openecu.org

Now that I have your attention. I would like to provide some feedback from my dyno experiences so that we can get this program polished up for these cars.
I use this program everyday and I LOVE IT!

The deff files are so very very incomplete and haven't shown any improvement in months. They don't work properly with the unknown maps on the ecu.

For example: when adding boost to an evo 9 you get serious issues.
1. Knock retard FOR NO REASON. There is no possible way whatsoever that the car is knocking and it still pulls timing. So, I increase the MAF size and it stops pulling timing. This only works for a few extra psi.
Maybe an overactive knock sensor. Subarus had that for a long time. Maybe this car does too. However, adjusting the knock filters does absolutely nothing. Unlplug knock sensor? Same thing.

My conclusion, THERE IS SO MUCH MISSING from the ecu defs that the evo9's are practically not safely open source tunable for more than say 60hp on a STOCK car. Sure you could do more, but you have to lie to the ECU and bypass safety features.

Subaru has implemented a "Desired Torque" map. When you exceed the desired torque it cuts Wastegate Duty Cycle. On the LGT's its 30.2%. So in order to properly tune those bastards you need to get that map or add your own boost controller.

My guess is that mitsu has a similar map. That instead of cutting wastegate duty cycle, it cuts timing.

Can ANYONE look into this? I am at the end of my resources. I can give EVO9 ROMS out the yin yang if they are needed for evaluation.

I am one of the larger monetary contributors to the enginuity program and I'd be willing to contribute monetary to ecuflash as well if it helps make results.

Desired Torque maps have been around for quite a while now however from what I've seen of the evo rom's they are not that complex.

Surely if you've hit a torque limit than the ignition timing would be reduced by a torque source not from knock retard. Also retarding ignition as a torque limiter on a turbo charged engine is just asking for problems. If there is such an algorithm then it would be more than likely reduction in wastegate duty cycle to reduce the boost.

I would agree with you except one thing.

One the dyno, we could only gain 60 peak HP. We could get way more than that up at redline(80-100), but peak HP was only 60. This was in a boost range of over 6psi and an AFR of 9.5 all the way up to 11.5. We varied all the parameters in all sorts of combinations. No matter what we did, the ecu would adjust timing to limit the peak HP and torque. It would retard timing as far at 4 degrees AFTER TDC to limit the power. All runs had the same output with very different timing, boost and fuel strategies.

So again, I would agree, but the evidence is overwhelming.

Mitsu has done worse things in my opinion, but this still sucks.

Ecuflash has worked absolutely fine on our Evo IX tarmac rally car.
( Australian spec GSR, ver 13). We have been very happy going from 170 all wheel kw to 230 awkw. The ecu only pulls timing (according to the knock learn function in the datalogger) if the the engine starts to knock (is obvious as well). This appears at about 23psi on our Aus spec 98 ron fuel. As soon as we use the race unleaded Elf fuel, the knock learn hovers around 1 -3 counts and all the horsepower reappears on the dyno! So far it has done everything I have wanted to on our ver13 def files.

Cheers Mike

Interesting.

Been tuning tons of evo 9's.

Some making over 400whp on a mustang. All using ecuflash. Are you doing something wrong?

Just wanted to stress that this is on a stock evo. So we are dealing with all the limitations of the stock exhaust and boost control system.

I have flashed MODIFIED EVO9's to 350whp, but they were modified with a seperate boost control system.

I still had issues where the car was pulling timing for no reason.
10.5 AFR, 22psi, 5000 rpm and the car was pulling timing at 2 degrees AFTER TDC. Now that isn't how I tune, but it was an experiment to see what the heck the ecu was doing.

If you cheat the ecu and make all the spark maps the same, sure you can get more timing out of the car. However, this isn't how it is supposed to be done.
The car should be able to have all 6 timing maps (or is it 8?) set up in the factory progression so it can swap between them like it is supposed to. It should be able to provide protection for ACTUAL knock events, not these mystical ghost knock events.

For example, the same car, with the same AFR, same boost, same timing did NOT pull as much when I upped the MAF Size value. This leads me to believe that this isn't supposed to be MAF size, but instead, MAF Limit table.

Anyway, I know the title sounded bad, but I am actually trying to provide some positive feedback so that we can get this all to function properly.

BTW, thanks freon.

oh one more thing.

Someone had mentioned that it makes more sense to just pull WDC to limit torque and not to pull timing. That is correct. However, I know as well as you that Mitsubishi knows of this tricky little device called a Manual Boost controller. The only way mitsu could protect the engine from a poorly adjusted MBC would be to retard timing. WDC aint going to help. Making the car lean, isn't going to work, making the car rich can only work until the injectors are at 100% IDC then it's lean and boom.

That is why the car pulls timing when one of these limits is reached.

I have seen logs and dyno runs from some of the "famous" evo tuners. And they all cheat the ecu. (I don't want to give names, but if I get flamed for this I'm throwing them under the bus too) Some of them even do what they say is publicly a bad idea (making all the timing maps the same for example). In ALL these flashes, the logs show 12 to 20 degrees of knock correction.
Is the car actually knocking? Out of the ones I've seen NO. The AFRs and the engine were safe. But the ecu is pulling timing anyway.

That's not how it should be. The ecu should pull timing for knock. Not for airflow limits or other garbage.

I am willing to submit a handful of logs and ROMS to Freon if it helps the cause.

Since I suck at XML, hex editing and the such. I'm asking for anyone's help. I would like to contribute more to the ecuflash project; as I currently contribute to enginuity on a regular basis.

imprezarsx wrote:oh one more thing.

Someone had mentioned that it makes more sense to just pull WDC to limit torque and not to pull timing. That is correct. However, I know as well as you that Mitsubishi knows of this tricky little device called a Manual Boost controller. The only way mitsu could protect the engine from a poorly adjusted MBC would be to retard timing. WDC aint going to help. Making the car lean, isn't going to work, making the car rich can only work until the injectors are at 100% IDC then it's lean and boom.

That is why the car pulls timing when one of these limits is reached.

I have seen logs and dyno runs from some of the "famous" evo tuners. And they all cheat the ecu. (I don't want to give names, but if I get flamed for this I'm throwing them under the bus too) Some of them even do what they say is publicly a bad idea (making all the timing maps the same for example). In ALL these flashes, the logs show 12 to 20 degrees of knock correction.
Is the car actually knocking? Out of the ones I've seen NO. The AFRs and the engine were safe. But the ecu is pulling timing anyway.

That's not how it should be. The ecu should pull timing for knock. Not for airflow limits or other garbage.

I am willing to submit a handful of logs and ROMS to Freon if it helps the cause.

Since I suck at XML, hex editing and the such. I'm asking for anyone's help. I would like to contribute more to the ecuflash project; as I currently contribute to enginuity on a regular basis.

The factory ECU has a boost cut to stop any manual boost controller from introducing too much boost into the system. From tuning other NA sytems torque reduction is based on amount of torque reduction request vs spark retard. If they implemented a system like this then you would end up with the majority of the combustion event within the exhaust which would only aid in spooling the turbo to produce more boost = more torque.

The gains that you have quoted seem quite reasonable to me on a standard exhaust system. Also if your running around 4deg of timng ATDC then you may have too much boost for the specific fuel you are using.

From what i've seen the factory exhuast system does't seem to pose too much of an airlfow restriction but more of a thermal restriction - as you will see once you upgrade.

If they implemented a system like this then you would end up with the majority of the combustion event within the exhaust which would only aid in spooling the turbo to produce more boost = more torque.

Kinda. A lot of energy is lost out the exhaust instead of pushing down on the piston. More boost doesn't mean more torque if the timing is pulled to offset it that much. Like anti-lag for example.

I have seen cars with Crank angle sensors retarded (owner fitted it) and made 1bar boost but went nowhere. The owner could not believe the boost guage was right until I refitted his crank angle sensor correctly so the timing was right. Then it went like snot. It was an RX7 but same rules apply.

"more boost = more torque" only works if timing remains constant.

I don't know what is missing, I've got everything that is commonly available for US 8/9 for the 9 GSR/GT/RS JDM models (except knock filters which they don't have), in addition to secondary air solenoid controls, reprogrammable request IDs, custom patches. My 9 loves tuning and gives good results, fuel pressure changes, methanol, 1.9 bar boost etc. It is knock resistant, smooth, predictable, spools up to 1 bar at 2200 RPM etc. Responds superbly to MIVEC mapping. I have some maps running in RAM, and a live editing protocol to adjust them. I'm going to work on rewriting the comms a bit to do high speed multi-byte request IDs etc. Nothing sucks about the 9 with Ecuflash

rob82 wrote:
The factory ECU has a boost cut to stop any manual boost controller from introducing too much boost into the system. From tuning other NA sytems torque reduction is based on amount of torque reduction request vs spark retard. If they implemented a system like this then you would end up with the majority of the combustion event within the exhaust which would only aid in spooling the turbo to produce more boost = more torque.

The gains that you have quoted seem quite reasonable to me on a standard exhaust system. Also if your running around 4deg of timng ATDC then you may have too much boost for the specific fuel you are using.
.

it definitely doesn't make more torque. Because if it did, I wouldn't be trying to figure it out.
However, the fuel up here is pretty crappy. Combine that with the poor engineering of the 4G63 and we probably have a winner.

Here's a dyno graph from yesterday. This car has an aftermarket exhaust (test pipe and back), Fujit-something intake and hard pipes. Its an EVO9.
This is with the factory boost control system.
These 9's put 220 on our dyno in stock form.
This car showed up with another reflash by someone else. It wasn't tuned, it was just off of another similar car. The plugs were factory and it was occasionally blowing out spark both before and after the re-tune. The most noticeable event is the one right near the end of the pull.
we were going to install a boost controller, but the plugs weren't hanging and the injector duty cycle was maxed out as it was so we canned that idea. Ordered some NGK iridiums and we're going to give it a shot in another week, but the fuel is still going to be maxed out.

imprezarsx wrote:I would agree with you except one thing.

One the dyno, we could only gain 60 peak HP. We could get way more than that up at redline(80-100), but peak HP was only 60. This was in a boost range of over 6psi and an AFR of 9.5 all the way up to 11.5. We varied all the parameters in all sorts of combinations. No matter what we did, the ecu would adjust timing to limit the peak HP and torque. It would retard timing as far at 4 degrees AFTER TDC to limit the power. All runs had the same output with very different timing, boost and fuel strategies.

So again, I would agree, but the evidence is overwhelming.

Mitsu has done worse things in my opinion, but this still sucks.

If you are experiencing knock the ECU will pull timing. Have you tried adding octane to your equation. Are you logging your dyno pulls to see your load and timing values? on my car if there is too much initial timing or not enough fuel or octane it will cause knock and then a downward spiral from there. This causes my ecu to pull timing. The cure could be less initial timing,better fuel or adjust your fueling maps to compensate for more power.

with the mods you list 60 more whp seems acceptable. do you know if you are hitting fuel cut? I am basically a noob but maybe bigger injectors and a fuel pump are in order?

PS I saw aDynograph of a tuned 9 with only a cat back and Electronic boost controller and a professional tune/remap hit 295 whp on a Dynojet so I beleive the factory ECU is very tunable.

there are 500 HP EVO's running around on stock ECU as well

Your initial complaint frankly doesnt make sense. Further, the fact that you have a car pig rich also cant be helping. Having spent almost a year tuning the Evo both 8's and 9's I cant say that I agree to most of your observations. It would appear that you might be in California in which would explain much.

My personal IX has run as much as 25psi of boost on the stocker (stock turbo) without knock and at 6* of timing (really good air, typically it hits higher load cells and runs in the 3-4* range). I run Shell 92 with no additives or methanol or the like. My AFR's tend to be in the 11.3-11.5 range under boost as well. My MIVEC settings are unique which I believe has made a significant difference in the boost threshold of the Evo on pumpgas. Quite by accident I hit 27.5psi on a Green the other day (I have upgraded recently) and with timing at 1* and AFR's at 11.3 it still didnt knock. I lifted rather fast, but still it shows that you are apparently facing variables that you havent taken into account.

For the record my boost is logged via a 3bar MAP sensor and EVOscan. It coincidentally backs up my autometer boost gauge as near as I can tell perfectly. I wish I could help answer some of your questions but I really havent seen the problems you are talking about so I have no frame of reference.

*edit*- I was thinking about this more today at work and there are a few things I do agree wholeheartedly with you on. The inability to disable CEL like you can with the Subaru Maps is probably the biggest thing. I think a lot more development is going into the recoding of the Evo ECU than you may realise though. I mean there are cars running on nothing but a 3 bar MAP sensor and the stock ECU without a translator, Blowthru MAF cars on the stock ECU with no translator, etc.

Hi all,
I have been trying to open JDM ecu maps fro the evo 9 with ecuflash 1.40 and it always crashes. Any help on how to fix this would be highly appreciated.

your def file is corrupt ,,what rom id is it.