Page 1 of 2

closed loop cruising problem.. need help please

Posted: Sun Apr 25, 2010 7:13 pm
by mike93hatch
ok my car has been tuned now for 4 years without a single problem until recently.. i dont know if it had anything to do with any new versions of smanager or something else.. i really need some insight i cannot seem to fix this.

here is the problem: while sitting and idling my air fuel stays steady around ~14.7....when cruising in closed loop on hwy the air fuel goes from 14.7 then jumps up to 16 or 17. it does this repeatedly back and fourth every couple seconds. you can feel it going down the road when it goes lean every 2 seconds and then back to 14.7

im using my bosch wideband sensor for the closed loop. dont think it has anything to do with the sensor because air fuel is steady at 14.7 when idling...it only does this when moving. im thinking maybe something to do with a new version of smanager and or soemthing in the tune??

thanks for any help

Posted: Sun Apr 25, 2010 7:18 pm
by mike93hatch
i had soem datalogs but erased them, ill have to get some new ones but i didnt find anything in the logs showing what could be causing this problem...

Posted: Mon Apr 26, 2010 6:58 am
by scttydb411
had same issue w/ car feeling like it was surging while cruising after switching to new version. you'll have to go into the closed loop paremeters and adjust the response tables. go to the help section and look for closed loop. it has good values to use.

Posted: Mon Apr 26, 2010 8:30 am
by mike93hatch
are you talking about the 'closed loop rate of change' parameters?

see all that stuff is new in this versions and ive enver messed with them or understand them.

Posted: Wed Apr 28, 2010 12:11 pm
by XPEHODATA
What wideband you are using? LC-1?
Did you used narrowband output for closed loop before, when you haven't problem? And switch to wideband closed loop mode?

Posted: Wed Apr 28, 2010 3:35 pm
by mike93hatch
XPEHODATA wrote:What wideband you are using? LC-1?
Did you used narrowband output for closed loop before, when you haven't problem? And switch to wideband closed loop mode?
i have a plx m300.. no ive been using narrowband for years now with no problem until now... i noticed the added wideband for closed loop feature and tried that... does the same thing whether im using narrowband or wideband. its something to do with a new features of smanager because it didnt start doing this until i think i upgraded to 1.4.5. i now have 1.4.7 beta and still same thing... wish i could find an older version of smanager so i can prove this.

Posted: Wed Apr 28, 2010 6:01 pm
by ryanelm
I'm having the same problem. the closed loop seems to be going far into the lean and rich while doing its normal cycling and takes forever to correct. It has changed form crossing stoich 2-3 times a second to taking a full second to go from lean to rich.

With this new closed loop upgrade, I wonder if hondata has completely rewritten the closed loop search algorithm, or if they are just tweaking the parameters of the factory one.

In either case it seems less able to deal with the somewhat shaky part-throttle tune I have set up, and of course for those of us running mad injectors a 2% fuel change at low rpm can put you way out.

I've had to set all the cells in the rate of change table to 1% just to keep it somewhat steady.

INFO:
p28-A01 ECU
b18c
e85
1000cc injectors
T3/T04e Garrett 57 trim
81.5mm 9.8:1 CP Pistons, Eagle Rods
300whp @ 8psi
400whp, 331tq @13psi

Posted: Wed Apr 28, 2010 6:37 pm
by XPEHODATA
Maybe new algorithm will good work with stock narrowband sensor. But wideband sensor have a bigger lag time. And new algorithm have very optimistic fast rate of change short trim.
I set slow rate to 2% per second and fast rate to 4% per second. AFR not fast to adapt now but smoother and without spikes.

Posted: Wed Apr 28, 2010 9:23 pm
by mike93hatch
so its working good for you with 1.4.2 ? i knew it had to be the new software and something in its caliberation..

Posted: Thu Apr 29, 2010 10:20 am
by Spunkster
Please look at the settings in my own calibration which work very well.

Posted: Thu Apr 29, 2010 7:02 pm
by ryanelm
Wow those are aggressive values, can I see a datalog of it running?

In any case I'll copy them over and post a data log of what it does on my car.

Posted: Tue May 18, 2010 10:37 pm
by jun1or
ryanelm wrote:Wow those are aggressive values, can I see a datalog of it running?

In any case I'll copy them over and post a data log of what it does on my car.
Sorry to bring this topic back from the dead (19 days).

I have seen this slow transition in closed loop as well, with the new version. I have a pretty good part throttle tune, so no big impact here, but curious why the change from oem. (don't get me wrong, control over closed loop is a nice addition).

Mike

Posted: Wed May 19, 2010 9:18 am
by Hondata
We did not change the settings, only exposed them so that they can be altered.

Posted: Wed May 19, 2010 2:26 pm
by jun1or
Hondata wrote:We did not change the settings, only exposed them so that they can be altered.
As is the case with just about any other new feature implementation in the past from what I've seen; however, my wideband is also showing slower sweeps. Just wanted to tell the OP, and you hondata, that I have noticed a slight change (without modifying these parameters myself).

Regardless of the aforementioned, thanks for sharing your adjusted values.

Mike

Posted: Sun May 23, 2010 9:26 pm
by TTR
I found the "rate of change tables" to be extremely helpful. I have 5 for a value in the "slow" boxes, and 99 in the "fast" ones. It makes a huge difference on the freeway above 4000 rpm (gas mileage went up) and at light throttle low rpm around town over the standard values.