FI Open\Close Loop

[Al Navecky Jr]

Can someone explain open loop, close loop and how it relates to a Power Commander? I read about this before but cannot find it.

Thanks

 

[smiller]

Open loop refers to a fuel injection system selecting a fuel mixture from a pre-programmed map based on various external conditions (engine temp, air temp & pressure, intake air flow, etc.). In closed-loop operation the ECU measures combustion efficiency via an O2 sensor in the exhaust and uses that information to precisely control fuel mixture. Closed loop can provide more accurate fueling but is generally only usable under steady-state operating conditions (such as a steady cruise speed) due to the inherent delay in O2 sensor response time. Most modern fuel injection systems use a combination of open and closed-loop modes based on operating conditions. There's a lot more to the details but that's a basic explanation.

 

The Power Commander (for BMW models) is usually used to vary fuel mixture from the stock calibration in order to improve driveability and is unique in that it can provide adjustment in the closed-loop operating mode, which is the most effective way to address surging issues. There are other devices that perform a similar function but they can only add fuel in the blind. This method works as well but is IMO a sub-optimal solution. Here is a thread with some more information on the topic.

[Al Navecky Jr]

thanks for the info. I have PC and need to make some changes becuase it is running too rich. so i am going to make some changes to the close\loop setting.

[Joe Frickin' Friday]

Open loop refers to a fuel injection system selecting a fuel mixture from a pre-programmed map based on various external conditions (engine temp, air temp & pressure, intake air flow, etc.). In closed-loop operation the ECU measures combustion efficiency via an O2 sensor in the exhaust and uses that information to precisely control fuel mixture.

 

The "loop" in open/closed-loop refers to information flow: in closed-loop operation, information about the results of previous combustion events is looped back upstream to be figured into the calculations of how much fuel to inject for subsequent combustione events. All of the other information used in open-loop operation - throttle position, air temp, air pressure, etc. - has not flowed in a loop from downstream of the engine, which is why it's not considered closed-loop operation.

 

The O2 sensor in the exhaust measures (funny enough) exhaust oxygen content. Assuming complete combustion, the exhaust oxygen content is an indication of air-to-fuel ratio: very little exhaust O2 infers a rich mixture, while substantial O2 content infers a lean mixture (a stoichiometric or "stoich" mixture is neither lean nor rich, but has air and fuel in exactly the right proportions to consume all available O2 and all available fuel).

 

Most O2 sensors in production vehicles are "switcher" type sensors: they put out a constant high voltage (800 mV?) anywhere lean of stoich, and a constant low voltage (100 mV?) when the mixture is anywhere rich of stoich, and there is a rapid transition in sensor output voltage as the mixture crosses stoich.

 

With those sensor output characteristics it's impossible for the fueling system to hold a mixture steady exactly at stoich, so the system deliberately dithers back and forth from rich to lean. When it's rich, the computer gradually reduces injector pulsewidth (fuel quantity) over a number of cycles until the O2 sensor suddenly indicates "LEAN," and then the computer starts increasing injector pulsewidth until the sensor indicates "RICH," at which point the process repeats. Because of the time it takes from a particular fuel injection event until the results of that injection event reach the O2 sensor, a small amount of time elapses between when the injected mixture transitions across stoich, and the time when the O2 sensor registers that transition in the exhaust. It's RPM and throttle-dependent, of course, but this time is typically such that the rich-lean-rich-lean process typically repeats a couple of times per second.

 

An analogy would be a cruise control system that is aiming for some target speed: if the vehicle speed is sub-target, it puts the pedal to the floor, and if the vehicle exceeds the target speed, it hits the brakes. At highway speed, it takes a moment to accelerate to "TOO FAST", and it takes a moment to decelerate

to "TOO SLOW."

 

Modern exhaust catalysts require very precise regulation of air/fuel ratio right around stoich for maximum effectiveness. It's just not possible to do this well enough with open-loop fueling, but this is where closed-loop oepration excels. In fact, the catalysts actually like it when the air-fuel ratio dithers back and forth just a bit: when the mixture is lean and there's excess O2 (and NOx) in the exhaust, the cat strips off the oxygen and hangs onto it, and when the mixture is rich, it uses that O2 to oxidize the raw hydrocarbons and CO. And of course, that's exactly what the closed-loop fuel system is doing: rocking the A/F ratio back and forth across stoich, a few times per second.

 

I know next to nothing about the Power Commander in particular, but anything that's going to alter the A/F ratio of an engine operating in closed-loop mode has to somehow interfere with or break that loop.

 

In the other thread Seth linked to, he indicates that the Power Commander allows continued closed-loop operation. I'm guessing here (can someone confirm?) that it (the PC) is supplied with a wide-band O2 sensor (unlike the "switcher" o2 sensor, this will have a smooth voltage-versus-mixture curve), and that you can supply the PC with a target A/F ratio and it will output a switcher-style voltage signal for the Motronic, with the transition happening around the target A/F ratio as measured by the PC's O2 sensor.

[smiller]

There's a lot more to the details...

And now Al has them... Good explanation, and also hints at some of the issues that a wideband sensor tries to address. A wideband sensor provides a constant output vs. the cycling operation of a standard O2 sensor and thus enables a faster response from the ECU to changing conditions. A wideband sensor can also directly measure A/F ratios from about 10:1 to 20:1 (vs. only being able to indicate whether a mixture is above or below 14.7, as with a standard sensor) allowing a variant target A/F ratio to be accurately maintained. Here is a technical description of how wideband sensors operate for those interested.

 

 

In the other thread Seth linked to, he indicates that the Power Commander allows continued closed-loop operation. I'm guessing here (can someone confirm?) that it (the PC) is supplied with a wide-band O2 sensor (unlike the "switcher" o2 sensor, this will have a smooth voltage-versus-mixture curve), and that you can supply the PC with a target A/F ratio and it will output a switcher-style voltage signal for the Motronic, with the transition happening around the target A/F ratio as measured by the PC's O2 sensor.

That is what I had originally guessed as well but not actually how it works. The Power Commander does indeed have its own wideband O2 sensor (replaces the stock sensor on the bike), and has a direct connection to the TPS, and controls the output to the injectors. This allows the PC to have primary control over engine fueling in closed-loop operation (and allows it to fully utilize the advantages of a wideband sensor.) The Motronic is normally kept spoofed into open-loop via its O2 sensor input (of which the PC has control) and when open-loop mode is required the PC utilizes the output of the Motronic (which has access to external data such as ambient temp & pressure, engine temp, etc.) and then modifies that output (if desired) based on PC's user-controllable map. So the Power Commander arbitrates between open and closed loop (by looking at the raw TPS output among other things) with the Power Commander alone controlling injector pulse width in closed loop and working in conjunction with the Motronic in open loop. (The preceding info was obtained from a product development guy at Dynojet.)

 

In my experience the Power Commander provides a significant improvement over the stock Motronic in terms of driveability. Whether this is due to the wideband sensor, or by simply running a slightly richer mixture, or both, I can't say. The rather crude single-lead port injection of the oilhead FI system remains but the PC at least upgrades the electronic portion of the system to modern capabilities.