Thanks again, the approximations may look like we're missing a lot of peaks, but the outliers are usually on transients and with just a few items recorded. I tend to place most weight on the cells that are accessed frequently.

So how about for Tephra's MAP VE calibration:
10kPa: 70% of 10=7
20kPa: 70% of 20=14
30kPa: 77% of 30=23
50kPa: 77% of 50=39
70kPa: 77% of 70=54
100kPa: 90% of 100=90
130kPa: 100% of 130=130
310kPa: 100% of 310=310

RPM VE:
About 94% at 500,1000 (this should help your lean idle depending on what you have there now)
100% from 1500 to say 5500 then taper (no log info, suggest follow mine).

It is quite surprising how flat the VE is from 30-70kPa, mine is the same.

 

John what mods do you have versus tephra's that yields greater VE?

Also, from those logs eric posted how did you determine RPM VE.

Be gentle i am very green with SD.

Thanks!

 

Just to throw the TPS data out there. This is TPS vs RPM vs the same VE (100*load/map):



I know you don't have a VE table for TPS, but I figured looking at the data for both TPS and map would be good for comparison.

 

Nothing much, more likely just differences in our induction systems that affect the (old) MAF sensor calibrations we're trying to emulate.

Also my load is not temp compensated but I expect Tephra's is. I used non-compensated load because MAP is also not compensated at this stage, so I was comparing like with like.

RPM VE - there is drop from about 77 at 1500 RPM to 72% at 1000 RPM at the same MAP.

 

TPS correlation is good Eric, if it wasn't for the ISCV (being a second throttle) it would be worth trying instead of MAP.

 

Forgive me, i still don't see where you are getting the 72% at 1000rpm. I am probably looking at the wrong table.

 

Post #180, 40kPa, 1000 RPM.

 

oh so you just take the lowest value. I was thinking you averaged out the values at 1000rpm. I will wait until the system is finalized before making most inquiries.

 

1000 RPM at 40kPa (idle) needs less VE than 2000 RPM at 40kPa (cruise). Since they are both at 40kPa we adjust the RPM VE to suit.

I'm using Tephra's suggestion of 16 bits for the MAP VE and calibration table, coming along nicely, although adds a few more tables to Ecuflash for the conversion...

 

Anybody ever play around with how MIVEC timing affects the engine VE?

It seems like it would be a pretty significant detail, but maybe it's not?

It's awesome how you have been able to fit SD so nicely into the factory code with just a few changes. Very creative.

 

Hey John,

Why did you choose those as your bins?

I understand the 130kpa last bin thing, just not the 30/70/80?

 

The post you quoted was just my quick note of where I felt the "breakpoints" were. Post #181 has my suggestion.

 

I've used the 10 bit MAP ADC and the load lookup to word. I've got rid of the 0x73 constant and made the code independent of JDM MAP sensor scaling now. The bottom row of the MAP VE table is now scaled like load with an adjustment to counteract the MAF size multiplication (0.596). So the load lookup is now crazy detailed and has silly limits. The MAP sensor can be calibrated by simply changing the scaling for the top row of the MAP VE table and then correcting the bins to the typical values.

I have also killed the other baro check and will see if that brings the LTFT back tomorrow.

 

MIVEC I think has large effects on VE, so if you change your MIVEC table you may need to adjust.

 

looks good!