NefMoto

hey... can anybody explain me how to calibrate PI controller for lambda closed loop. I m stuck with functions like kfrp, kfrap, framp, tramp and I am clueless.. can somebody explain all the functions and variables of Lambda controller PID...

I am currently trying to understand and tune the lambda PID, I have adjusted Y2MAX some(I gain max limit) with limited success and I feel I need to get a better grasp of which maps to focus on.

My "problem" is that the o2 correction sways too much, 5% back and forth, and that 5% seems to be fixed and doesnt change from what I can see from my logs?

This might be ideal with stock injectors but with 1000cc injectors it causes actual to sway over and under setpoint, going from 0.96-1.03 back and forth:
(http://img210.imageshack.us/img210/2630/bische120120928160944.png)

(http://img42.imageshack.us/img42/2630/bische120120928160944.png)

I have noticed in the FR that Y2MIN is mentioned in the basic application values, but not in the diagrams.. Then I noticed in my binary there is a 16bit value after Y2MAX that very likely could be Y2MIN, but I have not found this value defined in any other damos I have, it is just left out.

Anyone have any hints or tips on this?

I have been reading more(finding myself reading the german FR more and more) and believe I have found out which map to tune regarding this, LRSZAML - Zwangsamplitude bei ML

OR Forced amplitude based on mass flow for LRS, lol :)

Im having a look at this when get home, FR on the phone FTW

As I understand my stock LRSZAML, it forces o2 correction ~2.6% in each way over and under setpoint at low airflows(about 80kg/h) and up to ~3.5% each way on medium airflows(120-360 something kg/h) and goes down as airflow increases further.

I have revised this now and going for a log to verify.

I think those figures or overly large.

I edited my post as I brainfarted and wrote 7% when I meant 3.5%.

Application notes suggests 3%.

What did you find?

On the phone so I cant post no graphs, but the logs verifyed my theory. I got the sway down from 5% total to 2.5% total :)

I will post a graph and a screenshot of what I did later or tomorrow, having a romantic evening with the wife lol

(http://img818.imageshack.us/img818/3461/lrszaml.jpg)

Top window is stock LRSZAML and bottom is modified, ~halfed the swings corresponding to my changes.

(http://img338.imageshack.us/img338/3234/bische120120929194822.png)

Can someone explain why there is a forced amplitude?

Is it something emissions related or is it required for the PID to work?

EDIT: I may add, as I understand the FR it is for testing the cat converters function. Im going to try turn of the forced amplitude completely via CLRS.

Yeah, thats right, forced fr amplitude is to get sine curve for actual lambda and hence operate catalyst at its optimum.

Thank you, I tried to disable it in CLRS bit2, but no avail. According to the FR setting CLRS bit2 should turn off b_za flag which enables the forced amplitude.

I could zero the maps to disable the swings, but I would rather turn the whole function off. I need to go over that section again and manually translate, maybe I understood the CLRS wrong.

I was wrong about the CLRS bit2, setting bit2 will force b_za even if it isnt inited from LRSKA.

The bit I need to change is CLRSKA bit0, and I think I have done so with no success. I must have the wrong address to CLRSKA, the one I have is 1A2BA but it seems off as its stock value is FF?
(http://img211.imageshack.us/img211/541/clrska.jpg)

My stock binary: http://nefariousmotorsports.com/forum/index.php?topic=1103.0 (http://nefariousmotorsports.com/forum/index.php?topic=1103.0)