« What makes a 330 hp smog-legal Acura RSX? | Main | A supercharged year »
October 18, 2005
"Closed loop" tuning for full throttle?
While I generally specialize on Hondata's K-Pro platform for Honda/Acura's K-Series engines, I also follow other ECU tuning platforms. My turbo-charged Chrysler PT Cruiser PT, for example, benefits from some very economical Mopar performance solutions. In addition, there are ECU piggyback systems, like the PPQ from Psi-Fi Motorsports.
In the www.ptcrew.com Cuiser enthusiast website, I cam across a very interesting article on PPQ tuning. ModMike there mentioned closed loop tuning. In essence, he promotes using the add-on computer to keep the car running in closed loop all the time.
For those less familiar with ECU/PCM tuning concepts, here are the basics: modern automotive computers essentially have two modes of operation, "closed loop," and "open loop."
"Closed loop" where the computer takes control and aims for what it considers optimal running conditions. Closed loop is general used during cruising where the computer tries to maintain the stoichiometrially ideal air-fuel ratio of 14.7 part sof air to one part of fuel. That theoretically allows for complete combustion of all gasoline and thus cleanest running.
In "Open loop" the computer abandons its "big brother" approach and instead reads fuel values from fuel tables. In general, open loop kicks in when you exceed a certain degree of vacuum in the intake, i.e. under medium to full throttle. There are other conditions where the computer switches into open loop, like when the engine is cold. Some ECUs have three or four open loop modes,
My view of the situation was always that in a programmable ECU, all you have to worry about in closed loop operation is that your short term fuel trim isn't too big. Short term fuel trim is basically the difference between the fuel tables, and what the computer has to actually allocate in order to stay at 14.7:1. As long as the computer does the work, why should we worry about short term trim? Because short term trim often affects long term trim, which means the computer "learns" to always allocate more or less fuel so that the short term corrections in closed loop are as close to zero as possible. A side note here: If one can edit fuel tables within the table areas generally used during closed loop operation, always aim for being a bit richer than necessary. That way, when an oxygen sensor fails, you're still safe because you're running richer instead of leaner.
In general, the ECU tuner's task is to extract the most from open loop operation by optimizing fuel, ignition and , if available, cam angles, under full throttle. In an NA motor, you generally want perhaps a 13.2 to 1 AF in open loop under full throttle. With forced induction, you'll be aiming for 11.5 to 12. With ignition, you aim for highest possible advance without detonation (that is a whole other project, one that depends on the specific knock sensor algorithms in a platform).
The problem often is that the ECU isn't entirely in open loop and still butts in. So if you managed to have a nice, consistent 12.0 AF in third, you may find that it's nowhere near as perfect in fourth or fifth gear. Then there's temperature compensation. Some ECUs do a good job at compensation for intake air temperature, others do not.
Truth be told, I had always thought the ideal tune would consist of running entirely in open loop. That way you'd have the computer out of the way, and the car would run optimally due to masterfully optimized fuel and ignition tables. I simply had never thought of doing it the other way around, of keep the car in closed loop all the way, so that you simply tell the ECU what AF you want under every load/rpm condition. It almost sounds too good to be true.
Unfortunately, I don't know enough about the PT Cruiser ECU/PCM to form a conceptual tuning strategy in my mind. I believe the PT uses what essentially amounts to a knock sensor-based ignition map overlay that probably starts with a base map, and then is modified via short term and long term knock sensor input. I don't know if it is possible to edit that map, or the ignition map, with the PPQ software (I also know that it doesn't work right in my car).
So... plenty of questions and issues, an interesting concept (closed loop running under WOT), and a desire to figure it all out.
Posted by conradb212 at October 18, 2005 2:04 PM