Bigger is not always better...

 

are you referring to map size?

the kicker to this whole situation is that the car was so well maintained, and was "tuned" by a very smart and i would have thought capable member of this site.

guess i was wrong

this is the bastard of a mivec map i discovered tonight compared to a JDM RS map.

 

Correct, referring to map size... if you are on a stock frame and are only hitting around 320 load I see no point in switching to big maps.. tephra V5, yes if you need altmaps etc...

 

yeah dont worry ive got plenty of work o do before i can worry about any tephra features.

 

Unless you want direct boost per gear, then yeah V5.10

 

Might I also recommend since you are tuning everything, disable closed loop, and tune your maf correction factors with AFRERROR = AFRACTUAL- AFRMAP. If you can get AFRERROR near 0 then your maf is dialed in. I replaced my air filter and intake tube and was off quite a bit. You also have to log airflow (hz 2-byte) to know which column your changing with the AFRERROR. And for the logging, you have to maintain constant throttle, constant airflow if possible to get a good set of data. Any transient throttle/load/airflow readings will cause AccelFueling or DecelFueling to skew your AFRERROR by adding fuel that is not accounted for in the airflow. It was quite a lot of work, but now my AFRMAP is usually +/- 0.2 of my actual AFR, which makes tuning the fuel map a breeze.

No more "well if I type in 9.4 then it will be 10.7 in my car in the wideband".

Oh, and another thing that blew my mind is that 200Hz airflow at 2500 rpm is not the same as 200Hz airflow at 5000 rpm due to the VE of the engine changing with engine speed. I eventually had to decide where would I be using 200Hz airflow the most at? About 3000rpm on the interstate at 60mph. So thats what I tuned for.

We have the Kármán vortex sensor as a MAF, which doesn't measure mass flow rate at all

"An intake air flow rate detector of an internal engine using a Karman vortex sensor, wherein the mass flow rate is obtained by detecting a volumetric flow rate by the Karman vortex sensor"

 

Honestly I never looked into this until you mentioned it. Tuning with LS disabled is easier but it seems like there are really some benefits to LS, and they address things I was trying to fix with my tune. Thanks.

The big maps are useful for their resolution. I rescaled mine to give more resolution in the spool up range and up until where timing gets relatively constant.

 

like i said, i will get up on tephra eventually because there are in fact some ancillary functions i would like to have. i just feel that its a bad idea to convert to larger/high resolution maps, until i get it pretty close to where i want it to be on normal maps.

i made a HUGE improvement in the car today with just a small mivec change. have a new fuel map to try tonight.

 

Killerpenguin remember that you have to log two different load variables if you are still on a stock ROM. You need to log address 840B (1-byte load) for fuel tuning, and address 8960 & 8961 (easiest to log with 2-byte load) for timing, this is if you are located at any elevation above sea level ***BOTH THESE VARIABLE ARE FOR ROM ID 94170015, as you have a Evo IX I would locate the ROM addresses for your ROM ID.*** This would account for your messed up fuel map if the previous owner only logged one load variable. This would explain why your changes to the AFR Map did not do anything, you might be looking at the incorrect load address in Evoscan. Also it is much easier to tune your initial map with LS dissabled, then re-enable it after you get your maps close to where you want them.

LS helps in the following situation: You are on a two lane road and need to pass in a hurry and do not wish to downshift, mash the gas at 2500 RPM and suddenly your map will lean out to allow the turbo to spoil faster due to the hotter air entering the exhaust manifold. Also remember that after you enable it you have to address the trailing time table to determine when you return to the AFR Map. Look up Merlin.oz Ecuflash guide in the googler and it will help to explain what the different table do...

Hope that this helps.

 

^ Merlin's guide ftw. I've mirrored it on my google account: https://docs.google.com/viewer?a=v&p...thkey=CI2o36YO

 

your over thinking it in my opinion. i log 2 byte load. im only at about 300 feet above sea level, and have had no issues knowing exactly what cells im hitting. i also dont tune just the cells i hit, i smooth everything and keep the cells on either side of what i actually hit within a very tight variance. furthermore, i thought 1 byte was only in tephra?

the messed up fuel map is because the previous "tuner" did not use a wideband and added fuel for "safety" the map was requesting low 8's in areas/bottomed out as far as it could go. the reason the first 2 or so tries didnt make any change was because im making changes in baby steps, ie only a max of 0.3 (about two clicks of the increase button) per new map so that i can see map by map exactly what im doing and learn the degree to which my clicks make changes.

 

One byte load is a patch that can be added to any ROM as long as you edit your .xml file, and I'm glad to hear that you are finding your cells, I was just adding that point to make sure that you realize fuel and timing do not use the same load variable on the stock ROMs, nor do they on Tephra ROMs prior to version 7.

 

definite progress tonight.


Need to clean up that dip just before 4k. and i think im trailing off too far at 7k. im spooling 2-300 rpm's faster now, and as a result i am hitting one load column further over than i was previously.

some things i ran into:
-3 3rd gear logs, first 2 were absolutely knock free. 3rd run knocked in 3 seperate clusters, first 2 clusters were a 1 into a 2 count and the third cluster was a 3 count tapering off. im thinking the car might have heat soaked since its so damn humid and i did the 3 pulls in about a 7 minute window. thoughts?

-in 4th gear, i have a lot of knock starting at 6500 in the form of a 6 count dropping to a 5 count. pull was done within 3 minutes of the last 3rd gear pull that knocked. due to hitting higher loads, the pull held 8* for ~1200 rpm, logged timing went up 2 degrees and then i got the 6 count. could holding the same timing too long/not ramping up fast enough be the cause of this?

4th gear

 

Ignore the 1-2 knock counts unless they seem to pop back up in the same spots regularly. The 3 counts of knock in 3rd, does it occur in the same area that you got high knock counts in 4th? Sounds like you need to trim back timing a hair up top. You can't just tune for optimal conditions, heat soak is a normal occurrence and it's best to have the tune be consistent under all conditions.

I only use 2-byte load. I have had inconsistencies with using the 1-byte factor which was way off from actual load. And yes, the scaling factor was the same in the rom and in the evoscan formula.

Holding the same timing shouldn't be the cause of knock unless the timing was TOO low. If that was the case, you'd get the knock on almost every run. Yours seems like heat soaking and a slightly aggressive timing curve.

-Jamie

 

the knock i got in 4th gear is an area that the car has never really knocked before. i do also suspect the timing map since when the car was horribly rich, i had to smooth the timing map because it was so choppy it was almost knocking for an entire pull. i tried to advance anything but i guess i may have.

below is the timing map im using as well as a trace of the 4th gear pull. part of it im pretty sure is also because of jumping over a load column and getting interpolation off of the 270 column now too.