What the hell does this have to do with what you quoted from him. It is like you have this saved somewhere and you copy and paste it all the time. He just said what does the stock ecu do when low/hi maps are the same.

 

I think what he is trying to say here is that if there is knock with hi/low oct maps the same and the ecu detected it it would pull timing and of course if it pulled timing the torque curve would drop.

 

rpm----ks---tim
4343---0----8
4500---2----7
4750---2----8
4937---1----8
5125---1----9
5312---1----9
5500---1----9
5687---0----11
5875---0----11

Could you please tell me what are the timing values that are in your timing map at the above rpm and load levels? How far off are the above numbers from your timing map?

 

Thank you for the response Eric.

 

I apologize if i've come off the wrong way to you Eric, as this post above is extremely informative about the DSM ecu.

However, my independant testing with the EVO ECU shows different.

Mchuang got it exactly correct.

When you make the low det and high det maps the same provided that the fuel and timing maps are done correctly it will make the engine produce an extremely clean dynograph. Provided that you put in just the right amount of timing in the correct places it will expose an engines optimal Maximum Brake Torque curve for that given fuel.

Maximum Brake Torque Timing is defined as

If you were to change the low det map back to stock, the above optimal curve will be changed certainly, maybe not in all spots but there will be dips. Those dips are where the ECU believes there are knock.

Change the low det map identical to your high det map, and immediately those dips will be patched. The percieved knock is still detected by the ECU but it has been overriden.

This is fact.

While making the maps identical draws a perfect dyno graph, it affords no knock protection IMHO, which is gathered from my testing.

Leaving an engine on MBT timing on pump fuel should never be done unless...

1. It is your own car, and can man up to the consequences for running a hot tune.

2. You're customer specifically request it, and you have informed him of the consequences. (You'd be surprised how many people actually want this)

However that is not to say leaving the low det and high det maps identical are necessarily wrong or dangerous. If you detune the engine by retarding the timing a decrease in 5% - 10% in torque it affords a good safety margin.

This is how it is done tuning stand alone ECU's all the time without failure given a competant tuner.

In many ways this is a hack, and the guys who criticise this technique are justified because to tune with twin maps is substantially more time consuming and comprehensive.

Twin maps will afford an even greater safety margin, and if done correctly will not produce large dips in torque.

Still in some areas, parts of the map can be made indentical if your experience tells you the ECU is picking up phantom knock. These areas are found through sound judgement, and extensive testing on your own vehicle and not customers.

I consider the techniques above only to be tried by experienced users, follow at your own risk.

Thanks!

 

So are we saying that we think the ECU interpolates between the high and low octane maps depending on knock count?

 

That is a definate possibility

 

EFIxMR,

No hard feelings. We're just both relaying our experiences. We're all here to teach each other.


Perhpas more people can validate this when we get better logging abilities and dyno time, as will I soon. I have barely had any time to mess with the Evo Ecu as much as I would have liked, but I find it very strange that you would see no timing retard when you made the maps the same.

Knock correction uses different tables and calculations than just the low/high octane fuel/ignition tables. However, your experiences tend to show otherwise. Since most of my experience is with the DSM ECU, I'm not sure what is truly correct at this point.

At first when you posted about your torque curve, I didn't reply because you could have had timing being pulled throughout the run, which would still produce a clean looking torque curve, as long as the knock was consistent and roughly the same amount through the pull. But since you said that with the same exact conditions back to back making the low/high maps the same didn't produce knock and putting the low map back to stock did, that is extrememly interesting to me.

I will still take it with a grain of salt, though, until I can substantiate the claim. But I definitely appreciate the good data. Sorry to get off on the wrong foot earlier.

And I do agree with your tuning methods concerning MBT and pump gas. Without getting to technical, that is basically what I was trying to say about getting a safe, knock free tune. Obviously, you want as close to MBT as possible, but depending on your AFR and boost you may be nowhere close to that. And some people like to tune for low boost, high timing (close to MBT) and some like high boost, low timing, and don't even care about approaching MBT.

But regardless or the tuning method, the most important factor that I use is knock, as recorded by the stock ECU. I always tune up to the point where knock would begin to occur and then back it off in one of three ways, either fuel, timing, or boost.


Thanks for all your info,
Eric

 

Eric,

Please keep us informed on what you find with the EVO ECU, so we can truly figure out what's going on. I'll be heading out to the dyno today to tune a friends car, and ill will try to duplicate what I was describing earlier. I will document the process, and print out the actual graphs of the change between tuned identical maps and a tuned high octane map with a stock low octane map.

 

I wonder if ECUFlash could pull that number from the ECU? BTW, there are unknown parameters that it can pull. EVO4MAD gave the option to poll the values in the latest ECUFlash version. We've just got to figure them out. Has anyone messed around with them yet?

 

Yes, please do "truely figure out what is going on" - FIRST

And then, afterwards post on the internet lyour conclusions about HOW it works

This is more helpful than stating what you assume first and then finding out what is going on later. This is especially the case when what you are asserting is not correct.

Just my two cents


PS - Just for reference, the ideal way to tune an evo is so that there is a total absense of knock activity what so ever -and a significant margin of saftey below the knock threshold. Of course on 91 octane vehicles and with the new summer gas that some areas are getting this can be a very tedious process. Knock can be RESPONDED TO by the ecu in a very quick manner however it does create physical stress upon mechanical parts when it occurs no matter how briefly and it also will result in an unsmooth feeling as the stock ecu pulls the timing only in small areas of the map. I prefer to set up my tuining so that the cars run totally under the knock threshold.

No matter WHAT you set the so called high and low octane maps to - if the ecu picks up what it calculates as knock activity it can and will immediately PULL as much timing as needed to remove the knock. I have seen some evos with engine noise problems go up to - 9 degrees of timing advance all the way through the power band with both ignition maps set the same.

Again, the only way to test or confirm this theroy is to fix both maps the same and then either lean out the car or raise the boost enough to cause detonation. IF you log your timing you will see that the ecu is pulling timing in direct relationship to the knock count - IRRESPECTIVE of what the relationship between the high and low octane maps is.

Of course all of this gets more dicey in a IX evo which has 6 - 7 ignition maps depending on which ecu version is used.

Finally - there is nothing wrong with setting the multiple ignition maps up with different levels of timing and many of my maps are set up in that manner.

However, in certain situations it is very advantagous to fix the maps the same for reasons which I do not care to discuss.

My main point is to disagree with the claimed assertion that the ecu can not pull timing if the maps are fixed the same - this is totally false.

 

I wish my e-flash agreed.

 

This is very true. I actually saw this happening numerous times with evoscan logs on my car. Not only timing.... fuel also get super rich.

 

E FLASH! is a base flash which is geared towards do it yourself tuners so that they can have a close base starting point which matches their combination of modifications

Many customers actually use it as their final map and just install it and drive and the E FLASH! will generally work very well for this kind of application


However - E FLASH! is not a replacement for a full custom tune which takes me 2 hours of testing and adjustment to complete

Particularly on 91 octane california fuel getting the car to achieve maximum power, avoid knock induced timing retard and have a great "feel" takes a lot of time and careful balancing. 91 octane fuel in particular is a tough challenge.

My concept in creating the E FLASH! maps was to make it easier for do it yourself tuners to have a good starting point and a safe starting point. It is not a replacement for fine tuning as indicated.

 

Correct me if I am wrong, but an E-flash is the same thing as the mail in flash except the customer actually applies the flash to the ECU.