Page 2 of 3

Posted: Thu Aug 14, 2008 1:42 pm
by Razathorn
magnEsium wrote:Why am I seeing ass backwards results here?
I dunno, but it's gonna suck to have to leave him 5 degrees retarded in the map when you're done just in case k-pro decides to add those 5 degrees back at some point down the road. Perhaps there is a setting that defines high/low load switch and that setting is wrong?

Posted: Thu Aug 14, 2008 11:14 pm
by Razathorn
I have an appointment to tune a k100 guy at the shop on saturday -- I'll have to pay special attention to if the ecu does this since I'll be using my k-pro for the tuning and we tend to see some decent intake temps on the dyno.

Posted: Fri Aug 15, 2008 3:16 am
by magnEsium
Yeah its really gay

Just keep the IATs consistent between 80-110

No retarding between those points

Posted: Fri Aug 15, 2008 10:01 am
by Razathorn
magnEsium wrote:Yeah its really gay

Just keep the IATs consistent between 80-110

No retarding between those points
One of the datalogs has it at 105F! My car sees no retard at 140+!

I wonder if this is brought about by the car knocking for a long period of time and the retard tables in the ecu. I remember reading some analysis from hondata on the K series ecu and how it based everything off retard tables or something along that line. I wonder if these are being used since removing power from the ecu for a while gets rid of the retard according to some folks.

Posted: Sat Aug 16, 2008 5:21 pm
by Razathorn
I think I have some definitive data. I tuned two cars on the dyno today, intake temps of 106+, both with the same k-pro board, both with different kproified ecus (one of them was from a k100 install). One ecu was an a06, the other was an a05. Both had powered removed for minutes before being used (k100 swapped etc).

The a05 ecu (mine) never removed timing. The a06 removed timing compared to the map.

This was same day within hours. This was with brand new #8 plugs and a reset ecu. Removing power from the ecu for 10 minutes didn't even reset the timing.

I left the timing in the customer car and told him to be on the lookout for CEL flickers since I know the timing maps used were good for daily driving since it was based on a very safe map I had used before, but the bottom line is with the a06 ecu, it never ran all the timing I asked it to, and with the a05, same kpro board, it did.

Wayne

Posted: Mon Aug 18, 2008 9:44 am
by Spunkster
ECU type has nothing to do with this.

Posted: Mon Aug 18, 2008 10:37 am
by Razathorn
Spunkster wrote:ECU type has nothing to do with this.
Well, then I'm at a loss. Two ecus, same k-pro board, same day, same dyno, same intake temps, same coolant temps. One was my s/c rsx, the other was a customer i/rh/e bolt-on rsx. We both have CAI and my IAT is before the blower, so we're seeing the same conditions.

I don't know what other information the community can provide. I'm ready and willing to do any reconnaissance I can -- I have to clean up a portion of this guys map again since the humidity change, so I'll have access to his car again and his ecu that likes to retard timing. I can swap ecus and run the same calibration and see if one retards and the other does not.

I can do anything -- just let me know.

Wayne

Posted: Tue Aug 19, 2008 7:49 am
by Hondata
If it is something which we have altered, then I'll fix it, but for many issues the first course of action is to ask Honda/Acura to look at the problem. ie this is something that all ECUs will do, but you would not know without datalogging.

First I think you should narrow your conclusion by trying the same calibration on two ECUs in the same vehicle - no two cars are actually going to be the same.

Second you can try editing the calibration file and change 'ProgrammedKnockRetard=1' to 'ProgrammedKnockRetard=0'. The way Honda implements knock control uses a long term trim, which may be the problem.

Posted: Tue Aug 19, 2008 8:42 am
by Razathorn
Hondata wrote:First I think you should narrow your conclusion by trying the same calibration on two ECUs in the same vehicle - no two cars are actually going to be the same.

Second you can try editing the calibration file and change 'ProgrammedKnockRetard=1' to 'ProgrammedKnockRetard=0'. The way Honda implements knock control uses a long term trim, which may be the problem.
I will try the above mentioned suggestions.

Assuming the knock retard long term trim is to blame, how can we reset it? I imagine clearing security / erasing the ecu would do the trick. Visibility to the knock retard trim in the sensors window would be a treat as well, if at all possible.

I will have access to this car and ecu again in the coming days.

Posted: Wed Aug 20, 2008 8:52 am
by roland
Razathorn wrote:I wonder if this is brought about by the car knocking for a long period of time and the retard tables in the ecu. I remember reading some analysis from hondata on the K series ecu and how it based everything off retard tables or something along that line. I wonder if these are being used since removing power from the ecu for a while gets rid of the retard according to some folks.
About the power reset - I am not 100% sure because I didn't do it myself, but I do believe my ECU had the power removed by the dealer when they did the 3rd gear TSB. I made the datalogs about 10 days later so I guess that is plenty of time for the retard table to be rebuilt but I had tuned all of the knock out as far as I know.

When I get some free time I will double check this theory and I will also be looking into the ProgrammedKnockRetard value that was mentioned.

Posted: Thu Aug 21, 2008 11:43 pm
by Razathorn
I did some cleanup on the car that was pulling timing. Seems my ecu does it too -- it was running about 1 degree less than was called for in columns 8 and 9.

I would really like some information on the knock trims etc. I remember reading something way back that seemed to indicate that there were knock retard tables... like there are modifiers all over the map to pull timing in specific places of the map, almost like long term tuning of ignition for detonation prevention.

Doug, can you shed some light on the operation of the ecu in this area -- is this something we can zero out for new tunes? I've never seen it happen above column 10.

Wayne

Posted: Fri Aug 22, 2008 8:48 am
by Hondata
The knock retard is exceeding complex - there are not just a few simple tables controlling how it works. I really don't think that editing the behaviour is something that will ever be possible, just from the sheer size of the code and number of tables. To give you some idea, it would be several times larger than all the existing tables and parameters currently persent in KManager. It would also require the corresponding development time to expose the workings.

Remember that this is not something we have added or altered - this is how it works from Honda.

However, there are some relatively simple ways to control the knock retard. Try the ProgrammedKnockRetard setting and let me know if that makes any difference.

Posted: Sat Sep 06, 2008 12:35 pm
by ebatangan
Any update on this???

Posted: Tue Sep 09, 2008 7:22 am
by ebatangan
I have tried changing that paramater to 0 instead of 1 and I still get ign retard. Other then that, any other thoughts? Thanks

Posted: Sat Sep 20, 2008 12:34 pm
by snowman
i get around 5 degrees of retard with the iat around 60 F. even though ive set the kal file ,iat retard high all to 0.
what else do we need to change, i still cant get the desired ignition timing?