Previously I had started a thread regarding part throttle boost tuning (now a sticky!), so I thought I would drop back in and start a new thread detailing the latest efforts related to tuning for FI. I hope it is as successful a discussion as the previous thread turned out to be. In the previous thread, we embarked on a discussion of the iEMS3 and it's roll in the tuning matrix, in particular closed loop boost fuel correction. Now I have ventured to the next level of the tuning matrix. CMR and iEMS3 tuning combined, a new direction................ I have been working with Josh at HHP and Ted/Jeff/Jamie of allusamotorsports on a combined tuning effort that has been an interesting project. The vehicle was a lowered compression 6.1 based block with a very high lift cam and high flow heads. On top of that, it had a twin turbo kit installed, so it had a lot of tuning issues to overcome. What we did, was to combine the capabilities of a CMR tune and the iEMS3 together. The CMR tune was configured to address the air flow changes needed for the lowered compression, high lift cam, and high flow heads. The tune was scaled for the stock MAP sensor in lieu of a typical 2 bar sensor. The tune was scaled for the larger injectors, and the other typical changes such as torque management, idle, throttle response, etc. The timing changes were less aggressive than in a typical FI tune as well. In this fashion, the iEMS3 was essentially in a pass though condition whenever operating in a vacuum state. When the vehicle would enter a boosted state, the iEMS3 would take over fueling requirements providing part throttle boost fuel corrections and increased fuel as the boost pressures increased through direct injector control. The iEMS3 provides control over the stock 1 BAR map sensor to prevent ECU MAP voltage over range, and control of the O2 sensors for part throttle boost fuel correction. It also provided the timing retard needed when in boost so the timing tables could be left closer to stock configuration. The iEMS3 was controlling boost levels through it's internal electronic boost control. This setup worked out very well on this vehicle, and provides a much more civilized driving environment than it previously had. It is inter-cooled and is using water/meth injection for combustion chamber cooling rather than dumping a lot of fuel and a lot of timing retard. At 7psi I believe the final numbers were 525rwhp and around 530 rwtq, and this was close to the limit of the fuel injectors and stock fuel system. The primary goal was conservatism and drivability This has proven that the iEMS3 can work directly with a specialized CMR tune and address the challenges of FI on the Hemi's by combining the technologies to address the issues and concessions that have been made up to this point. I invite discussion on this technique and any questions regarding it. I am including a response that I had made outlining some basic "how did you do that" questions. Originally Posted by name removed "how did you hook it up? what qualifications are expected for a tuner to know how to work with the iEMS2? It only corrects fuel?" I will answer your questions at the end, but first let me clarify a couple things. You do not have to run a CMR tune to use the iEMS3, it is fully capable of handling the FI tuning on its own. Conversely, many people are running FI without using an iEMS3, relying on CMR for their tuning. Using CMR alone for FI will not allow for fuel correction when a closed loop boosted condition is present, so you are left running a 14.7 AFR under low boost. Commonly a larger than stock MAP sensor is used to prevent MAP voltage over range and resulting engine codes. In some cases other drivability issues can remain as well. If a vehicle is running an aftermarket cam and in some cases very high flow heads (or combination of the two) where it creates low vacuum idle issues, that situation can only be corrected by modifying the ECU parameters via a CMR tune. The iEMS3 works very well on it’s own for vehicles that do not have airflow issues and resulting low vacuum conditions (vehicles with stock or near stock head and cam profiles). The iEMS3 does not write to the ECU, so it can not correct for such airflow issues, and it is not intended to do so. So both methods have their own limitations. This is a new approach combining the two methods together to address these challenges, and to provide as near stock like integration of FI into a modern HEMI powered vehicle as currently possible. CMR can handle the airflow modifications needed in certain vehicles, injector sizing, plus the usual tweaks such as idle, throttle response, torque management, etc. The iEMS3 eliminates the need to run a larger MAP sensor and any drivability issues that could be related to using a larger range MAP sensor. The iEMS3 provides the closed loop boost fuel correction that can not be provided by CMR alone. It also reduces the need to run very rich AFR,s and very aggressive timing retard to reduce combustion chamber pressures to reduce heat and potential knock. The iEMS3 can handle fuel addition and timing retard throughout the boosted range of operation. It can also handle injector sizing as well. Combining the two together allows us to use the best of both to further refine the tunability of Forced Induction. This is what I see as the ultimate setup; CMR tune that uses the stock MAP sensor. CMR scaled for the new injector sizing. CMR has less aggressive timing changes. CMR changes to correct for any airflow needs related to non stock heads and cam. CMR normal tweaks for idle, fans, torque management, throttle response, etc. Essentially, an optimized N/A tune running larger injectors. iEMS3 set for 100% of ECU fuel pulse and no injector offset. iEMS3 configured to provide MAP sensor voltage control to prevent ECU signal over range. iEMS3 configured for O2 sensor control providing closed loop boost fuel correction. iEMS3 fuel map configured to add additional injector pulse width based on load and rpm (boosted range operation). iEMS3 ignition map configured to provide timing retard as need in the boosted range minimizing overall timing changes in the CMR tune. iEMS3 configured to provide electronic boost control (turbo vehicles) based on load, rpm, and TPS position. iEMS3 configured to provide progressive water/meth injection based on load and rpm. The two additional digital outputs configured for user options such as shift light and additional accessory control. Wideband 0-5v analog output connected to the iEMS3 for datalogging. Additional required; Wideband AFR, EGT, and vac/boost gauge mandatory so you know what is going on. Water/meth injection to control combustion chamber temperatures when in boost, controlled by the iEMS3 or a progressive external controller. Water/meth injection point downstream of the IAT sensor Optional, would be to run an intercooler or add a second smaller water/meth injection nozzle upstream of the IAT (1/4 size of the main nozzle, about 24†away from the IAT). In this configuration, you would be running an optimized NA tune providing the best functions and features of CMR so that the ECU tune handles vacuum state operation. The iEMS3 eliminates many of the drivability issues associated with FI tunes and non stock MAP sensors, provides the missing closed loop boost fuel correction, allows less aggressive fuel addition and timing retard needed to stay safe in low boost closed loop situations, provides full control of the injectors for boosted range operation for fuel addition and timing retard, integrates boost control, water/meth injection, and any other user functions into one unit based on actual load and rpm conditions. Regarding your questions, How did you hook it up? The iEMS3 is plugged into the ECU injector output via male injector connectors which plug into the factory connector going to the injector. The fuel pulse runs into the iEMS3 where it is read and a new calculated fuel pulse is sent to the iEMS3 internal injector drivers and out to the injectors through female injector connectors plugged into the injectors. The ECU can not see any changes made to the fuel pulse after it leaves the ECU, and in actuality, it is not firing the injectors, the iEMS3 is firing the injectors through it’s own internal injector drivers. The ECU injector output is tied to an electronic injector load board so that it sees what it perceives to be the injector spike, so it is firing dummy injectors. The iEMS3 is spliced into the ECU harness near the ECU connectors. The connection that made are; Crank signal lead cut, In/Out to the iEMS3 (2 wires) MAP signal lead cut, In/Out to the iEMS3 (2 wires) O2 sensor return lead cut, In/Out to the iEMS3 (2 wires) iEMS3 chassis grounds (2) connected to the ECU chassis ground case mounting bolt iEMS3 signal ground leads (2) connected to the ECU signal ground lead (T- Splice) iEMS3 power lead to 12V+ battery terminal (constant on) Additional connections based on configuration For electronic boost control input, iEMS3 lead connected to ECU TPS lead (T- Splice)(also provides datalogging of TPS voltage/percent open based on sensor definition in the iEMS3 datalogging setup) For electronic boost control, iEMS3 output ground lead to boost control solenoid (12v+ power supplied to second solenoid lead by installer) For IAT datalogging, iEMS3 lead connected to ECU IAT lead (T- Splice) For Wideband AFR datalogging, wideband 0-5v output spliced to iEMS3 input lead For water/meth injection control, iEMS3 output ground lead to ground input lead on optional power switch (high load electronic relay) Two additional user configurable digital outputs for additional relay control, shift light, etc. What qualifications are expected for a tuner to know how to work with the iEMS3? I can provide input to your CMR tuner for the ECU tune, or put an individual in contact with a tuner to create the tune file. The iEMS3 tuning is very straight forward as all tuning is accomplished through direct value inputs in the ignition and fuel tables, ignition in degrees retard (-.2 increments), fuel in +/- milliseconds of additional fuel pulse (.2 milliseconds increments), O2 sensor control in fixed signal offset value (volts), MAP voltage control in volts (.3 volt increments on a single column rpm table), electronic boost control configured in psi target boost on a single column rpm table and percent of target boost on a single column TPS voltage table, water/meth pump control based on percent of duty cycle on a single column load table. It only corrects fuel? No, the iEMS3 has the full capability to control the following · Injector overall sizing. · Injector offset in three configurations. o All 8 injectors. o Left bank and right bank offset (useful to correct left/right fuel trim differences when using the factory fuel rails). o Advanced applications can correct individual injector offsets.· Fuel control from idle to full boost. · Timing retard from idle to full boost. · O2 sensor control for closed loop boost fuel correction. · MAP voltage control. · Electronic boost control. · Progressive Water/meth (or nitrous) injection control. · Data logging of all inputs and outputs connected to the iEMS3 as well as injector in, injector out, fuel added. · Additional user configurable data based on calculations in the iEMS3. On my 2008 Turbocharged Jeep Grand Cherokee (all stock internals, stock heads and cam), I am running a stock ECU and the iEMS3 is handling all fuel and timing corrections for the forced induction. At full boost, I am running the stock timing curves with a maximum of 3 degrees retard, and I have no knock retard present. I run an 11.2 to 11.8 AFR under wide open throttle boost, and I normally see around 12.2 for closed loop boost AFR. My EGT readings are around 1200 degrees at normal cruise (no boost) and around 1300 degrees in max boost. My IAT temps run around 140 degrees under full boost (spray downstream of the IAT only) and around 115 degrees (partial spray upstream of the IAT). If I spray upstream of the IAT, I will see about a .8 to 1 increment increase in AFR (richer) then when I do not spray upstream of the IAT, so I am still out to lunch on the benefits of upstream spray at this point. Questions, comments, discussion? Lets have at it and enter "The Tuning Matrix"
The only issue I have with adding in a piggy back ecu (so to speak) is that in the past, they proved to be inefficient. I can only comment on my old Split Second Ecu back in the days before PCM tuning was available. The PCM almost always seemed to over ride the piggy back ECU when it came to timing and part throttle. What I mean is, the PCM would add or subtract fuel/air and timing when the car was in boost, thus nearly zeroing out the ECU of its partial tuning. I am concerned that the iEMS3, PCM, and CMR combination will ultimately cause the same effects? Maybe with CMR additions, it will maintain its integrity with the iEMS3 partial tune? I still think that the part throttle issues can be resolved with the need of only one PCM and tuning tool, but that more R&D is needed to accomplish this? Using two computers to keep a car properly tuned scares the crap out of me. lol Great read and congrats on "Getting" there with your testing. I hope this is something that will prove to be viable and worthy for the long run...
Yes I understand where your concern comes from, I spent a long period of time tuning with SMT-6, Split Second, Unichip, etc, and they all did exactly as you described. The iEMS3 is totally different, it is not based on MAP sensor voltage skewing technology as all the previous mentioned units use. I have been running an iEMS3 for three years, and never has the ECU detected it or tried to change anything to offset it. It has direct control of the injectors, and there are no inputs to the ECU that are outside the expected parameters so it simply goes about its business doing the same thing day in and day out. When we take control of the O2 sensors, we do so in a manner that the ECU continues to see the expected voltage swings based exactly on the input voltage provided to the sensors by the ECU. We simply offset the output voltage in the normal expected range so all the ECU sees is a normal fluctuating signal that indicates a lean condition thereby providing the additional desired fuel in closed loop boost conditions. Unfortunately without the ability to reprogram the factory ECU functions (not simply the parameters) it can never understand a boosted condition and it can not be made to ignore it's target 14.7 AFR in closed loop. Those who have seen and used the iEMS3 have been amazed at the seamless nature of how it works with the ECU instead of fighting it as the old MAP sensor skewing piggybacks would do.
sweet. I am assuming you are NOT using a Boost A Pump application either? hmmm I think I need to get you in contact with a Turbo Kit maker friend of mine. This might help his TT application? As I believe that Turbo cars are greatly effected by the part throttle issue more so than a S/C'd car?
No, I am not using a boost a pump. Depending on the injector sizing used, the iEMS3 can take into account via the fuel map lowered fuel pressure and actually extend the effective range of the stock fuel system to a certain extent up to the limit of the injector duty cycle,
The problem with the closed loop and ECU's overriding piggyback units is that the PCM is hunting for 14.7 with feedback from the narrowbands. So the piggy back adds fuel and the PCM adjusts it out. This iEMS3 sits in between the narrowbands and the PCM. The PCM sees a normal signal (let's say 14.7) while the piggyback is adding a variable amount of fuel based on other inputs like boost. Same for the map signal, etc. It's the only way without changing the programming in the PCM -- "programming" means the executable code and not just the data constants and values in the tables. As far as two computers... think about it. There are about a dozen or more independent modules in our cars. Only two are mostly involved with the engine (PCM directly and TCM indirectly). In theory it's the same thing. The question would be the quality assurance and testing on the iEMS3 (hardware and especially software).
It would probably help to provide a discussion of the differences between the older “piggyback†tuners and the iEMS3 which we refer to as working â€In Series†with the ECU. I refrain from using the term “piggyback†in reference to the iEMS3 because of the negative connotations that piggybacks have had, in particular with Chrysler based ECU tuning. This is also due to the way the iEMS3 functions which is different than the older piggybacks. Before CMR/Predator and the iEMS3, there were few choices to use when tuning the Chrysler platforms. The fact that Chrysler went through a transition period from 03 up to 06 with the ECU’s did not help matters either. And Chrysler continues to change wiring layouts on what seems to be a whim, maybe just to keep people guessing! When I first started playing with FI on the new 5.7 Hemi in late 2005, the choices were slim. You had the SMT-6, and a few Split Second products. Later in 06, along came the Unichip which showed great promise (at least according to the manufacturer), but that also proved to be lacking in overall capabilities. All of these products were based on the ideology that if you make the ECU “think†the vehicle is under less load than it really is, then you can get the ECU to reduce the injector fuel pulse and thereby run a larger injector on a smaller fuel pulse. As load would increase, the tuning device would be programmed to allow the ECU to see more of the actual true MAP sensor voltage thereby driving more fuel through the larger injectors. But this still required the ECU to see artificially low loads across the power band. This is what I refer to as indirect injector control via MAP sensor voltage skewing. This process works on older applications that don’t have the adaptive capability that the modern ECU’s have. In our case on the Chrysler platforms, the adaptives would see the changes and begin modifying the fuel trims and other ECU tables to counteract the changes taking place from the MAP voltage skewing. Fuel trims from closed loop operation will slowly over time bleed over into changes in the open loop fuel tables as well. The cumulative effect was that after a period of time, you would lose the fuel correction you were trying to achieve, and in the case of forced induction, if you were not monitoring things, you would run a leaner and leaner AFR while in boost. Sometimes leading to very bad results. I would have to play around with the SMT-6 software every few days to keep my AFR’s somewhat in line, and constant resetting of the ECU to clear the adaptives out and revert back to the base fuel tables. Split Second controllers functioned in the same fashion. Then along came the Unichip, and at first, it seemed to be different, and the manufacturers rhetoric made it seem that it was different. In some aspects it was, but not in the way it went about it’s business. This unit actually combined several functions and had what appeared to be load based fuel and timing tables as it had an external MAP sensor connected to it. But in the end, it was based on the same old method of skewing the stock MAP sensor voltage, and the same ECU overwriting would take place. The bad part was that the software was proprietary, and only an authorized dealer could have access to it, and it also required the use of a special “tuning†box being connected between the Unichip and the laptop. This prevented the end user from being able to tweak the unit as could be done on the SMT-6 and Split Second controllers. The owner if even aware of what was taking place, could only remedy the situation by killing power to the ECU to force it back to the base tables, sometimes as often as once a day. The Unichip did try to address closed loop boost fuel correction by interfacing with the O2 sensors, but the technique used was not in line with how the sensors and the factory ECU function, so it was only partially effective and very inconsistent, The unit also tried to be a full plug and play device, but the manufacturer did not realize that Chrysler would change wiring schedules for the ECU’s frequently, so even though it was the same ECU, depending on the year and model, wiring would be vastly different. It became apparent rather quickly that this was not the “great solution†that the FI industry thought and hoped it would be, and it left a lot of individuals and companies in quite a situation. This was when I started looking deeply into how things worked and for another solution, and this research eventually led me to the manufacturer of the iEMS3. The company had been producing Stand Alone engine management systems for quite some time, and had modified their stand alone to function along with a factory ECU. Back in the late 90’s, their technology of the time was used in the KB Optimizer units used to drive supplemental injectors, and control timing on pre third generation Hemi engines in the dodge line at the time. This evolved later into the G1 and then the G2 “in Series†units which were also used on well over 1000 Viper FI systems. In 2006 the manufacturer came to the US and spent three days at my shop figuring out how the Chrysler ECU functioned on the Hemi and made both electronic and software changes so the unit would function as we desired on the Hemi platform. Once the changes were made and the unit installed, my 2006 Turbocharged 300C was fully tuned in under 2 hours. I still have the original G2 unit which has since been upgraded via a firmware flash to the current G3 system. Other than some additional tweaks over time to minimize fuel trims in a couple places in the vacuum range, the original file ran the vehicle without any adjustments for nearly three years and over 30,000 miles. The iEMS3 does not use MAP voltage skewing to try and control injector fuel pulse. As Quick mentioned, the iEMS3 is wired “In Series†between the factory ECU and the actual fuel injectors. The iEMS3 is programmed in “close to hardware†machine language so that it operates faster than the Factory ECU computations that are taking place. The ECU fuel pulse is fed into the iEMS3, and that is where it stops. Those inputs are tied to an electronic injector load board so the factory ECU is firing dummy injectors. The iEMS3 uses the calculated fuel pulse from the ECU as a base reference. It then applies the percent of original fuel pulse to use (based on old divided by new injector size) to calculate the new base fuel pulse. It then applies any additional injector offset values and any +/- values in the fuel table to arrive at the correct fuel pulse to deliver to the injectors. The iEMS3 has its own internal injector drivers, so it is firing the injectors directly. This technique allows the factory ECU to continue to do exactly what it was designed to do as far as environmental adaptives and adaptives for fuel pump voltage, aging components, etc., calculating a base fuel pulse for the iEMS3 to use that allows near stock like drivability in the vacuum range of operation. Since the ECU can not see what the actual fuel pulse is once it leaves the ECU, it has no idea as to what is taking place at the injector. Once tuned, the closed loop fuel trims will be in the single digits just as if the vehicle was stock. Unlike the older MAP skewing controllers, it is impossible for the factory ECU to see what is taking place and it can not override the changes because nothing is outside its normal operating parameters. In regards to the O2 sensors, we figured out how they function, how the factory ECU controls them, and how it uses the data stream coming from them. That allowed us to create a control logic so that whenever boost is present, the iEMS3 reads the dual fluctuating signal in real time, and applies a programmed offset value to the fluctuating signal thereby telling the ECU that it simply has a lean condition (which is true in our eyes), so it in turn increases the fuel pulse width to add the additional fuel. Since the signal it is provided is a normal signal it expects to see with the dual signal fluctuations intact, the ECU sees nothing abnormal and functions exactly as it is programmed. This allows us to achieve the desired AFR’s whenever we have boost present in closed loop operation. The iEMS3 also applies a voltage clamp to the stock MAP sensor output preventing a MAP voltage over range condition from taking place. The stock sensor is only used by the ECU in the vacuum range, so by using the stock 1 BAR sensor, the ECU has its normal range of function based on the full 0-5v signal output. In some cases, we have seen where the use of a 2 or 3 BAR sensor has left drivability issues, so in this fashion we avoid that potential. The broader the voltage span available for the factory ECU to use in the vacuum range, the more refined the normal operational changes are based on the values. A 1 BAR sensor usually operates from 0 – 4.9 volts, a 2 BAR from 0 – 2.45 volts, and a 3 BAR from 0 – 1.6 volts.
For those of you who missed "part 1" it's right here and a very good discussion providing the background for this. http://www.srtconnection.com/forums/showthread.php?t=14657 .
