Closed loop vs Open loop tuning

XoX1de

Member
:
2000 Mazda 626 LX
I heard some people here have unsuccessful tuning with e-manage. The ecu "fightback" modified information by those piggyback ems.

This is because ecu is working in closed loop mode which will maintain a afr of 14.7 no matter how big your injectors are. This is based on feedback of narrowband oxygen sensor. ECU use long time fuel trim and short time fuel trim for the calculation.

Here's a detailed explaination quoted from here:


" I am going to assume the reader has a basic understanding of how modern GM EFI works. However, I will provide a brief overview. In general you have three unique modes of operation, open loop, closed loop, and power enrichment. Open loop is used for startup running and the PCM will remain in open loop until a specific operating temperature is reached. In open loop, the PCM takes the reading from the Mass Airflow Sensor (referred to as MAF or sometimes MAS), the Manifold Absolute Pressure sensor (MAP) and an RPM reading. Based on these factors plus air temperature, the PCM makes an educated guess about the correct amount of fuel needed for engine demands. In open loop, the primary function of the PCM is to keep the engine operating despite the fact that temps are not optimal yet. The PCM must also keep emissions as low as possible, because startup is the time of greatest emissions production. As a side bar, this is why the Air Injection Reaction pump runs only for a few minutes upon startup. its job is to pump unburned oxygen into the exhaust, which will then combine with fuel remnants in the catalytic converter, further reducing emissions and getting the converter up to operating temperature much more quickly. In Open Loop mode, the oxygen sensors (we?ll call them O2s) are not yet up to full operating temperature either, and are therefore not used by the PCM for fuel corrections. Thus the term "open loop."


Once an appropriate operating temperature is reached, the PCM then brings the O2 sensors on line and uses them for feedback about the fuel adjustments it is making. The O2s complete the feedback loop, thus it is called Closed Loop operation. In Closed Loop, the PCM takes dozens of readings from the O2 sensors each second and uses that data to fine-tune the amount of open time the injectors are on. In this system, the MAF, MAP and RPM are used to make a base fuel guess, then the O2 sensors give the PCM feedback on how well it has guessed, and it adjusts. During Closed Loop, the PCM is constantly trying to maintain 14.7:1 air to fuel mix. 14.7:1 is important because that is what is called the stoichiometric ratio for air and gasoline. In other words, it is the most complete burning (i.e. cleanest from an emissions standpoint) mixture ratio. If a car was equipped with a wide band O2 during closed loop operation, you could see the air/fuel ratio swapping back and forth between 14.4:1 to 15:1 as the PCM adjusts back and forth constantly.

The adjustments from the PCM are not just thrown away however. The adjustments are fed into a storage system that is used for future reference. For OBDI cars these are called Integrators (INT) and Block Learn Multipliers (BLM). In OBDII, these are referred to as Short Term Fuel Trim (STFT) and Long Term Fuel Trim (LTFT). For our purposes I am going to refer to them under their OBDII names, because they are more meaningful. Each "bank", or side of the engine has one oxygen sensor (in OBDII there are two per side, on in front of the catalytic converter which is used for fuel adjustments, and one behind the converter, which is simply used to verify the cat is there and is working). Each bank also, therefore, has its own STFT and LTFT values and the PCM can adjust injector pulse width (simply the amount of time the injector is open each engine revolution) independently on each side.


The STFT is simply a correction value for the fuel requirements which changes several times per second, which then average out into the LTFT value, which is recalculated every 20 seconds or so. These values are fed into a grid of 16 cells, with each cell corresponding to a different range of RPM X MAP reading for varying driving conditions. By using this system of cells and LTFT values, the PCM can be more accurate in its fuel adjustments. All of this fuel trim data is deleted when power is removed from the battery, which is why you often hear about the PCM having to ?relearn? a mod. The PCM can make permanent adjustments to overcome changes in fuel or airflow caused by engine modifications?.until the battery cable is removed. Then the learning process begins again.

OBDI cars operate on a system of INT and BLM values from 108 to 160, with 128 equaling a perfect 14.7:1 mixture, with 108 being max rich. In OBDII, the system is percent based in terms of read out. The PCM is still using the 128 based system for calculations, but you and I see readouts in the form of percentages. 0 is 14.7:1, -10% would mean the engine is running rich and the PCM is removing 10% from the fuel requirements. Next time the PCM encountered that same RPM/MAP situation, it would look in the table for the appropriate cell and see that ?10% fuel is required.

The process of tuning a PCM is the attempt to eliminate this learning curve so that engine performance is not poor until the PCM re-learns the modifications. Also, if the modifications are significant enough, as in a camshaft change, the PCM may not be able to adjust enough to overcome the modification. Camshaft changes, more than any other modification, alter the efficiency at which the engine converts air and fuel to power. The optimum RPM ranges for max efficiency will change because of that, and the PCM can sometimes run out of room to adjust.

Now, you might say at this point, ?If my car can adjust within a certain range, why would I need tuning?? The answer is not so simple. Yes the PCM is designed to compensate and learn, but even with the learning the car can run rich or lean. I have never gotten a complete answer from anyone but it seems as if even the learning capability is not complete. For example, after my GTP Stage 1 heads and CC304 cam install in 2001, I did not tune any part throttle variables. When it came time for emissions testing, the car failed?miserably rich. When scanned it showed Bank 1 ?9% and Bank 2 ?5%. Looking at those values, you would think the PCM was pulling fuel and thus compensating?but the car was well past the hydrocarbon limits. The car only passed after I tuned it to near 0% LTFT?s. This is just one incident but it shows that even the PCM?s learning capability is limited in its effectiveness. The best way to achieve optimum driveability at part throttle and to ensure emissions compliance is to tune for 0% fuel trims. In my personal opinion if you are expecting the PCM to compensate more than 3% lean or rich you are outside of its range off effectiveness and you need to get the PCM tuned.

By now you should have a good understanding of how the PCM governs fueling for your engine. In Part 2 we will discuss what parameters are commonly ?tuned? and the methods used to tune for maximum power. NEXT: On to Part 2.

In part one (see last issue) I discussed how the PCM manages fueling for the engine, and the various sensors and software schemes employed to do so. In Part 2, I?m going to give you the ?meat? of tuning a GM fuel-injected engine.

At a minimum, this article should give you the knowledge to understand what a tuner is modifying in your PCM to give you that max power you are looking for. And, after all, a better understanding of how stuffworks is half the reason we enjoy these cars so much in the first place.

There are many, many variables that can be tuned in the PCM to govern things from fueling, spark curve and idle speed to things like emissions controls, fault codes and even transmission shift variables for automatic equipped cars (GM vehicles prior to 1994 did not put transmission control in the engine computer, hence the distinction between a 94 + Powertrain Control Module and a 93 & earlier Engine Control Module). There is no way I can address the various aspects of tuning each variable, or even most of the common ones. Instead, Part 2 covers how to tune your car for maximum power at full throttle, whether on the dyno or at the track.

Most obviously, to tune the PCM, you must have some kind of scanning software to read sensor outputs such as O2 sensor readings, knock counts, Fuel Trims, etc. Assuming you have modified your car enough that you think you need tuning, sourcing some kind of scanning equipment (whether laptop based or stand alone handheld device) is the next step. If you are going to tune the car yourself, you?ll also need software like LT1/LS1 Edit or Tunercat, or take it to a reputable performance shop with a chassis dyno and tuning capabilities.

When you stomp on the gas, most of you know the PCM commands WOT, or wide open throttle, operation. This condition is governed by throttle position and varies based on RPM. When you go WOT, the fueling scheme changes somewhat, and becomes a kind of mix between the closed loop and open loop modes. The new mode is called PE, or Power Enrichment.

When you enter PE, the PCM takes the last known Long Term Fuel Trims and uses them for the base fueling, then looks in the Power Enrichment table for a percent value to add or subtract fuel from that baseline. The most important thing is to have the LTFT?s near zero or slightly rich before entering PE. If the trims are zero or rich, the PCM defaults back to zero for the baseline before adding the PE fuel. If the LTFTs are lean, the PCM will continue to add the extra fuel it was using to compensate for the lean condition, in addition to the PE fuel, to prevent you from blowing up your motor. The point here is that the car should have a good part throttle tune first, otherwise future part throttle tuning changes will throw off your PE fueling.

Again, it is important that part throttle fueling be tuned for near-zero LTFT?s before beginning to tune for power. The simplest way to tune the fuel trims is to add or subtract from the Injector Constant variable, which is a value telling the PCM how big your injectors are. If you are running rich, and you increase the injector constant from 24.8 lbs/hr (stock setting) to 25.3 lbs/hr, the PCM now thinks the injectors are about 2% larger and will dial back the injector open time in order to provide the correct amount of fuel. Likewise, if the car is lean across the board, decreasing the constant can richen it up.

We?ll assume, from this point on, that you are dyno tuning, but the same principles will apply if you are track-tuning. Once the Fuel Trims are where they need to be, you need to get a baseline dyno pull to evaluate where the current tune is and where you need to tune. I strongly recommend utilizing a dyno that has a wide band O2 sensor. The reason for this is the wide band is much more accurate than the O2 sensors on your car. OEM O2?s are only accurate around .450 mV, which is roughly the 14.7:1 air fuel range. Many people will tell you that .880-.900 mV is where you want your O2 readings to in order to reach 12.8:1 or thereabouts (12.8:1 to 13.3:1 is optimal for power in a naturally aspirated car, for forced induction, you will want to be closer to 12.5:1), and while this is generally true, the O2 sensors on your car are not consistent enough to be trusted ultimately. I If track tuning, always try to start with a dyno tune first to get the fueling as close as possible, making only incremental adjustments at the track based on the on-board O2s.

The dyno?s wide band will give you an air-fuel ratio readout. Depending on whether you are rich or lean on the first pull, fuel will need to be added or removed via the Power Enrichment table in the PCM. There are multiple cells in the table, each one representing about 400 rpm from 400 all the way up to 7000 rpm. An ?8? in a given cell would represent +8% fuel in PE mode. A ?-5: would represent ?5% from the base fueling in PE mode. I recommend gradual changes, 3% increments is a good rule of thumb, until you get a consistent 12.8:1 reading on the dyno graph all the way through the RPM range. 12.8:1 is the preferred air-fuel number to shoot for on the dyno. Since a dynamometer does not load the engine the same as does the car moving down the track, it is a good idea to tune richer on the dyno, because the greater real-world loads will tend to lean the car out more.




As to makin e-manage work, you must find a way to make ECU work in open loop mode when boost, either by hackin the TPS into 100% or Maxout the MAF... the AEM fuel/ignition controller have the function for oxygen sensor output, it makes the closed loop tuning easier.
 
Last edited:
so in conclusion, even if we wanted to tune closed loop position we cant because of the pcm is too strong? guess it makes tuning a little easier for everybody, since you only tune after 4000 rpms, lol.
 
you can, but you dont have much control. As long as u send a 14.7 stoichio* The ecu will be happy with it. ecu may not like a sensor signal that dont switch forward and back and throw a cel, but im not sure on this. stock ECU is very smart, they are soooo smart, but not useful on performance purpose.

So here's the first idea, find a way to trick ECU in closed loop mode with o2 signal send to pcm a .5 to .6v = 14.7. ECU will keep the fuel trim stable. And boost the fuel pressure, as injectors pulsewidth stays always same, more pressure = more flow, and you will get a richier mixture. Because ecu is still reading 14.7 so it will not bother to adjust the LT-FT or ST-FT. If you boost fuel pressure with a FMU in closed loop, it wont work, because without trickin ecu, o2 reports a richier mixture and , oh s***, switch to a narrow pulsewidth to maintain 14.7.

The second idea, find a way to send lean signal, and the ECU will automaticaly adjust the fuel trim to maintain 14.7. If we can send a 0.15-0.2V, thats close to a 16-17 AFR, ecu will read the engine is runnin lean and dumps more fuel in to maintain a stoichiometric mixture, but in reality its runnin richier than 14.7, maybe 12 or 13. The problem here is, we have to find a way to make a variable output , if we stay sending a .15V the ecu will be confused, and the trim will be out of range, a cel will result. AEM's FI/C have the function, a table of input o2 voltage vs RPM range, it intercepts the o2 signal and change it to the volt you want at given rpm range. The good thing is, the output switches back and forward just like a narrowband 02 sensor.

sry for my eng skill. hope you understand it
 
Last edited:
The voltage clamp is what boosted P5's use with the MPI if I recall correctly...The Flying Miata Voltage clamp...Not 100%, so if anyone knows...bump
 
yup i have dat clamp buht in parralel wid Emanage, buht now my problem is i run 2 rich. the clamp works
 
my splitsecond afc can not tune closed loop. like the guy said, the pcm is too strong.
 
Ok her it goes I got 97 probe with fs9 setup an machined for dist but loser 3 cyclinder couple wired think got grounded don't know if fried computer or rings new head an gasket no compression tester please help guys
 
Back