7000 RPM IPW step on US Evo IX

[trinydex]

JM8859-13,14,15

At &H1285 you have &HE0 (=7000 RPM).

On my JDM if I move this to &HF0 (=7500 RPM) I move the IPW step accordingly getting rid of a rich step where I'm making useful power and moving it to 106 RPM before the stock rev limiter, which I don't mind.

All the Evoscan datalogs that cross 7000 RPM on US Evo IX that I've seen (and also the VIIIs as well but I haven't disassembled those in this area to find the routines) show this step.

If you want to change it, have a play at your own risk and please post your before-after IPW vs RPM plots. Obviously make sure your AFR is OK - if you've already compensated your fuel map for this effect you'll probably want to flatten it. As previously mentioned, getting rid of the cause is better, because if you lean your 7000 RPM zone, you'll lean the 6500-6999 area by interpolation, which it doesn't need.

If this works as JDM it will improve the smoothness of your top end AFR curve and should maintain power from 7000-7500 a little better.

All the lean spool tables and code are there in the US ROMs as far as I can tell from disassembly. The values for the actual adjustments are similar too.


So, doing some quick digging in my '03 ECU (94170008), it seems like &H1685 might be the corresponding value.Â* The nearby values look like this:

JM8859:
Row 1268: 0040 003D 0068 00E6 001A 000A 0006 0028 0028 0040 0010 005A 0050 0091 00E0 0080 00FF 00FF 00FF

94170008:
Row 166A: 003D 0068 00E6 001A 000A 0006 0067 0028 0040 0010 0075 0054 0091 00E0 0080 00FF 00FF 00FF 0017

I basically searched for "009100E00080" and that was the only hit I got in the '03 ROM...

Thoughts?

l8r)


I think that's the one to try. Not only because the data is so convincing, but because the subroutine that ]

[quote=Ludikraut;3426558]I guess I'll put some more race gas in it ... just in case. ;)

Any idea what the values around it are associated with?

Also, if:
&HE0 = 7000 RPM
&HF0 = 7500 RPM

then what is the equivalent of 8000 RPM?

l8r)

One more question:

Is this a uint8 or uint16 value?

l8r)


It is the generic 8 bit RPM scaling that is in lots of ECU maps. RPM=value*31.25
Max is therefore 255*31.25=7968.75 RPM (&HFF or decimal 255). Although the stored values in the map are words, I don't trust the code to do it properly, so keep any existing byte values to bytes to be safe.


This is the same exact location (1285) for this value on my 05 ROM, too, internal ID 96940011.


Eric


Don't have 9694, but I loaded 9693 and it seems to be fine at 1285, again ]

[quote=jcsbanks;3427853]JohnBradley, yours also is at 1285 on the ROM you previously sent to me.

The byte at 1285 is 224 (decimal). If you change to 240 (decimal) this changes the effect from 7000 to 7500 RPM.

In ECUflash:

Right click parameters and select new map. OK to inherit Evo9base.
Change type to 1D.
Put in a name like "IPW step". Put 1285 in the address. UINT8 in the scaling. Click Apply then OK.

Then file menu... save models.

Exit ECUflash and reopen and reload your ROM.
Then your new map should appear. Seems to go under misc even though I said to go under fuel. Then you can confirm the value is 224 and change it to 240.

Then reflash and test on Evoscan as you go through 7000 and 7500 RPM on full throttle. The IPW step upwards should have moved.


Once I confirm that this works on my '03 ECU (94170008), I will post updated XML files.Â*

I can also put together another "how-to" on adding map definitions to ECUflash...Â* that way all the non-9417008 ECU owners won't be left out in the cold. :p

l8r)


Because I dont believe in pussyfooting, I reflashed with this value set at 254 and then went for a drive.Â* I have my boost set rather high right now anyway (WG @ 112, Boost load at 145 curved to 159.4 by peak, 25-22psi) for pumpgas but I noticed an immediate improvement.Â* I didnt log any knock increase in the 7000+ range, but I also hadnt ever tried to tune this out before.Â* I have target AFR set at 9.4/9.5 and it seemed to love it.

Ludikraut-How to's would be really cool.Â* As soon as JC explained it, it was a duh situation and I immediately added the definition like I had been doing it forever.

LEAN SPOOL


Has anyone tried to change lean spoolÂ* starting point also? address 1281 value 80(2500rpm)? does it work on USDM evos?

Quote:
Originally Posted by ST
is this applicable to all USDM model maps? Still can't see new map def after much tinkering.

yes it works on USDM model.
it usually gets created under MICS for no reason. try to edit and change it to fuel again.


yes it works on USDM model.
it usually gets created under MICS for no reason. try to edit and change it to fuel again.


This worked for me by right-clicking IPW Step and choosing "Edit Map", setting the category to anything else, hitting apply, changing it back to Fuel, hitting apply, saving and then reloading the ROM would finally put IPW Step in the right category.


I am wondering instead of using Unit8, couldnt for ease of use use rpm8?


I think it is "uint8" instead of "unit8" and it stands for 8-bit (1 byte) unsigned integer.


Nevermind, thru experimentation I see why I was wrong.Â* However on my MR it shows that the enable (address 1281) is set at 192?Â* Thats 6000rpm if I did my math correctly.


Looking at the scaling definition for RPM8, the conversion formula listed wouldn't match and the max display value would be off as well since 255 is the IPW maximum and 2000 is the RPM8 max. I guess one could define a new scaling type instead of the generic uint8. I'm not sure if I'm interpreting all this stuff correctly as I am a novice to all this.


The JDM has the enable at 3500 RPM. The US or UK models are set to 2500 or 3000 IIRC. There is no need to change this at all.

Changing the disable is the topic of the thread, simply to get rid of this rich IPW step at 7000 RPM and gain a tiny bit of top end perhaps by not choking it with fuel.

You can use any scaling you like, just verify that it does the right thing with the bytes. I wouldn't worry about it - change and forget it. This is a feature we don't need disabling at 7000 RPM and richening us up.


i know that...i followed jcsbanks instructions (see about 10 posts above)..

- parameters: new map
- name: ipw step
- 1d
- 1285 addy
- scaling
- uint8 scling
- apply
- save model
- close and reopen

nothing under fuel or misc =/




Looks like I missed all the fun. ... that's what I get for driving my car around, beating it into the ground, logging the IPW step.

Well it definitely works. I'll post some logs later when I get more time. For now, here's one way to add the new value to your existing maps:

1.) Backup your evo7base.xml and <<ECUID>>.xml file. The <<ECUID>> should correspond to the Internal ID of your ECU. In my case it is 97140008, which would make the filename: 97140008.xml

WARNING!!!
You must know the correct address to enter for your ECU ID!!!
The address of 1685 that is used in the example below is the correct address for ECU ID 97140008. This address may not be the correct address in an ECU with a different internal ID. Make sure you find out what the correct address value for YOUR car's ECU ID is.

2.) Open up the evo7base.xml file using NOTEPAD (right-click on the file, open with ... pick notepad). The file is usually found in C:\Program Files\OpenECU\EcuFlash\rommetadata\mitsubishi\evo

a.) add the following line after the last table, but before the </rom>:
<table name="High RPM IPW Step" category="Fuel" type="1D" level="2" scaling="uint8"/>

b.)
... this should result in the end of the file looking something like:

Quote:
</table>

<table name="High RPM IPW Step" category="Fuel" type="1D" level="2" scaling="uint8"/>

</rom>

c.) Save the file
3.) Open up the <<ECUID>>.xml file using NOTEPAD (right-click on the file, open with ... pick notepad). The file is usually found in C:\Program Files\OpenECU\EcuFlash\rommetadata\mitsubishi\evo

a.) add the following line after the last table, but before the </rom>:
<table name="High RPM IPW Step" address="1685" level="2" scaling="uint8"/>

b.)
... this should result in the end of the file looking something like:

Quote:
</table>

<table name="High RPM IPW Step" address="1685" level="2" scaling="uint8"/>

</rom>

c.) Save the file
4.) Launch ECUflash and take a look at your ROM image. You should have a new map called High IPW Step in the Fuel section. If you cannot see the new map, please check to see if you have set ECUflash to at least "Advanced" level access (Options -> Options -> User Level = Advanced).

========================
If you want to look at the value in RPM instead of converted HEX, add the following line to your evo7base.xml, in the same section as all the other <scaling> items:

<scaling name="IPWStepTargetRPM8" units="units" toexpr="x*31.25" frexpr="x/31.25" format="%.0f" min="0" max="8000" inc="1" storagetype="uint8" endian="big"/>
Once you've added this line, you can set the scaling of the High RPM IPW Step table to "IPWStepTargetRPM8"

l8r)

[trinydex]

this is my first attempt at creating defs and this is what I came up with for this value :)
http://forums.evolutionm.net/attachment.php?attachmentid=95951&stc=1

corrected
http://forums.evolutionm.net/attachment.php?attachmentid=95954&stc=1

[trinydex]

why is data endian different in some of the scalings?


I think it refers to which side has the most significant bit, but I'm not sure.


So, I just went through adding this to my deffs for on 03 94170008 and the value is comming out to be 7969 rpm....... the hex value was 255. Anyone else seen this???


Ok. Well I changed the address from 1285 to 1685 and now have 7000 so I guess for the 94170008 ecu the correct address is 1685.


is this the reason why the car goes suddenly rich at 7k+rpm??
so if i change this to 7500 rpm, then i should also richen up my fuel tables at the 7000rpm?


This is actually the lean spool disable rpm. My rom has the enable rpm at 2500 and disable at 7000. I have already tuned out the rich transition at 7000 so I have not tried moving the disable rpm up cause I don't feel like retuning the afr curve at the moment.

In general, if you have tuned out the step in afr, you will need to richen it up when you move the lean spool disable up to 7500 or so.

I plan to move the disable rpm to 2500, same as enable rpm to turn off the lean spool all together, but I'm not sure what the result of that will be down low so I'm holding off till i have more time to retune the afr's

[trinydex]

Well here's the missing piece of the puzzle

table name="Boost Enhancement (Anti-lag) Table" address="39b6" level="4">
<table name="RPM" address="6400" />

7 items and rpmstatlimit is the scaling for the axis

http://www.alpine-heat-ac.com/mitsu/antilag.jpg

[BHCevo]

This helped me a LOT.

I corrected the IPW stepup at 4531rpms...this can be done by looking at the ecuflash
"Boost enhancement (anti-lag) Stop Rpm" table and changing this number from 4531 to 2500.

I no longer have the very inconsistent AFRs that I used to have...this hobbled my previous tuning efforts considerably.

My afrs are now between 10.8 and 11.2 regardless of gear or driving conditions.

I still need to do a back to back doublecheck that this was the issue...I did some hardware changes at the same time too. Im going to go out on a limb here and say it anyway.

I described this in a previous thread in some detail...let me link to it here: http://www.socalevo.net/forum/index.php?topic=32017.0

My evo is a 03 with cpuid 94170008. Use the latest version of ecuflash with developer access enabled.

I left the IPW step at 7k rpms as is...maybe ill get around to dropping it to 2500 rpms at some point and see if that helps stabilize my afrs even more.

[trinydex]

can you go into more detail about what the problem was and how this definition could correct that specific problem and what your theory is on how it corrected it? or the reasoning behind why it works.

[BHCevo]

Sure triny.

I explain the problem in exhaustive detail, including logs etc. in the link I provided above.

In a nutshell, my power delivery was erratic over driving any real length of time. Examining my logs, it was clear that I was a full point leaner than i tuned for on the occasional WOT event. Sucky, no?

Turns out that our ecu has a Boost Enhancement (anti-lag) table that allows for more fuel delivery as rpms climb. Other people on evom with a similar problem to mine were reporting better results by messing with this table.

Oriiginally it was thought that the enleanment was started at 2500 rpms, then got richer at 4351 rpms, then finally got even richer (enleanment turned off entirely) at 7k rpms. These rpms are changeable in ecuflash, so by changing the 4351 and 7k rpms setpoints to 2500 rpms you turned off this feature entirely and could tune without worrying about sudden enleanment/enrichment.

Alternatiely, you could raise the 4351 and 7k values to 8k and also not worry about sudden enrichment/enleanment between 2.5k and 8k.

My logs suggested that more often than not my sudden enleanement happened whenever my revs dropped below 4k or so...which is pretty much what you would expect with the boost enhancement table value set at 4351 rpms.

I changed the 4351 value to 2500 and presto the problem disappeared (verified with logs and driving experience). My 7k value is still at 7k but I rarely eceed 7k so I haven't messed with it yet.

Finally, it looks like the Boost Enhancment table actually has more than three values ...it actually is a table that gives values every 500 rpms or so. This is shown in the last quote trinny gives from 3gturbo from evom. So I really should be messing with this to make my correction smoother ...this is what I recommend others do.