I forget what CEL's I've disabled. I didn't initially have any when I first installed it and ran 9653. Now I don't have emissions and I'm running SD. So I've turned off quite a few CEL checks.

Oh, and my problem was a bad coil pack. I put a new one on my car and it's running great. I even managed to blow a small16g with the car since then. Too much boost and not enough oil pressure to the turbo. I think I had it up around 21-23 psi before the turbo went kaput. An HX35 is defiantly in my future for this summer.

Ill check the codes. How hard was it converting to SD? OR how hard was it making your VE maps.

It didn't take long at all to get the VE maps roughed in to get it running. I'm still tweaking the maps to get it dialed in more.

I've been trying to get the load to equal kPa and it's work well so far. Cold startup is still sketchy, but I just need to add more fuel.

I finally got my cold start to be happy recently, it mostly involved finishing dialing in the SD tables down low but there are some enrichment maps and periphery bits that make a big difference. I'm in class now but I will try to post the relevant maps later along with my rom in case anyone is interested.

Even though I'm on a 98/99, I'd love to hear what you're doing to help your cold starts. Mine are way lean and the car is barely driveable until its at around 150* - 160*.

I saw big improvements just from getting my VE at 500rpm and 1000rpm correct.

I've also increased the values in a couple of the fuel related coolant trim tables.

I can't wait until the 3d VE tables are implemented. It should make things easier to tune.

I guess I should have clarified that i'm not even on SD. I could just use some cold start ideas.

I don't even remember now what all I've done for cold starts. I know I added some enrichment to the cold start tables and then lengthened out their decay timers but since the table names are so similar I'm not sure which is which anymore. All the necessary information can be found in MrFreds advanced fuel control thread on evolutionm.net I though I knew what I did but now I'm less sure. The periphery bits also helped significantly at one point but that was quite a while ago so I'll just post my rom and evo8 ecu guys can compare cold start tables and periphery bits to find the changes and then hopefully reply as to what works.

Included are the rom and I believe the correct definitions to find everything you need. Keep in mind for anyone that would try to use this for anything else that this rom is untuned above about 200 load because I've significantly changed my VE tables since last season and have been tuning from the bottom up. If you run this on your car it might break. That said the VE tables are probably pretty close for a stock longblock e316g car with a TBE an open element filter, and an evo3 IM.

Hubbard, cold start for you is pretty simple probably. Basically all of the cold start enrichment tables assume stock size injectors. Check out MrFreds advanced fuel control thread on evolutionm to understand how everything works but basically you'll need to edit the enrichment tables by approximately the same % that your injectors are different from stock. The only other thing you likely could have done is to change certain periphery bits that affect cold start fueling.

I've changed a few things, and it's helped a little. It still needs some work though. From biggest change to smallest effect.

1.) Get VE maps dialed in for idle.
It can be a pain. I can't just set it to open loop because it will actually change the target back and forth between 16.x AFR and 13.x AFR. Not sure what's up with that, but it cycles almost like it's running in closed loop. I just went off closed loop fuel trims at idle.

2.) Increase MAF Comp vs. Coolant Temp #1 and #2.
I can't remember how much I bumped these. There are 2 different scalings I've seen used. Just increase the colder temps little by little.

3.) Increased the Post Cranking Transient Enrichment vs. Coolant Temp #1 and #2.
I ended up bumping #1 by ~15. And bumped #2 by ~10.

That's where I've focused so far for improving cold starts. The car actually fires up just fine when cold, first crank. It still stumbles a little when cold though, but no longer completely stalls out from going lean.

I seem to remember there are a quite a few Automatics here.

Are there any issues or changes from removing all the Auto Trans pins from the harness? (Torque Reduction 1&2, Torque Execution, and Park/Neutral)

Does your Auto behave the same with the Evo ecu as it did with the DSM ecu?

If anything there is harsher engagement into drive but it could be in my head. Otherwise I haven't noticed anything bad going on.

I've noticed no difference. I've added a shift kit since then and that has helped my car a TON from stock. It even has the girlfriend seal of approval. It gives a little kick when shifting gears instead of that sluggish flare.

IIRC, the torque kickdown wiring pulls timing while brake boosting to help get the revs up before the car wants to pull through the brakes.

You can now do that with the launch maps.

Test my Evo9 ecu in a 2G DSM patch harness today.

Started up on the first try.

I don't really have a use for it since I run a 99 DSM ecu, but the Evo9 ecu was so cheap on ebay I had to buy it.

(I have no idea why DSM ecus are more expensive then Evo ecus.)

The only issue I have is it runs very rich the first 30 seconds after startup, other then that it drives perfect.

Haven't even got any CELs yet, but I suspect the Mivec CEL is on its way.


Also I don't have the bouncing Tach problem people talk about. Does it only happen at certain RPMs?



<object width="425" height="344"><param name="movie" value="http://www.youtube.com/v/OweFwd2r9bo&hl=en_US&fs=1&rel=0"></param><param name="allowFullScreen" value="true"></param><param name="allowscriptaccess" value="always"></param><embed src="http://www.youtube.com/v/OweFwd2r9bo&hl=en_US&fs=1&rel=0" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="425" height="344"></embed></object>

Very nice.

I was thinking when I read your post, what about MIVEC throwing a hissy fit.

Nice harness, looks just like the ones I built. Aren't they fun to make? Haha.

Ceddy said:

Test my Evo9 ecu in a 2G DSM patch harness today.

Click to expand...

I have one of those sitting in my tool box just waiting...

Also I don't have the bouncing Tach problem people talk about. Does it only happen at certain RPMs?

Click to expand...

If you disconnected pin 58 like in the updated diagram on the first page of this thread, that problem won't happen.

So have you programmed one of the outputs on the 99ecu to run the cyclone valves? That would be spiffy.
I've thought about trying to do something similar eventually (Using the BCS since I'm happy with a MBC.) Me like torque.

knochgoon24 said:

So have you programmed one of the outputs on the 99ecu to run the cyclone valves? That would be spiffy.
I've thought about trying to do something similar eventually (Using the BCS since I'm happy with a MBC.) Me like torque.

Click to expand...

Yes I have added Windows Switches for the FPS, EGR, and Purge Solenoids.
Controlled by: TPS, RPM, Knock, Boost, Load, or Speed. Thats in version 2.0, not released yet.

The Cyclone definitely helps the low end. Especially if you have an old worn out 7.8:1 compression motor.

Ceddy said:

Test my Evo9 ecu in a 2G DSM patch harness today.

Started up on the first try.

I don't really have a use for it since I run a 99 DSM ecu, but the Evo9 ecu was so cheap on ebay I had to buy it.

(I have no idea why DSM ecus are more expensive then Evo ecus.)

The only issue I have is it runs very rich the first 30 seconds after startup, other then that it drives perfect.

Haven't even got any CELs yet, but I suspect the Mivec CEL is on its way.


Also I don't have the bouncing Tach problem people talk about. Does it only happen at certain RPMs?

Click to expand...

sounds like we need to start another thread...

Ceddy said:

Yes I have added Windows Switches for the FPS, EGR, and Purge Solenoids.
Controlled by: TPS, RPM, Knock, Boost, Load, or Speed. Thats in version 2.0, not released yet.

The Cyclone definitely helps the low end. Especially if you have an old worn out 7.8:1 compression motor.

Click to expand...

how close to an evo rom is the dsm rom? i believe i read somewhere that 9055 was similar. what are the chances that the evo8 ecu community can get these for our ecu's? or have you already started working with someone on it over at the EvoM forum?????

mike96 said:

how close to an evo rom is the dsm rom? i believe i read somewhere that 9055 was similar. what are the chances that the evo8 ecu community can get these for our ecu's? or have you already started working with someone on it over at the EvoM forum?????

Click to expand...

The 98-99 DSM use a H8 cpu which is a CISC (normal) instruction set.

The Evo7-10 use a SH cpu which is a RISC (Reduced) instruction set.

Even though they are both Hitachi/Renesas cpus, they have totally different languages.


I suck at SH code, one of the Evo coders will have to tackle this. I think RoadSpike on EvoM has started something similar.

Ceddy, which ROM did you end up running on the IX ECU, out of curiosity? (An 8 ROM, or are things working fine with the MIVEC code?)

who has the fastest or made the most power on the stock ecu so far?

logic said:

Ceddy, which ROM did you end up running on the IX ECU, out of curiosity? (An 8 ROM, or are things working fine with the MIVEC code?)

Click to expand...

88580714, its working fine with nothing connected to the MIVEC CAS. Not even a CEL yet. But I only had it running for a few minutes before I got rained out. (I did zero out the MIVEC map.)

I think all the MIVEC code is separate from the regular code. All the MIVEC CAS does is control the MIVEC oil pressure actuator, it has nothing to do with Ignition Timing, etc.


Don't think an 8 rom will work in a 9 ecu. Because of different cpus. Unless I'm wrong?

Cool, glad to know the MIVEC code can be basically ignored. I suppose that makes sense, thinking about how it works; I just hadn't spent a whole lot of time considering it.

As to compatibility: the CPUs are different generations of the same basic processor. The subset of the instruction set used by Mitsu (that I've disassembled, anyway) is identical. All of the interrupt vectors are identical. A small number of registers moved around, but nothing so substantial that you couldn't easily programmatically fix them up. Available ROM and RAM space is enlarged, but again, it's a superset; the address space expected by existing VIII code is there, with extras. (You can probably see why the idea is intriguing to me.)

Someday, I'll stop being spastic about what projects I'm working on, and actually settle in and do one thing for long enough to see it through to completion (ie. actually get the damn IX ECU swapped into my wife's 1g, and confirm my suspicion about VIII code running properly). I think my dogs have a longer attention span than me.

Right now, my biggest concern would be finding all the cooling fans (rad and a/c) tables. I'd like to set both fans to come on a bit sooner. Plus, someone has mentioned on the EvoM forums that the Evo's control the fans using PWM, and that the DSM's relays might not like that so much. If I could get all the correct table addresses (for 9653) I would like to set the tables to either 0 or 100 so it wouldn't pulse, just steady on or off to the relays.

Cool that you got that working cedy but is there any actual gains from using a 9 ecu over an 8? I don't really see much advantage. On a sidenote knochgoon24 did you ever work out the fan tables like you wanted?