closed loop anormal variation

s300 and SManager software questions & answers
User avatar
Spunkster
Site Admin
Posts: 17861
Joined: Fri Sep 20, 2002 5:06 pm
Location: Hondata
Contact:

Post by Spunkster » Mon Mar 23, 2009 2:13 pm

Look at the attached datalog and you will see how a properly functioning wideband/narrowband setup will look. This was using the recommended PLX wideband.
Attachments
showing closed loop.s3d
Showing how closed loop will look when datalogging...car not fully warmed up
(471.4 KiB) Downloaded 52 times

GTMcoupe
Posts: 110
Joined: Sat Dec 10, 2005 6:39 am
Location: In a Kitcar in Oxford in the UK

Post by GTMcoupe » Mon Mar 23, 2009 2:44 pm

Sounds like you did the same as me when I first tuned my car, and I did post about this and got Hondata to confirm my tuning method, now I know it says put your ecu into open loop to tune, but if you want to run closed loop you need to be in closed loop and tune that to suit.

This is how I tune my car

Closed loop, short term min max set to 0, i.e the ecu thinks it is in Closed loop but as the trim is set to 0 it does not adjust to the lambda sensor

go and tune your map to 14.7

when done reset your max and min trims to what you want. mine are set only to +/-15 % and the trimming done by the ECU is minimal about 2%.

I think the Closed loop tunning needs to be a bit clearer explained. obviously if you are tunning on the dyno for max power you need to be in open loop as this is what your tunning the open loop part of the map.

WEST
Its Small, Its Red, Its Mid engined, It has a B18 fitted, 250hp+per tonne + nitrous

GTMcoupe
Posts: 110
Joined: Sat Dec 10, 2005 6:39 am
Location: In a Kitcar in Oxford in the UK

Post by GTMcoupe » Mon Mar 23, 2009 2:53 pm

as im feeling helpful this evening here is my post when i was tunning closed loop

http://www.hondata.net/forum/viewtopic. ... 3551#24526

WEST
Its Small, Its Red, Its Mid engined, It has a B18 fitted, 250hp+per tonne + nitrous

User avatar
Hondata
Site Admin
Posts: 8723
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Post by Hondata » Mon Mar 23, 2009 3:59 pm

hondapowaa wrote:what i can tell is =>

When in open loop i have exactly 14.7/14.8 afr, with the closedloop activated my afr don't stop to move back and forth example form 13.8 to 15.5.................

What i think is in each cycle, the correction made by the hondata to goal the correct voltage is really too big. the result is one time we are too lean, one other, too rich etc etc etc.

(just an example, if to match we need to injec juste 2ms more but the hondata can correct by step of 5ms we never can match what we want.)

Sincerely i really think the closed loop "module" of the s300 can really be upgraded and receive some améliorations.

Why don't add options that permit to us to set the time correction at each cycle, a minimum target voltage AND a maximum target voltage, and maybe other things?
The ECU does not use absolute values for the fuel change in closed loop - rather is adjusts the fuel based on a percentage of the current fuel value. Thus the injector size and duration do not affect closed loop.

Have you considered providing a datalog and details of the lambda to narrowband analog voltage conversion?
Hondata

VT-Doo
Posts: 119
Joined: Tue May 09, 2006 2:49 pm
Location: Stirling, Scotland

Post by VT-Doo » Tue Mar 31, 2009 4:12 am

If I tune my car in open loop it's fine.
If I tune in closed loop it can also be fine.

BUT, when in closed loop, on cold start up the A/F ratio is about 9:1 and is a pain to start. It's about 50% too rich according to the S300 :shock:
When it warms up the A/F ratio goes back to normal.

I've got the S.Term & L.Term corrections set to 0% with the target voltage at 0.45v (as instructed by the LC-1 manual).

When tuning in closed loop, should the O2 Heater be disabled :?:

I can't understand why closed loop operation would dump so much fuel in on cold start up when the S.Term and L.Term corrections are at 0% :?:

The temperature here when cold is about 10degC. The water temperature compensator is sitting at standard settings; 0=25.0, 20=20.3.

User avatar
Hondata
Site Admin
Posts: 8723
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Post by Hondata » Tue Mar 31, 2009 8:08 am

Post as a new problem with datalog and calibration. If closed loop is enabled then the heater must never be disabled.
Hondata

VT-Doo
Posts: 119
Joined: Tue May 09, 2006 2:49 pm
Location: Stirling, Scotland

Post by VT-Doo » Tue Mar 31, 2009 8:29 am

I posted here as I thought it was relevant to what's being discussed.

GTMcoupe
Posts: 110
Joined: Sat Dec 10, 2005 6:39 am
Location: In a Kitcar in Oxford in the UK

Post by GTMcoupe » Tue Mar 31, 2009 1:29 pm

Hondata wrote:Post as a new problem with datalog and calibration. If closed loop is enabled then the heater must never be disabled.
Whats the reason behind this? As I trust many have plx/lc1 which no longer need the heater, I know it can be simulated with a resistor or just pluging in the sensor, but I would like a technical explonation please is this needed for correct function of the ecu?.

I have my car set up in closed loop with the NB input from the PLX fed into the ECU but have the heater disabled the car is running fine and cycling around lambda during part throttle etc.

WEST
Its Small, Its Red, Its Mid engined, It has a B18 fitted, 250hp+per tonne + nitrous

User avatar
Spunkster
Site Admin
Posts: 17861
Joined: Fri Sep 20, 2002 5:06 pm
Location: Hondata
Contact:

Post by Spunkster » Tue Mar 31, 2009 2:06 pm

The heater can be disabled and still read correctly as long as it is an aftermarket heated wideband with narrowband output. He was confusign this with the K-series ECU which you cannot disable the heater on.

Locked