Cool FlashPro Easter Egg - Two modes of closed loop!

[Gernby]

I made an accidental discovery yesterday that there are 2 different modes of closed loop. There's a part throttle closed loop where trims are applied to target the AF values in the Closed Loop Lambda tables, and there is also a WOT closed loop where trims are applied to target the AF values in the WOT lambda adjustment tables. I always thought that WOT lambda adjustment tables were just used in open loop, but they aren't! The WOT lambda tables are used for open AND closed loop!

This provides the ability to set your closed loop AF targets in whatever crazy way you want! In my case, I want my target AF at idle to be rich, since my fuel injectors don't have a consistent flow at the small pulse widths required for 14.7, so I want to target 14 at idle. In order to get better gas mileage, I'd like to target 15.5 at light loads, and slowly get richer as load and RPMs increase. All this can be done by using the WOT closed loop!

To do this, I set the closed loop target lambda values to be my peak AF (15.5), and set the MAP WOT determination pressure tables so that it goes into WOT above 100 mBar (effectively always), and selected "Use MAP to determine WOT". Then I set all the AF targets that I actually want to achieve in the WOT lambda adjustment tables. The datalogs confirm that the ECU follows the WOT lambda table very well until the ECU goes into Open Loop. I haven't figured out exactly what triggers Open Loop, but it seems to be the target AF value. I suspect that the ECU goes into open loop whenever the target AF goes richer than about 13:1. See attached datalog and calibration.

A couple things to note are that FlashPro Manager does NOT use these target values for its fuel adjustment suggestions. Therefore, you have to adjust these tables to agree with the settings in FlashPro Manager BEFORE you do the datalogging, otherwise, all of the fuel suggestions will be very wrong. Another important item is that the "regular" part throttle closed loop AF target needs to be as lean or leaner than the leanest value in the WOT lambda adjustment tables, it will go with the richer of the 2 targets.

[ajpturbopittsburgh]

Ok seems cool. But you said you aren't sure what exactly triggers open loop right?

I thought WOT = Wide open trottle i.e having the gas floored. Which means open loop or so I thought


You have your TPS WOT determination low set at 59% so that tell me below 59 you are in closed loop but using the WOT lambda tables and above 59 open loop right? In which case you normal fuel table values.

And on the high cam you have a TPS of 60% for determination of WOT. Which tells me above 60 you are in open loop and using normal fuel table values and below 60% TPS closed loop but using WOT lambda adjustment tables?

I just noticed that I don't have the TPS level switch for WOT under closed loop options. Is this because I'm using a MAP calibration?

Now that I look at the software should the WOT lamda adjustment tables be called CLOSED LOOP PART THROTTLE THROTTLE TARGET LAMDA? I thought basically if you were in closed loop you could only be around the labmda that you have programmed into closed loop parameters for target lamda based on coolant temp?

This all just sounds like WOT closed loop basically elimanating the need for open loop? So Gernby you're saying you can adjust lambda in closed loop based on TPS and load and rpm? Jeeze!

[Gernby]

I really have no idea if the other FlashPro'ed ECU's will behave the same as my S2000, but I would expect them to.

I have my WOT trigger set to MAP instead of TPS, so it really has nothing to do with TPS percent. I'll probably try the TPS trigger just to make sure it behaves the same. However, I think a MAP trigger is better.

I haven't done any testing on the high cam, since I want to minimize risk to the motor.

I kind of think that the existing Closed Loop Target Lambda tables need to just go away, and the name for the WOT Lambda Adjustment screen should be changed to "Target Lambda" All of the target values in the WOT Lambda Adjustment tables are very important regardless of whether the ECU is in open or closed loop. The only difference in behavior between open and closed loop with regards to those targets is that the ECU always applies a "trim" to the fuel table based on the percent difference between 14.7 and the target, but only in closed loop does it also apply S.Trim and L.Trim based on that target. So if your fuel tables are well tuned, your actual AFR's will be very close to the WOT adjustment targets regardless of whether you are in open or closed loop.

[Gernby]

On second thought, I guess the Closed Loop Target Lambda tables are still important during warmup, so it seems that it would make more sense to call them "Warmup Target Lambda".

[ajpturbopittsburgh]

.

I have my WOT trigger set to MAP instead of TPS, so it really has nothing to do with TPS percent. I'll probably try the TPS trigger just to make sure it behaves the same. However, I think a MAP trigger is better.
..

Oh I see, I thought you could use MAP and TPS.

What version are you using? I don't have the TPS for WOT tables, not sure why

[ajpturbopittsburgh]

Ok now I'm kinda confused. Under the contents and reference for the software Hondata puts a special note for the lambda WOT adjustment tables saying don't set the values above 12.50 a/f or .85 lambda or the ECU will stay in closed loop all the time. But right above that note all the values are above that! And all the stock calibrations I open including my own, which I've never touched are all also above 12.50 a/f.

[Hondata]

This is the way Honda uses the WOT correction tables - they alter the target lambda at WOT too. We do not recommend that you force the ECU to run with WOT setting all the time, for the simple reason that the ECU is not designed to run like this. Various functions will never run when the ECU is in WOT mode - which may be important, like purging the evaporative canister so that it does not fill up with fuel over time. There's make other functions, than is one example. Another reason is that we plan to alter the WOT compensation tables in the future, so you'll want to stick with conventional settings in the meantime.

If you want lean cruise you can do that through the closed loop target lambda tables. From my experience, 14.7 gives better fuel economy than 15.5, but it will vary with the vehicle.

ajpturbopittsburgh: The qualifier should be 'don't set the WOT compensation tables over 0.85 / 12.5 in regions where the ECU is under full load'. From the factor the ECU will run closed loop to 3000/3500 rpm, and that is evident in some of the calibrations.

[mucter]

original post

A lot of us tuners have known about this for a long time. I ran my car using this functionality for over a year to tune out misfires on my ID1000's in the midrange at part throttle by targeting a 13.7 AFR in closed loop from 4000-4500 rpm at 60kpa and above. I had a very specific misfire bs issue with my ID's, it was strange.

I agree with what Hondata said though. It should only be used to fix small / specific scenarios... you don't want to keep the car from running in standard closed loop for extended periods.

This isn't an easter egg. It's just how the system works. The MAP WOT determination table defines when to use the WOT Lambda Adjustment tables, and the WOT Lambda Adjustment table values tell the ECU what to target. If the target is richer than what the ECU knows is reliable on the stock O2, it goes to open loop. It's as simple as that.

[ajpturbopittsburgh]

original post

.
This isn't an easter egg. It's just how the system works. The MAP WOT determination table defines when to use the WOT Lambda Adjustment tables, and the WOT Lambda Adjustment table values tell the ECU what to target. If the target is richer than what the ECU knows is reliable on the stock O2, it goes to open loop. It's as simple as that.

Do you mean the target in the WOT tables or the target in the closed loop menu?

Still not clear to me when the car goes into open loop? I thought if you are at WOT that means open loop.

And Hondata, why do the WOT tables stop just at about boost? If you are boosted wouldn't it be good to have those tables extend into to boost to help alter a/f's in boost too if your car is poorly tuned?

My car is tuned pretty well I think, but I have my WOT pressure set at about 930 mb or just before boost so I target about 14.3 when out of boost then I thought when I exeecd 930mb and get into boost I go into open loop and use the normal fuel tables. But I thought I could do what Gernby suggests and progressively richen up the fuel the closer I get to boost by using the WOT lambda tables, but as it stands now I think I either have to target 14.3 and stay in closed loop and I can't change that unless I lower the WOT pressure then all of my pre-boost fuel tables will have to be tuned much better becuase I would be in open loop and I know my tuner didn't spend alot of time because I thought it was silly to be in open loop while in vacuum or prebooost.

[ajpturbopittsburgh]

original post

.

. If the target is richer than what the ECU knows is reliable on the stock O2, it goes to open loop. It's as simple as that.

Mucter I don't think it works like that.

I just read in the contents section " At WOT the ECU will switch to open loop and use the WOT lambda adjustment tables."

So with that being said, if being at WOT makes you run open loop and the WOT tables are being used, then at what point are you just simply using the normal fuel tables with a MAP based tune when boosting?

This was under closed loop parameters and for the explanation for MAP WOT DETERMINATION. So that tells me that when at WOT I am in open loop. And for me to use the WOT tables I need to use a much lower WOT pressure than 930mb or otherwise I will be in closed loop all the time until I just about hit boost.

[Gernby]

Thanks for the feedback! It's good to know that this would prevent the ECU from performing some functions. However, it would be nice to know which functions, and what percentage of running time would be required to keep it healthy.

I really don't care about a lean burn enough to mess with it right now, but I do want to improve my idle with these injectors. I altered the strategy above to only trigger the WOT mode prematurely below 2000 RPMs, so that I can target a rich AF for idle. Hopefully that will smooth out the idle and eliminate the misfired DTC's. Is this a bad idea?

[Hondata]

I'd still keep the ECU running normally (not WOT) at low rpm.

If the injectors don't idle well, and you have the dead times set correctly, it is from the spray pattern. Unfortunately the solution is to change to a different injector with a better spray pattern. Without know what injector you are using now, I'd recommend the RDX injector if NA, and the DeatschWerks for forced induction.

As a side note, the manufacturer supplied dead times for injectors often are nowhere near accurate - we found many were calculated electrically rather than by measuring when the injector actually supplies fuel.

[Gernby]

I'd still keep the ECU running normally (not WOT) at low rpm.

If the injectors don't idle well, and you have the dead times set correctly, it is from the spray pattern. Unfortunately the solution is to change to a different injector with a better spray pattern. Without know what injector you are using now, I'd recommend the RDX injector if NA, and the DeatschWerks for forced induction.

As a side note, the manufacturer supplied dead times for injectors often are nowhere near accurate - we found many were calculated electrically rather than by measuring when the injector actually supplies fuel.

My injectors are Siemens Deka 630's. I spent a ton of time on the lag times, which wound up being very different from what the supplier provided. My idle is fine when cold, but as the engine warms up and the idle RPM falls, the injector pulse width drops down to about 1.4 ms and the S.Trim starts bouncing all over the place. The manufacturer says the minimum pulse width is 1.5 ms, so I figured that is the problem. When I enrichen the idle and raise the RPM a bit to get the pulse width above 1.5 ms, the idle is fine.