I consider myself a beginning student in the art of ECU tuning, as I would like to become good enough at it to do it myself. Right now, I’m experimenting with stock maps and data logging with EvoScan to see how my car responds. Since I don’t have a wide band, I am not messing with anything else.

I would like some input on a particular acceleration run that I am attaching.
The mods on the car are 70mm ceramic coated downpipe, 76mm test pipe, and 80mm cat back exhaust, WORKS drop-in filter with airbox cover removed.

The ECU is stock JDM IX GT/GSR fuel/timing with WORKS proprietary ECU boost control set-up and the JDM IX RS MIVEC map.

One thing to note is I have a huge boost spike on spool (up to 26 psi) which gets down to 20 psi by 4500 RPM or so and is rock solid to redline. This was the case before as well with the WORKS map.

Specific questions:
1. Is it okay to run this map on 92 octane?
2. How much of a concern is the 26 psi boost spike?
3. Is it surprising that calculated load gets to over 300?
4. How worried should I be about the knock that is present?

I thank everyone for any insight they can offer. I hope to become a resource in the future, but like I said I’m still pretty green.

 

I wouldn't trust that load calc 100% and I also wouldn't worry about the 2 counts of knock. Do multiple pulls and look at the knock after everything is heat soaked. Your timing doesn't look very high.

(Take everything I say with a grain of salt I'm a newb too, but I say it looks good)

 

I checked your injector duty cycle and you are running pretty high towards the end of that pull. You go 90% at 5750 rpm, climb past 95% at 6600 RPM and it keeps going up. I presume if you would have stayed in it to 7500 or more you would have passed 100% IDC.

I recently logged my car on 25 psi with 264 cams, TBE, Walbro FMIC, etc and I was running 104% IDC in the upper RPM. This was with the boost tapering to 22 or 23 psi. I turned the boost down to 23 and ordered up some injectors.

I assume this is a common problem on these cars, but as EvoScan only logs pulse width I'll bet not too many people take the time to do the conversion.


Here is the formula: InjPulseWidth / ((60000 / RPM) * 2) or (IPW * RPM) / 120000


Or if you are super lazy you can plug this into row 2, any column of your excel file: U2/((60000/G2)*2)

 

Here is what I would do if I were you:

1. Put the airbox cover back on. There is really no need for this kind of a mode at this point of your learning curve. Stay with the stock airbox and the Wroks drop-in.

2. What octane was the map designed for by Works? Was it designed for 91 octane or 93 octane?

3. I would be concerned over a 26 psi spike. But do a re-test with the air box cover on and tell us what the spike is now.

4. Post the actual log in csv format. Zip it and post it. Makes for a document that is very easy to read.

5. Works boost control is done by using a less than 60 size pill on the turbo side and then manipulating the wastegate duty cycle table in the ECU to achieve desired and steady boost. I KNOW it works because I have tried it on my old Evo 8. What you need to do is lower the number in the WGDC table in your Rom @ peak boost to get rid of that 26 psi spike. But first try and test with the air box cover back on.

6. Your load calc is hitting 300 because you removed the air box cover. Put it back on and your load will go down and MAYBE your boost spike as well.

 

Thanks for the input everyone.

I reattached the run as a zipped CSV, and included IDC calculation (thanks coolnick).

Just as some additional background, the car was probably relatively heatsoaked as this was after a few other runs (not all-out, but still pretty aggressive). I also realized that I lifted off the throttle subconsciously even on this run.

The map from WORKS is for 92 octane. I'm assuming they just interpolated their 91 and 93 maps. Keep in mind, though, that the only "WORKS" stuff in my tune now is the boost control.

I am going to try and do some more logs tonight, but first I'm going to
1. Put the intake cover back on
2. Play with WGDC to try to kill the boost spike

The 98% IDC has me a bit scared. I'm not sure how it is possible to have over 100% like you said you had, coolnick. Maybe you can enlighten me. I am guessing I am still running pretty rich since its a stock JDM IX fuel map, so leaning it out more once I have a wideband may help that condition.

Thanks for the feedback so far, keep it coming!

 

26 psi with modest ignition and no knock means that you have real good gas or not reading correct boost.

Tuning without WB is like driving your car at night without headlights, I can't believe I did that for 6 months myself

 

No need to worry about the 98% IDC at 6900 rpm. That's pretty typical for Evo IXs. I would be concerned about 26 psi spike even if there is only a few knock counts. What gauge are you using to measure that?

 

Keep in mind that this is a 26 psi spike, though it pretty quickly lowers itself back to 20ish.

Yeah. I wouldn't really consider myself as tuning. I'm just messing with various stock maps and making sure the knock isn't too high. I realize it's a slippery slope...which is why I think a Zeitronix unit is in my future

I am using a Defi BF gauge. I realize this isn't as accurate as a MAP gauge, but the Defi gauge was pretty expensive. I hope that wasn't all from the bling factor.

Can you explain why living so close to 100% IDC is okay? From my small background in electronics, it isn't possible to have a duty cycle over 100%.

 

The easiest answer to defend is that its ok because no one has had a problem thus far with going lean even on a IX with the usual stage 1 mods (TBE, air filter, LICP, tune).

More hand waving is required for a technical explanation. The IDC is obtained from the IPW and RPM. A simple possible explanation is that the IPW is probably less than what the ECU is reporting making the actual IDC lower than reported. Another angle some people take relates the available IDC to the injector rating. The stock injectors are rated at 560 cc/min, but the injector scaling used by the ECU for the stock injectors is 513 cc/min. The actual rating is about 10% higher than the rating entered into the ECU. Thus, some people say that the reported IDC reads high by about 10%.

I forgot to mention that its not uncommon to see calculated IDC values of greater than 100%.

 

You also need to set the Boost Desired Engine Load tables correctly as WERX sets them maxed out at 159 basically taking out the closed-loop error correction out, this is more the reason why you are spiking as the error correction isn't working correctly.

 

Can you give me any information about what would be a good starting point for this? I suppose turning these values down could only serve to decrease the boost I'm seeing.

 

Put in the Stock values in and see from there/

 

Boost Desired Engine Load + Boost Offset = target load.

so, let's say you are hitting 250 load at 4000 rpm, you would set your BDEL at 150 and your Boost Offset at 100 = 250.

Now, your target load is 250, if the actual load goes over or under 250 load, then the Boost Error Correction table comes into play. Let's say the load is 5% over the target load, look at the 5% column in the error correction table, there you will see how much waste gate duty cycle the ecu will take away to bring it closer to the target load.

 

Before you fiddle with the boost tables, put back the air box cover and go logging.

Start when the car is cold and drive around for 10-15 minutes to get it warm. Then go a flat piece of road and do THREE back-to-back WOT 3rd gear logs.

Then zip them and post the results. I am betting that your boost spike and your load calc spike will drop.

If that does not work, then you can start fiddling with the boost tables as I and razorlab have suggested.

 

The numbers that you enter in the Boost Desired Engine Load tables are dependent on the load that your car hits per rpm.

So lets say that your hitting a max load of 240 @ 3000 rpm. Then you can enter 150 in the BDEL table @ 3000 and set the Boost Control Load Offset to 100 or 80. Go with 100 so you can have more room in case you hit higher loads. You have to do that for the entire BDEL table.