I'm working off a dyno tune that turned out to be pretty far off (too rich, negative timing).

I've done probably 2 dozen pulls and have adjusted fuel and timing. This is my result after a week.

Any input is appreciated, as this is my first self-tune. Feel free to glance at a snippet of the log here: http://images.puremhz.com/Tune/log.xls

Thanks in advance.

 

Looks very good, a little hard to tell what your afr are but they couldn't be far off.
your timing indicates something less than 20 psi(?). oh , I see you have a ix.

have you done any part throttle logs, 10 psi or so, up a hill or with a little brake. it can be a good thing to know as it represents a lot of road course time or even some"normal" driving.
in my case I lost some timing in the 90 100 120 areas.

 

Looks very good. You do have up to 3 KS at 4500 at 230 load or so. AFR seems OK but hard to see..Im thinking 11.2 11.5 in the peak torque 3500 and then it holds nicely on up tapering to about 10.5 . You could try getting it to 10.8 11.0 up top

If you take out 1 degree at 4000 5000 in the 230 load, you'd be knock free .

Is this on 93 ?

Also the little dip in 2byte load occurs just before the knock, meaning you went to a lower load, and probably more timing. see if you can add a bit to the WGDC to get into the higher load column, alternatively. If you have a way to log the boost that will really help i think. probably see the dip in boost at that point just prior or at the beginning of the knock.

Hmm just wondering if the mivec timing table is changing at the boost dip too. Can you post that ?

Milburn

 

I noticed in the xls that your timing is jumping around badly.. does it feel smooth? I also noticed that you've got smoothing turned up to 12 in DLL which is probably masking a problem.

can you show me what it looks like with the smoothing at 5 and SAE turned off?

 

It looks pretty good so far.

You might be able to smooth that timing curve a little by setting up your low octane fuel maps. I found that there is still some interpolation between the high and low octane ignition maps, even with 0 knock count. I copied the stock high octane timing map #1 to my low octane tables. Map #1 is the tamer of the 3. When you do this, expect to run a little richer to avoid knock.

Are you hitting the boost cut when you go over 270 load %? I assume you've got a MBC, correct? I'm just not sure why you're getting the valley in the load curve after peak boost. If you haven't already, try increasing the values in the boost limit table. I know you've gotten rid of the stock boost control system, but I don't know what actions the ECU can still take to prevent overboost. Maybe someone more knowledgeable can chime in.

I expect you're peaking at about 26 psi, holding 21.5 psi or so, then tapering to around 18 psi toward redline. That all seems right, minus the spike. No way to tune that out with a MBC though.

 

After reading the log, Mellon is right the timing is bouncing around badly. AF looks great. You also have a data point showing 319 load. And the dip past peak load is pronounced. Something needs work there. This was on a dyno ? Problem with being strapped down perhaps?

Still very nice work. I'd recommend logging boost to help see whats going on.

Milburn

 

A few comments:

1. Why are you block tuning on your fuel map and ignition map? Since you are logging the 2 byte load, then you can accurately hit the cells that you want to hit. Follow the model set by the stock Rom of decrementing the fuel and timing as the load goes higher. Personally, I do not like block tuning, but some people do it.

2. As Mellon mentioned your timing is jumping around even w/o knock. It could be that you did not make ALL three high octane ignition maps the same. And the ECU is interpolating between the maps. Make ALL three high octane ignition maps the same and see if that solves the problem. I do not have this problem on my Evo 9.

3. Your IDCs are almost maxed out. Did you extend lean spool to fuel cutoff to reduce that? Evo 9s tend to get really rich @ beyond 7000 rpm due to lean spool disable. That might be affecting your IDCs.

4. Why is your wideband skipping a row when it logs? Are you using Evoscan or Logworks to log? If it is Evoscan, then the LC1 has a tendency to do that with Evoscan. There is a reason for it, but it is too technical and I do not understand it. But logging the LC1 with Evoscan tends to cause this.

5. Your hp curve on the DLL looks too peaky. Try smoothing the rpm before creating the log. You can do this if you are using Logworks, but I do not know if you are.

 

I wouldnt really call it block timing but it definitely can be smoothed and increased in some areas. Fuel map looks about right for a car with LS off and an intake that is causing a rich spot (nevermind I see it has a drop in)? You'd be amazed how "block" tuned you can end up with fuel and it will be perfect on a car without a pump.

If it is still a stock airbox I would hack part of the lid and open up the airflow. I am not sure if its just boost control related or MIVEC (I have seen it already and dont think thats really it) but your torque peak looks late for what the MIVEC was last I saw. A JDM map should be making torque and full boost earlier generally.

What are your barometer readings if you are logging them? It would be nice to see your calc load in the log or 2 byte since you log that to kind of see whats going on as far as actual curve. It can do alot to tell whats happening during a pull. Its cool to see that you are getting 3 samples per 1/10 of a second at times and at least 2 sample pretty regular. What speed are you logging at?

All in all it looks like you are getting it done, keep it up.

 

Here are some of the parameters plotted and HPTQ vs rpm. That's a very unusual timing curve. First time I've seen one jump around so much like that. Besides what everyone else said, things look pretty reasonable out past 5500 rpm, but it seems like some optimzation of boost needs to be done at lower rpms to get a more consistent load curve. With your mods, it should be pretty easy to peak at 270 load and smoothly taper down to 230 by 7000 rpm. It should even be possible to build in a 270 load plateau that can be held from 3500 out to ~4250 rpm.

 

I have done a few and plan on doing more. Thanks for the pointer.

AFR is 11.5 @ peak torque with a slow taper to 11.0 at redline. This is 93 with Torco accellerator. I believe I have the JDM MIVEC map.

This may be due to only adjusting one timing map. I was under the impression #1 was cold engine, #2 was warm engine, #3 was limp home. I was only working off #2, but I may clone over to 1 and 3, and also adjust the low octane maps as well, as they are all 100% stock except #2 high octane.

I don't believe I am. I am using a Hallman Pro MBC. I'll try increasing boost limit to be safe. I'm peaking 24psi according to my gauge, and holding about 23, with about 20 or 19 at redline.

I wasn't try to "block tune", I'm just trying not to adjust timing too much over a group of cells. I will fix all ignition maps to prevent the jumpiness. Lean spool is off and I am using EvoScan to log. Not sure if there's a way to fix the gaps.

I'm not logging Baro, and I am actually still using my K&N Cone. I have my stock airbox with K&N Drop In that I may attempt to tune for, I'm undecided as to what the best way to go is. I'm logging with basically the default speed settings for EvoScan and Tactrix cable drivers, not sure what speed that is.

 

I made the same assumption for my WGDC maps. Mrfred straightened me out though. It's more reliable to tune all 3 the same. There is definitely interpolation between the maps and thats where that variation is coming from. Duplicating the high octane maps will get rid of most of the variation, and setting the low octane maps should make it smooth as glass. Good luck and I look forward to your results.

 

This is probably a dumb question but I assume that only the IX's have 3 ignition maps due to MIVEC timing?

 

Well, they might be related, but there is a separate map for MIVEC labeled "VVT."

 

OK, so after fixing my timing maps, I've got a nice smooth timing ramp to go along with my smooth AFR ramp.

DataLogLab is showing ~320whp, so I'm really making progress.

My only problem now is my 2 Byte Load... I can't seem to figure out why it's so humped (or why I get a spiked reading of 319)

Again, I'm using a Hallman MBC, which spikes to 24psi and tapers to about 21 at redline

 

My 2-byte load has spikes like that too. Sometimes 3 or 4 per run.

The Hallman you have has the porcelain ball, correct? It seems to me that the BC is releasing pressure rather quickly at peak boost causing that dip afterward. I thought the Hallman Pro was supposed to help with that issue ...