When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Skeese Informational - Open Vs. Closed Loop Fuel Tuning
Disclaimer: This is meant to be an informational guide to those looking to learn. The information within is correct. Please refrain from negative criticism, trolling, nit-picking, and general dickheadedness. If you wish to express any of the aforementioned emotions please PM me and I will gladly discuss your issue privately. Please keep the thread focused on the topic and don’t detract from others.
I constantly see questions and misinterpretations of the difference between open loop and closed loop tuning. There are some really in-depth seminars out there behind the actual mechanics of how it all works but I'm going to lay it out here in functional useful terms. While this will specifically discuss fueling, the open vs. closed loop theory and function is the same when used for other outputs. I'll be posting another thread before long detailing the actual mathematics of the PID calculation and how the Proportional, Integral and Derivative gains affect the output.
Open loop fuel tuning is where the ECU reads the fuel value from the main fuel table and then applies that value directly to the injectors. If it is an injector ms map then the physical time of the injection goes directly to the injector and the injector sprays the value from the map. If it is a VE tuning map the VE fuel map value goes into the VE fuel calculation where it affects the calculated fuel output which is then applied to the injector. Done. Fuel has been sprayed. Value in table is directly proportional to value output injector.
Closed loop fueltuning takes it a step further. The ECU reads a Target AFR from a target AFR table and compares this to the AFR reading it is getting from an O2 sensor in the exhaust. The ECU calculates the error between the O2 sensor sensed AFR and the target AFR, determines the necessary change in output to correct the error from what is on the main fuel table, and then applies the corrected output value to the injector for it to spray what the ECU has calculated will result in the Target AFR. This error calculation is driven by what is known as a PID controller.
There are a few important things to you need to understand regarding how these two tuning methods are tied together.
1. Open loop tuning serves as the basis for all tuning regardless of platform, ecu, engine, setup or application. The base fuel table that the ECU reads from during open loop fueling serves as the base for closed loop fueling as well.
2. If the main fuel table is not close to what is actually needed for the engine to hit a Target AFR, then the closed loop PID function will have to make a large correction. Large corrections will be slow and inaccurate which can cause you to overshoot or undershoot the Target AFR which can be dangerous.
3. The closer/more well-tuned the main fuel table is in open loop, the better the closed loop fueling will work. If the ECU is only having to make a small fueling correction the result will be much more accurate and you can run your closed loop system more aggressively which will speed up reaction and correction time.
So for example with closed loop…
You are making a pull where you have a map reading of 15 PSI at 6000 rpms and at this point your wideband is showing a 10.8 AFR. At 15 PSI and 6000 rpms the target AFR set in the tune is 11.8. The ECU calculates the error to be 11.8-10.8 = 1.0 meaning the value in the main fuel table is causing you to be 1.0 AFR point rich from the target AFR. It will then take the fuel value from the 15 PSI x 6000 rpm cell in that table and reduce it (using the internal PID calculation) in an effort to correct the 1.0 rich AFR and spray less fuel to move you closer to the leaner 11.8 AFR target.
Let's just consider 1.0 as a large error in this example, at it is a large error. If your closed loop PID controller is set too aggressive it will likely overshoot the required fuel reduction that it is calculating to take you from the rich 10.8 to the target 11.8 AFR, which will leave you leaner than 11.8. You can tell when this is happening in a log when you see that you are lean from the target AFR and the ECU is negatively trimming fuel. This was confusing to me at first as all I could think was why the hell is it PULLING OUT fuel when I am already lean from the target? The answer was the fuel map had placed me WAY RICH from the target, and the closed loop PID settings were too aggressive for an error that large causing overshoot.
If your closed loop system isn't setup aggressively it will likely undershoot the required change causing the resulting AFR to still be richer than the 11.8 AFR target value. The less aggressive your PID function is, the longer the correction will take, and given a high acceleration rate across the RPM columns/rows the fuel change may be applied too slowly to affect the correct range for which the calculated correction was to be applied. You can see this in a log when you keep adding/subtracting fuel in a set of lean/rich cells, but the resulting AFR there doesn't change. The fuel is skewed from the closed loop trim applied from the error calculation from a lower rpm range. You may still overshoot the target with an un-aggressive set of PID values, however the overshoot will take longer to occur, and longer to oscillate down to the target after leaving you in the leaner/richer than ideal range for longer.
The smaller the error that the ECU has to calculate a correction for, the more aggressive you can run your PID values without substantial overshoot in either direction. This is why it is IMPERATIVE that the base fuel map be tuned in open loop to closely reflect the fueling requirements of the engine across the full range of operation. Only once this base map is "close" to the need should you turn on the closed loop function which will then serve to account for on-the-road variances the car/engine is exposed to and it can do so fast enough to have a positive effect. Fast and accurate closed loop fuel control is directly tied to the open loop main fuel table's proximity to the actual fueling target.
The use of Long Term Fuel Trims, or adaptive tuning as some would call it, is simply an expansion on closed loop tuning where the ECU saves or banks the O2 corrections that were made to me fuel map during closed loop operation. In some cases the ECU will write the changes to the fuel map live time, and in others it will save the trims to a table that can be referenced on top of the closed loop control to reduce instantaneous correction size or applied to the base table at a later time. These are extremely useful tools, but they are only as accurate as the correction that is being saved which as we know is directly related to the size of the correction based on the open loop map's proximity to the actual target need.
Lastly, I want to explain what closed loop fueling WILL NOT DO. This really gripes me as you will soon know.
First and foremost it will NOT tune or run the car for you. Some believe closed loop fuel tuning is simply plug and play. They blindly believe that if you put it into closed loop the ECU will calculate the fuel and output what the car needs. This is not only wildly incorrect, but wildly lazy and ignorant to me. It is not a replacement for the open loop tuning process where you sort out fuel map based on AFR readings to sculpt it towards a target output. I've chased AFRs and fuel trims across a large number of logs trying to figure out why is the ECU pulling fuel if the wideband is showing a LEAN condition from the target only to find the main fuel table was so far off the closed loop function was grossly overshooting the fuel calculation such that the resulting trim caused the car to be extremely lean from the target.
YOU HAVE TO TUNE AN OPEN LOOP FIRST SO THE CLOSED LOOP HAS SOMETHING ACCURATE TO WORK WITH.
While some may have learned this already, I do find this type of conscise explanation very helpful for those of us who like to delve into our tunes amateaurly. . Skeese, would you consider doing beginner tuning tip posts such as this on varying topics? I'm not looking to become a tuner but I appreciate threads like this which allow me to have better understanding of what goes into putting together a properly running vehicle.
I try to post stuff like this when time allows. My life is so insanely busy right now outside of all things rx7 that most of this stuff I write bits and snippets of at a time and then at some point feel is complete post. I'd like to make an informative post to de-mystify rotary timing and split as well as one that goes into detail on how to take the haltech basemap and mod it to suit your car as well as some best-practice tuning modifications you can make to the supplied map. Hopefully time will allow shortly. Having a newborn, a demanding job and a half finished rx7 of my own all combine to be a HUGE time suck...but all good. If you have any direct questions I'm more often free via whatsapp or facebook IM to chat and sort tuning related **** out, just PM me.
I've started tuning my car and its kind of fun after you learn the basics and the why behind fueling and timing. Just recently did a log and drive in open loop mode and when I play back the log tracing afr's, when I highlight a cell, it still shows a quick trim value, which I would think indicates that the car was still running in closed loop fueling. Did I not set the ECU to open loop, I changed the closed loop drop down to open loop before I started logging. Is this normal?
Furthermore, I have fuel cut activated and I think that's why I see a lot of lean conditions in the table, mostly in high vacuum areas. I'll have to either turn fuel cut off in decel or when logging jump to neutral during deceleration.
Anyway, slowly learning and staying out of boost for now
09-06-18, 07:00 AM
