Desired AFR

Tuning discussion for all first generation Mazda RX-8 (2004-2008)
Post Reply
tophe88
Posts: 29
Joined: May 6th, 2022, 8:31 am

Desired AFR

Post by tophe88 »

Hello

I would like know why my AFR is too less of my AFR desired ?? when i log ?

Regards
User avatar
wcs
Posts: 61
Joined: May 12th, 2021, 12:46 am

Re: Desired AFR

Post by wcs »

My personal experience is that Desired AFR parameter does not work.
In Open Loop the Desired AFR parameter IMO is completely broken, and in Closed Loop there seems to be enough of a variance to make it suspect.
tophe88
Posts: 29
Joined: May 6th, 2022, 8:31 am

Re: Desired AFR

Post by tophe88 »

yes i agree, in closed loop its perfect !

So have you a idea for respect ? or need to do some step by step for have the good value ???
Attachments
AFR.jpg
AFR.jpg (63.71 KiB) Viewed 3134 times
Walle
Posts: 51
Joined: August 29th, 2021, 7:27 pm

Re: Desired AFR

Post by Walle »

wcs wrote: June 8th, 2022, 12:06 am My personal experience is that Desired AFR parameter does not work.
In Open Loop the Desired AFR parameter IMO is completely broken, and in Closed Loop there seems to be enough of a variance to make it suspect.
when the VE table is cleaned up using ltf, stf and deviation from desired afr, the desired table is followed perfectly afterwards when I changed it. a lot of patience but worth it. made tables in excel to use logs to generate the corrected tables, and was a brief afterwards.

was able to completely disable close loop and drive with less than 0.1 afr deviation. The low load and high rpm part is really hectic in oem (very lean! )
dcox11b
Posts: 25
Joined: March 9th, 2021, 1:28 pm
Location: USA
Contact:

Re: Desired AFR

Post by dcox11b »

Walle wrote: September 18th, 2022, 7:15 am
wcs wrote: June 8th, 2022, 12:06 am My personal experience is that Desired AFR parameter does not work.
In Open Loop the Desired AFR parameter IMO is completely broken, and in Closed Loop there seems to be enough of a variance to make it suspect.
when the VE table is cleaned up using ltf, stf and deviation from desired afr, the desired table is followed perfectly afterwards when I changed it. a lot of patience but worth it. made tables in excel to use logs to generate the corrected tables, and was a brief afterwards.

was able to completely disable close loop and drive with less than 0.1 afr deviation. The low load and high rpm part is really hectic in oem (very lean! )
Hey Walle, would you happen to be able to walk me through or share a bit about your process to get the VE table cleaned up? This would be immensely helpful for my learning.

Cheers
2005 Rx8 Shinka - Black Cherry Mica
Walle
Posts: 51
Joined: August 29th, 2021, 7:27 pm

Re: Desired AFR

Post by Walle »

dcox11b wrote: May 9th, 2023, 12:16 pm ....
Hey Walle, would you happen to be able to walk me through or share a bit about your process to get the VE table cleaned up? This would be immensely helpful for my learning.

Cheers
Hello,
yes I would be able to do that. I am "back" was sick and thus inactive. :welcome
dcox11b
Posts: 25
Joined: March 9th, 2021, 1:28 pm
Location: USA
Contact:

Re: Desired AFR

Post by dcox11b »

Walle wrote: September 18th, 2023, 8:56 pm
dcox11b wrote: May 9th, 2023, 12:16 pm ....
Hey Walle, would you happen to be able to walk me through or share a bit about your process to get the VE table cleaned up? This would be immensely helpful for my learning.

Cheers
Hello,
yes I would be able to do that. I am "back" was sick and thus inactive. :welcome
Welcome back! I hope things are going well now. Please do tell me the process you used for veining up those tables.
2005 Rx8 Shinka - Black Cherry Mica
Walle
Posts: 51
Joined: August 29th, 2021, 7:27 pm

Re: Desired AFR

Post by Walle »

we have to stay strong and :burnrubber

before I find time to fully explain everything take a look at this video Sir:
https://youtu.be/DBFn1RDa3eM?si=3Xha1i8Swh9_c8DI

The biggest "problem" is the ECU doesn't export target AFR as soon as it exits the closed loop, so approach as in video by HP academy is not possible

also MegaLogViewer HD is something that is great for pro users, and ones that can afford it.
I'm going to acquire it to speed up my work as a professional tuner.

seams that you can try it now for free (I don't know how much is "open" and how much it is useful in demo/free version)

So far I have been using a self made excel table which made me take car and be much more understanding of the work that is done.

1st
You need a high speed logger.
You have the Obd Link SX I assume or Versa Link (I hope you are not doing any ECU remapping with remote connection)
if you have the obd Link SX, get a usbA>usbC adapter to use it with your phone and install the OBD Link app. if you have an iphone, get an android :D :drive

2nd
Prior to this you need to make your car run like "shit"
Ie, you need to make the car ready to be Logged.
1st table - clear the Closed loop enrichment table - make it all 0
set a AFR target table at the rich side (oem one is quite goodfor this purpose) but make it "smother"

now here you have two options, do once at all ore part in closed loop part in open loop:
2.1: - disable closed loop altogether, flash , record ONLY #1, "2 and #5 from list in step 3. (which makes it faster) be careful and monitor the AFR as low load and high rpm is off on the ECU map (VE is underestimated, and engine gets way to lean).

2.2: - split the logging in to two areas. closed loop, open loop.. this is how I did it initially but I would argue it is much more complicated and more prone to errors in calculations.

3rd
You need to select several PID's to log (will check them all later and confirm if I used more than I have here listed) and this is irrespective of the table editor / visualizer you are going to use, but obviosly MegaLog will require slightly diferent setup.
  1. Engine speed - RPM
  • Engine Absolute load - %
  • Short term ft - %
  • Long term ft - %
  • Lambda sensor - Lambda
4th
drive, drive as much as you can in various loading conditions / rpm, but try to avoid fast transitions of load to avoid transient enrichment / leaning.

5th
when you obtain data, depending on the approach 2.1 or 2.2 you need to analyze the data.
in either case you need to setup the excel to group the data in to the RPM / Load table cells , in same format as it is in the Versa tune target. That will generate the "achieved AFR" table. this is THE MOST complicated step in the manual data analysis.
if you are doing it with 2.2 approach you have to in addition to this modified the archived AFR by the LTFT and STFT to show what would have been the achieved AFR if those corrections were not applied as you want to applied the correction to the VE so that the ECU doesn't need them (you will have them active, but wont need them)...

now when you have your achieved, you compare it to the target AFR table you copy from Versa, and you appli those corrections to the VE table..

repeat 4th and 5th step! :D

after the VE setup is finished, update the AFR target tables to the appropriate and enable closed loop. I would keep the closed loop enrichment table disabled and Enjoy
:burnrubber :drive

I am working and need to write down an additional step to recognize if it is the MAF calibration that is out of alignment and not the VE table. For that MAF would be to be logged and compare to the corrections, that would require different excursive on the target AFR data vs maf (which we dont have in "native" form)


afterthought,
Enabling Closed loop up to high load / high rpm but have to think what would be the implication with such.

PS, if you find this excessive, I can provide you with an e-tuner service and explanation why and how has been modified
Post Reply