joined:
    Apr 11, 2009
    See BOLD below

    I think the tuners have done a lot with what they have available to them, none of this is in no way meant to be a bad reflection on what they do or the software they use. Since they can't reprogram the ECU, they are left with manipulating table data. Obviously some are better than others. It is not an issue with the software either, it is no ones fault that the ECU programming code remains locked up in a vault in Auburn Hills!

    The whole intent of the thread is to elicit discussion of the limitations so people know what they are, know what to look out for, and to know what to do to monitor and protect their investment.

    Is 5-6 psi safe on these engines, yeah, I say it is, as long as the fuel, timing, and combustion chamber temps are in check.

    I put 32,000 on a stock 5.7 running up to 9 psi, and believe me, the foot was in it all the time. It was not babied around!

     
  2. Mr MoPar

    Mr MoPar New Member

    Messages:
    88
    Likes Received:
    0
    Joined:
    Apr 11, 2009
    That is a very good description of things. Previously when only the SMT-6 was around, you did just that. Either toe out of boost or go to the floor. You did not drive in part throttle boost at 14.7 AFR.

    Now things are different, and a supercharger is probably easier to tune than a turbo simply because the supercharger boost tends to come in at a point based on RPM, not load. Therefore it easier to lower the WOT threshold to cover a larger range of the boost envelope as well as some of the other tweaks.

    On a turbo, I can bring in boost as low as 1800 to 2000 rpm, or I can run it up to redline in the lower gears and never hit boost. Makes it much tougher to try and cover that broad of an envelope.

    Everyone's definition of drivability is different, what suits one person may not suit another. Myself, I want to be able to have a somewhat normal vacuum range of drivability, and if I go into a couple lbs of boost on the highway, I don't want it to just take off. I want to be able to control what is going on, and I don't want to have to worry about it either.
     
  3. Mr MoPar

    Mr MoPar New Member

    Messages:
    88
    Likes Received:
    0
    Joined:
    Apr 11, 2009
    I agree with what you are saying, but unfortunately unless Fiat decides to sell the ECU code, I don't think any other solutions are going to show up. I have been researching it for years.

    Hopefully I will be able to get to some of the events this year and some owners of FI vehicles can take my Cherokee for a spin. First hand impressions speak far more than words do.

    Since the turbo is based on loading, I can drive it right on boost threshold at 0 inches of vacuum, I can accelerate without going into boost, or I can drive at 2-3 psi of boost and the vehicle just doesn't do anything stupid. Go to the floor, and hang on.

    I bet if you were to check, at the point you see the AFR increase, you would also see it transition from closed to open loop, which is what most people call WOT, even though you are no where near the floor.

    I also would not be surprised that if you were to open up your bypass valve so it does not go into boost, at the same pedal/throttle position you would still see it transition to open loop. If so, then this tells you exactly that the fuel correction is based on throttle position percent open.
     
    Last edited: Jun 9, 2009
  4. Quick

    Quick Mgmt. - I can't help you

    Messages:
    7,549
    Likes Received:
    0
    Joined:
    Nov 6, 2007
    Location:
    San Jose, CA
    Yes, I understand that part.

    I'm having a hard time describing what I'm thinking about...

    When the ECU goes from closed loop to open loop it stops monitoring the narrowbands and adjusts fuel and timing from the PE tables. I'm under the impression (probably mistaken) that this is a discrete jump in parameters.
    Lets assume everything is stock and the ECU transitions from closed loop to open loop at 80% throttle. At 79% throttle you would be running close to 14.7 AFR. At 80% throttle the ECU would transition to open loop operation and factor in the adjustment from the PE tables. I'm under the impression that your AFR would immediately go rich to something like 12.8 (we're still stock).

    If I'm wrong and it's a smooth transition and AFR ramps smoothly from 14.7 to 12.8 as the throttle goes from 80% to 100% then disregard the rest below (as Emily Lutella would say... never mind).

    Continuing under the assumption there is a discrete jump when the ECU transitions from closed loop to open loop, you would need to accomodate that since you've already started making adjustments for boost at part throttle while the ECU was in closed loop -- the next cell in your table would have a larger delta from the previous cell than the delta between the n-1 and n-2 cells. The implication is that the unit is aware of when the ECU goes from closed loop to open loop.

    I was thinking that you could remove that dependency if you used input from the wideband (along with boost) to control the fuel add (cell selection in the table).
     
  5. Mr MoPar

    Mr MoPar New Member

    Messages:
    88
    Likes Received:
    0
    Joined:
    Apr 11, 2009
    See BOLD below

     
  6. lafrad

    lafrad Supporting Vendor

    Messages:
    17
    Likes Received:
    0
    Joined:
    Apr 10, 2008
    I've worked with a few boosted tunes, and I just wanted to chime in with some observations...

    A properly scaled tune can properly control Closed loop in ALL situations, so long as the MAP sensor has good resolution for your driving conditions. (Scale the Map sensor input to read a proper 2 or 3 bar map, and adjust the VE, PE, Spark, etc tables to match the expected engine range). The vehicles that I have worked with were able to properly fuel closed loop from Map PRatios of 0.10 all the way up to 1.50, with short and long term fuel trims working perfectly, giving a VE table that displays expected values from top to bottom. (if you have VE's jumping significantly in boost, you probably are near your limit on the injectors or fuel delivery.. the mopar ECM's read VE against the current PRatio, and there is no magic about boost when it comes to making an engine more efficient or not...)

    Closed loop with narrow band sensors is a "dance" around 14.7:1. with a good closed loop tune where the VE's have had a bunch of fine tuning, that should be bouncing between 13.8 and 15.2, rapidly, with a LOT of time spent between 14.4 and 15.0. This is also the ONLY way to keep catalysts living a LONG life (as mandated by the EPA).

    Factory tunes for SRT4's, Pontiac GTP's, Supercharged Mini's, Supercharged Colbalts all allow part throttle cruising with boost, and they have a federal requirement to have long term catalytic converters in them. That means they are able to cruise around at 14.7:1 in boosted situations.

    When it comes down to what a combustion chamber see's in boost, its really not *significantly* different than changing elevations from sea level to high in the mountains, or going down into death valley. the scale of the change is a bit more but: when the piston reaches TDC on the compression stroke, it didn't care that it was filled at 0psi of boost or 14 psi of boost. what MATTERS, is how much compressed pressure is in the chamber, and what temp did the mixture reach.

    Now, Compressing air heats it up. that is a KNOWN thing. Compressing Air with a piston, like a motor does, is one of the most efficient ways to do it: the volume of the containing chamber is reduced with relatively little movement of things around it. Unlike a supercharger, where the very act of moving air through the apparatus "wastes" energy by heating up the air. this means 14 psi from a supercharger (which is a compression ratio of approx 2:1 at sea level, btw) will have WARMER air than a piston compressing the same volume to 14 psi.
    That being said, a GOOD inter-cooler setup will return the intake charge temps to ambient, giving us a "standard" starting point again. we just a slightly denser air charge before the piston starts compressing the mixture.

    Now, the more pressure we have at the top of the compression stroke, the warmer the combustion chamber temps. Thats just the way it is. if you run advanced timing, you get a higher peak pressure, with smaller combustion chamber volumes, making the chamber temps THAT MUCH more. Stock block hemi's have really closely gapped rings and thin ring lands. that heat is able to move into the pistons and rings quickly. The increased cyl pressures also cause more ring friction (it seals better, but also heats up more as the piston travels in the cyl). again, heating it up a bit. All of this means: TREAD CAREFULLY with ANY setup that increases combustion chamber pressures. heat those rings up and they will pop the piston apart! (yea, that was said back in the first post)

    Well, that can be said about cams, boost, spray, all sorts of ways to increase engine efficiency. we have seen plenty of STOCK block N/A vehicles lift a ring land due to ring gaps closing. Boost may encourage the situation, but it IN NO WAY is the actual cause for it.


    Something that you DO have to worry about beyond chamber temps: Exhaust Gas Temps. depending on where your EGT probe is, it may not accurately reflect what the chamber temps are. if your EGT probe isn't directly at the exhaust valve, highly retarded spark timing will show INCREASED egt's... since the combustion event is started so LATE in the cycle, the mixture can still be burning when the exhaust valve opens... granted, that's a significant case, but it does point out to trying to make sure you are reading your sensors correctly.

    In regards to boost , closed loop, and engine design, it just all comes down to: what can the parts physically handle. With increased chamber temps & pressures, more energy has to be put into the tune to make sure the stress on the parts is kept to a minimum. There will be a point that stock parts cannot reliably handle it, no matter what effort is put into the tune.... but that is just not straightforward: cams, heads, blower types, intercooler efficiency... all make a difference.

    I do NOT think that a tune needs to stay away from boost in closed loop. reasonable levels of boost with properly matched VE tune, matched up with appropriate timing values for the MAP levels, will be save to drive on, for any motor. "reasonable" on stock will most likely be less boost than "reasonable" on a low compression, forged piston, big ring gap stroker. after all, combustion is combustion, no matter what it took to get there...

    If we did have options available for running a wideband for closed loop (not really available with the current ECM programming codes), and/or we had a good way to trigger open loop (*AND POWER ENRICHMENT*) via just PRatio, we would have a bit more flexibility with what can be done, but I would venture to guess we wouldn't buy much. there is only so much you can do with an engine that was originally targeted for N/A applications.

    ----

    as a side note, I DO agree that we don't have *everything* we need for 100% perfect control of a boosted motor... well, boosted motors that rely on pressurizing the throttle blade...

    our ETC (electronic throttle control) systems have NO WAY to know how much pressure is on the *outside* of the blade. The computer makes assumptions about how much air can flow across a throttle body at a given throttle angle. As soon as you start changing the pressure on the outside of the plate, those assumptions are no longer accurate, and you can get weird behaviour. "autodrive", "odd part throttle shift patterns", "weird down-shifts", "surging" can all crop up, depending on how much boost you run, and how all the supporting hardware works in providing consistent air delivery to the motor.

    Something that I would personally like to see, is adjustments to the tune (and PCM hardware) to support wiring in a TIP sensor, like the SRT4's have. that would give the PCM a reference and the ability to properly set throttle angles in any Closed loop, open loop, power enrichment, etc... situation.


    -mike
     
    Last edited: Jun 9, 2009
  7. 1bad4dr

    1bad4dr Mr. Meany

    Messages:
    19,670
    Likes Received:
    0
    Joined:
    Oct 10, 2007
    Location:
    Corn Country
    Now I know why I stay away from computer tuning. lol Carb tuning FTW. haha

    Very cool that you chimed in, Mike. Thanks!

    And Mike has tuned a few of the 10 second LX's. ;)
     
  8. greatone61

    greatone61 Full Access Member

    Messages:
    875
    Likes Received:
    0
    Joined:
    Dec 31, 2007
    Location:
    Santa Fe
    Props to all the contributors in this thread. This is one of the most educational threads I have ever read on ANY SRT forum....this is awesome stuff.
     
  9. Bud

    Bud GG EVO IX MR

    Messages:
    3,949
    Likes Received:
    0
    Joined:
    Oct 14, 2007
    I've been reading every post as well. Great stuff guys!!
     
  10. lafrad

    lafrad Supporting Vendor

    Messages:
    17
    Likes Received:
    0
    Joined:
    Apr 10, 2008
    One of them is a little faster than that... :hypno:



    Something to keep in mind: Unless the programming software (CMR, SCT, ..., etc) that I have seen is lying to me, ALL fueling, spark, load, torque, airflow, etc calculations are made based on the kPa reading that is spit out *after* the MAP scaling value is read. (some go one step further and are based on the kPa / baro = Pratio value) I have witnessed this first hand by putting a 2 bar map on my N/A motor, setting the appropriate scaling values, then going for a drive. There was *no* changes in shifting, fueling, spark... I didn't drive it long enough, but I would assume fuel economy would also be unchanged.

    This also goes for throwing *really big* injectors on. the only problem I have seen with those is some rare "idle quality" issues when MONSTER injectors are matched with a small cam... Otherwise, the fact that these PCM's seem to be able to control fuel delivery with pulsewidth accuracy WELL below a millisecond... (I would venture to guess its in the tens of microseconds before they loose accuracy) we can run whatever injectors we want on our cars so long as they flow enough for the given application. There are other things to think about of course... but for the sake this discussion its not relevant.
     
  11. 1bad4dr

    1bad4dr Mr. Meany

    Messages:
    19,670
    Likes Received:
    0
    Joined:
    Oct 10, 2007
    Location:
    Corn Country
    bwahahaha Oh crap, I forgot about Chris (Hemi C) and his 9 second Magnum. hehe
     
  12. Quick

    Quick Mgmt. - I can't help you

    Messages:
    7,549
    Likes Received:
    0
    Joined:
    Nov 6, 2007
    Location:
    San Jose, CA
    fixed it.
     
  13. 1FST4DR

    1FST4DR Supporting Vendor

    Messages:
    2,051
    Likes Received:
    0
    Joined:
    Oct 14, 2007
    Location:
    Ft Laud
    Wow,Great stuff here.Im really a moron when it comes to this stuff so forgive me if I am out of context.But if the main problem with tuning the new hemis is Part Throttle,could this be a good reason why N20 cars seem to hold up well because N20 is usually activated at WOT?
     
  14. Mr MoPar

    Mr MoPar New Member

    Messages:
    88
    Likes Received:
    0
    Joined:
    Apr 11, 2009
    See items in BOLD below.

    Mike, I appreciate you taking the time to join the discussion; this will be of great benefit to many people.

    In that light, possibly you could answer a few questions regarding how things actually work in an ECU tune for FI. Obviously I am not asking to divulge “tricks of the trade†but hopefully some generic explanations are possible.

    I understand the need to scale certain tables for the size of the injector, and size of the MAP sensor. When a vehicle enters into boost, at some point, additional fuel is added. From what I have seen, and the somewhat cryptic discussions I have had with others, this is primarily initiated as a function of the throttle position for WOT threshold, and not simply due to the fact that the MAP sensor value has triggered a transition over to WOT?

    From the description of the ECU, it only has two primary fuel tables (well three if you count deceleration), closed loop, and open loop. As such, since the closed loop tables are linked to the O2 feedback, you are basically restricted to a 14.7 AFR in closed loop?

    In some discussions, it has been alluded to that the commanded AFR can be modified, is this associated with closed and open loop, or only open loop?
    I would think that if commanded AFR was changed for the closed loop tables, it would be across the board affecting the entire table which would likely not be a desirable thing to do.

    Or, is the commanded AFR more fuel table cell specific where you could have a different commanded AFR in different areas of the closed loop map or even open loop tables (more interested in the closed loop since that is what the thread is discussing)?

    Also, if you would be willing to do so, There are a lot of terms associated with ECU tuning that are commonly used, but those who are not involved have very little understanding of what they are and what their importance is. If you could take the time to explain some of them, it would be of great benefit to those interested in the topic.

    For example, VE is Volumetric Efficiency, what role does it play, and why is it important to be updated/modified such as running a bigger cam in an NA motor, effect of running Forced Induction, etc.

    Not looking for an explanation of how, but just some general insight into what some of the terms that are commonly thrown around in general discussion related to the CMR and SCT tuning.

    Thanks again for the contributions.
     
  15. cvp33

    cvp33 Full Access Member

    Messages:
    428
    Likes Received:
    0
    Joined:
    Jun 22, 2008
    Enjoying the lessons here and I appreciate the experts weighing in on all sides. Is there a reason why you can't set your closed loop default to a reasonable AFR to help with FI setups? Or is closed loop by definition not able to be modified? Just curious. I mean if I have to settle for a band-aid fix I'd rather error on the rich side than the lean side. I'm sure I'm over simplifying a very complex problem. But as my co-workers would attest I have a gift for that. :whistle:
     
  16. Quick

    Quick Mgmt. - I can't help you

    Messages:
    7,549
    Likes Received:
    0
    Joined:
    Nov 6, 2007
    Location:
    San Jose, CA
    The issue is this.
    The PCM code will hunt for 14.7:1.
    The narrowband sensor is specifically for detecting 14.7:1 and not much else -- basically it reads LEAN <-> 14.7 <-> RICH.

    The PCM code (not data) cannot be changed.
    The narrowband callibration point cannot be changed.

    Ideally you want a richer afr but you only want it under boost. You really don't want to be running pig rich all the time you're not in boost. So you could do something like attenuate the signal from the narrowbands and you'd get something like LEAN <-> 12.4 <-> RICH but then you'd be running 12.4 ALL the time you're in closed loop even with no boost.

    Apparently this can be done using different approaches. Big division would be doing it entirely through manipulation of data and tables within the PCM or doing it with an auxiliary computer (piggyback) and control. If it's done with a PCM tune it doesn't sound like it's easy and maybe some tuners don't do such a good job of it. The piggyback solution that Tim has described sounds more straight forward but requres additional hardware, installation and cost.

    That's my understanding so far.
     
  17. Mr MoPar

    Mr MoPar New Member

    Messages:
    88
    Likes Received:
    0
    Joined:
    Apr 11, 2009
    This is in line with my question about the commanded AFR, what areas does it apply to, and it's affect. You certainly do not want to be running around all the time richer than 14.7, it would have a very detrimental affect on the emissions system, namely the cats, plus waste a bunch of fuel. I think it would need to be cell specific in the closed loop maps, but then again, what are the limitations of the O2 sensors that drive the fuel trims. While they have a 0-5v range, they only tick back and forth around the midpoint with about a 1 volt span. Not very useful trying to target some other AFR.

    Hopefully Mike will be willing to share some insight as to what can be manipulated and its affect.
     
  18. Quick

    Quick Mgmt. - I can't help you

    Messages:
    7,549
    Likes Received:
    0
    Joined:
    Nov 6, 2007
    Location:
    San Jose, CA
    We have bosch narrowbands. 0v - 1v.
     
  19. Mr MoPar

    Mr MoPar New Member

    Messages:
    88
    Likes Received:
    0
    Joined:
    Apr 11, 2009
    Part of this depends on an explanation of what can be done with commanded AFR and its affects. Hopefully we can get an answer.

    Regarding the O2's, the new O2 systems in use today are not like the old narrowbands, so the old bag of tricks doesn't work. This is because the signal floats in a 0-5v range. The ECU has to see the signal float, and it is constantly manipulating it as the O2's are one of the most critical parts of the emissions system.

    The manufacturer had to make some specific modifications to both the hardware and software in the initial unit I was running to make it work. Basically, it has to read the signal in real time, add or subtract the offset from the signal and feed that to the ECU.

    Due to the speeds involved, the unit I am using is programmed in assembler language rather than high level or object base language. Most of the programs used on computers are programmed in high level or object base language, and may have hundreds of executable machine language instructions. This would be too slow for dealing with the speeds involved in engine controls.

    Not being able to do everything by manipulating the ECU data drove me to find/develop what I am using. Unfortunately it is an added expense, but we are only looking at one aspect (closed loop control) of the many functions it can perform.
     
  20. Mr MoPar

    Mr MoPar New Member

    Messages:
    88
    Likes Received:
    0
    Joined:
    Apr 11, 2009
    No, they are not a 0-1v narrowband sensor, I have done quite a bit of research on them. The actual data is an approximately 1 v range high/low in the middle of a 0-5v span.

    The values fluctuate, the low might be 3.45v and the high 4.24v, and this constantly moves around, and the span between the high/low fluctuates, sometimes it might be a .75v span and sometimes a 1.12 span for example.

    Here is a description of both, the Zirconia sensor is the 0-1v narrowband you mentioned, the second is the type being employed in today's systems, including the Hemi. This is why we had to change how my initial unit actually functioned.

    Zirconia sensor

    The zirconium dioxide, or zirconia, lambda sensor is based on a solid-state electrochemical fuel cell called the Nernst cell. Its two electrodes provide an output voltage corresponding to the quantity of oxygen in the exhaust relative to that in the atmosphere. An output voltage of 0.2 V (200 mV) DC represents a "lean mixture" of fuel and oxygen, where the amount of oxygen entering the cylinder is sufficient to fully oxidize the carbon monoxide (CO), produced in burning the air and fuel, into carbon dioxide (CO2). An output voltage of 0.8 V (800 mV) DC represents a "rich mixture", one which is high in unburned fuel and low in remaining oxygen. The ideal setpoint is approximately 0.45 V (450 mV) DC. This is where the quantities of air and fuel are in the optimum ratio, which is ~0.5% lean of the stoichiometric point, such that the exhaust output contains minimal carbon monoxide.

    Titania sensor

    A less common type of narrow-band lambda sensor has a ceramic element made of titanium dioxide (titania). This type does not generate its own voltage, but changes its electrical resistance in response to the oxygen concentration. The resistance of the titania is a function of the oxygen partial pressure and the temperature. Therefore, some sensors are used with a gas temperature sensor to compensate for the resistance change due to temperature. The resistance value at any temperature is about 1/1000th the change in oxygen concentration. Luckily, at lambda = 1, there is a large change of oxygen, so the resistance change is typically 1000 times between rich and lean, depending on the temperature.

    As titania is an N-type semiconductor with a structure TiO2-x, the x defects in the crystal lattice conduct the charge. So, for fuel-rich exhaust the resistance is low, and for fuel-lean exhaust the resistance is high. The control unit feeds the sensor with a small electrical current and measures the resulting voltage across the sensor, which varies from near 0 volts to about 5 volts. Like the zirconia sensor, this type is nonlinear, such that it is sometimes simplistically described as a binary indicator, reading either "rich" or "lean". Titania sensors are more expensive than zirconia sensors, but they also respond faster.

    I surmise that the change was due to the tightening up on the emissions controls by the Fed starting in 2004. With the old style narrowband, the ECU did not have "control" of the sensor. With the new style, the ECU can vary the voltage applied, and look for a corresponding change of the output, all the while the sensor is still providing the normal control function. This makes it easy for the ECU to detect a failing sensor, or I think more importantly tampering with the sensor.

    So side topic on O2 sensors!
     
    Last edited: Jun 10, 2009