Closed Loop Mysteries

s300 and SManager software questions & answers
Post Reply
IdaSol
Posts: 21
Joined: Tue Jul 29, 2014 10:20 am

Closed Loop Mysteries

Post by IdaSol »

Hello! I have a few questions regarding the documentation of "Rate of Change" in regards to Closed Loop operation. How can I make it so the closed loop sweep is smaller? Currently it sweeps from about 13.6:1 to 17:1 AFR and it's very noticeable with cruise control on (small jerk that matches to the AFR sweep). I currently have my closed loop Min/Max S.trim adjustments set to -30/30 respectively. My IAC duty cycle is set at Normal. I have not yet attempted to set the idle as described in the documentation because I am the first, and only owner of the vehicle. The throttle body has never been touched.

Also, I have noticed that when I let off of the throttle to idle at a stop light the car will drop below idle RPM (Super Lean), then it will get rich and rev up to about 1300RPM until it settles back down at stoichiometric and idles at 1000RPM (Sometimes this happens twice). If the stop light is a longer stop light the car will fluctuate it's AFR from about 13:1 to 16:1 (increasing and decreasing RPM between 1000-1300) making it sound like the car is idle surging slightly.

I will attempt to get a log and calibration file, but capturing the event described is difficult without having a super long log to look through.

Thanks for all the help.
IdaSol
Posts: 21
Joined: Tue Jul 29, 2014 10:20 am

Re: Closed Loop Mysteries

Post by IdaSol »

Here are the logs and calibrations.
You do not have the required permissions to view the files attached to this post.
User avatar
Hondata
Site Admin
Posts: 10615
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Closed Loop Mysteries

Post by Hondata »

The closed loop rate of change is ok; the problem is a slow response from the wideband.
Hondata
IdaSol
Posts: 21
Joined: Tue Jul 29, 2014 10:20 am

Re: Closed Loop Mysteries

Post by IdaSol »

That makes no sense. I'm not using my wideband for anything but logging purposes. I'm still using a Honda OEM narrowband sensor, mounted at the same place the wideband is. (Right before the flex pipe)
Boost4Breakfast
Posts: 9
Joined: Sat Feb 14, 2015 4:31 am

Re: Closed Loop Mysteries

Post by Boost4Breakfast »

Did you set base timing?
Dial in your afr's in open loop and get them as close to where you want them. Then enable closed loop. I tend to use a small min/max trim settings (around -5/+10). I also use my wideband input for closed loop operations.
try lowering both of your fast rate of changes into the low 20's across the board.
Add a midpoint to your voltage/lambda table (instead of 0v=x, 5v=y, 5v=y, do 0v=x, 2.5v=z, 5v=y).
Lower your Idle speed vs temperature (stationary) at 177* to around 800rpm. Set the minimum idle speed to that as well
Then, while running, bump the idle duty slider up and down slowly, noting the idle. Log it. Also dial in your injector voltage deadtimes. There are a few other things in your settings i'd personally change that effect other things
IdaSol
Posts: 21
Joined: Tue Jul 29, 2014 10:20 am

Re: Closed Loop Mysteries

Post by IdaSol »

What other things would you change? How does one go about setting the injector dead times. I'm only familiar with the procedure on DSMLink.

I have done all my tuning in open loop and only switch it back to closed loop once finished.

I am also only using my wideband to log AFR. The OEM Honda sensor is providing the closed loop feedback for the trim adjustments. They are at the same exact location in my exhaust stream, just before my flex pipe. I'll check my base engine timing and do the other adjustments you noted.
Boost4Breakfast
Posts: 9
Joined: Sat Feb 14, 2015 4:31 am

Re: Closed Loop Mysteries

Post by Boost4Breakfast »

Dialing in deadtimes is similar to ecmlink (v3), but with more than one point (which v2/2.5 only has q). Set-up your voltage points to where the car will normally run (like11.2, 11.9, 12.5, 13.2,13.8, 14.2, 15.0), instead of what you have (0, 8, 10, 12, 14, 16, 24)....when do you ever run 0,8,16,or 24v while the engine is running?
Get the deadtime ms/volts for your injectors from the manufacturer/supplier for a starting point. Start the car, turn off ALL accessories and disable the alt control in the MISC tab in hondata. start, let warm up, and enable open loop. log and note which cell is being used at idle. you may want to adjust one of the voltage values to the same as the current voltage. Note that. adjust idle area of fuel table to get you to ~14.7 afr. Then Turn on all accessories, lights, music, etc. Note where the battery voltage changes to in the logs and which cell is then highlighted. Increase/decrease ms value (by 0.02-.04ms at a time) until afr's are back to ~14.7.
I then usually plot those 2 points on a graph (10-16 volts being one axis, 0.01-2.00 ms being the other) and draw a straight line thru them. Then use your other voltage points in hondata and plot them on the graph to find the approximate ms of each of them on that line. That should get ya in the ball park.

Other suggestions:
-Boost cut cold @ 4psi with hot/cold temp change at 155*
-Drop closed loop max engine speed to what rpm you are at in 5th at 90-95mph
-Copy the "Open loop- low engine load" table in your water temp compensation to the "Closed loop-low engine load" for a good starting point.
-Cold start, post start fuel adjustment in your "fuel trim" tab would be good to start around 8%
-Set your rev limiter to fuel + ignition

That's all I can think of for suggestions to work off of without seeing the car.
Post Reply