This thread was very interesting and had some good info at the beginging but the last two pages are starting to sound like battle of the egos and starting to bore me

To the customer who started the thread . . . .

I have found that Evo Scan is not doing a good job at reporting several parameters particularly on partial TPS position cruising. Other tuners have also reported other issues they have noted including inaccurate coolant temp readings.

I won't be participating in this particular dialouge further so if anyone has any specific questions directed to me PM me

Thanks and enjoy

 

EFIxMR


OK, obviously this is getting out of hand and getting into a pissing match.

I think both of us are reading past what each other is trying to say.


For the most part, I completely agree with you. The only part we seem to differ is that I trust the stock ECU knock detection and you don't.

This isn't even really about EvoScan. EvoScan is simply passing along the results of the calculations that the ECU has come up regarding knock....EvoScan isn't doin any calculations at all. I'm just saying that people shouldn't go tuning and ignore this value.

If you disagree, so be it.


Eric

 

Question, if the stock knock sensor is so good a detecting knock... what good does it do when both the low octane/high octane fuel and ignition maps are identical?

 

I don't think Al is responding in this thread anymore, but I can try to answer this for you.

The ECU doesn't simply use the low/high octane fuel and ignition maps for knock correction. That's what I was referring to earlier when I posted about the ECU's algorithms and calculations about knock.

It will pull timing dependant on the number of knock counts, irregardless of your low/high ignition maps. The DSM ECU was .35* per knock count.


Eric

 

So it will go to the low octane map only when knock gets really bad? Or am I reading this wrong?

 

I don't know anything about the DSM ECU so I cannot comment on that, but my hours of dyno testing on the EVO ECU shows no timing pull when the High/LOW octane maps are the same.

This is evidenced by the torque curve remaining in the same pattern on the dyno.

 

Judging from some of my logs, I think Eric is correct. I've had slight pulled timing due to low knock counts on my car. Timing was nowhere near the low octane map, but slightly below where it should have been on the high octane map - and the gradient does seem to match the number of reported knock counts from the ECU. I'll try to do some more tests when I have time.

l8r)

 

That's very interesting, but I would imagine the Evo to act similarly to the DSM ECU.

Did you happen to log anything on those dyno pulls, like timing or airflow, etc?

Are you sure that you were getting enough knock on the pulls to see the timing being pulled in the torque curve?

Perhaps other people can chime in with their experiences with the maps identical.


Eric

 

in that experiment, if you made the lower octane map different there would be an instant torque drop.

what tells me that is phantom knock is because it is FAR from MBT. MBT is what the original unaltered torque curve looked like. if you advance the timing 1 degree and it gains 15 whp that is far from MBT.

Also it occurs under high rpm where the boost has taper significantly, volumetric efficiency per stroke is way down compared to peak tq. Piston speeds are higher with less time for dentonation to develop.

The best knock sensor is between your ears.

 

Eric do you not want to answer my question:

 

You have a lot to learn

You do not verify the pulling of timing by looking at a dyno curve

You verify the pulling of timing by comparing your timing map to what the ecu outputs to the coils - this is how you deternmine when the ecu has pulled timing

Your position about the high and low octane maps and the pulling of timing is 100% incorrect

 

I’m not a super tuner like you EFIxMR. However I have about 4+ years and 60,000+ miles of studying the stock DSM knock sensor system in my personal car. If Mitsubishi didn’t get stupider since 1990 (which they may very well have) then it is extremely good. It even works well for modified engines. This includes ones with pistons, rods, cams, and even upgraded valve trains.

I have fabricated a hearing aide amplifier that I can clip on the back of blocks to “listen” to knock. To be honest, I like the stock knock sensor better. Too much other noise for my human brain to filter out.

To be honest, I’m an idiot so don’t listen to me. I don’t even own a wideband, I’m too cheap. But I’ve hacked my way to 11.60’s at 126 in a stock 161,000 mile motor tuning with nothing but the stock knock sensor. Yeah, I’m an idiot. That damm sensor has saved my *** numerous times though, something to think about. Not impressive, I know. But that’s the kind of faith I have in the sucker. No wideband, don’t pull plugs, nada.

Now, we have a complete misunderstanding of the evo ecu and how it pulls timing. We don’t even know what it does with the knock sum. We have to learn so as to better utilize the stock knock sensor to our advantage as has been so successfully done in dsm’s.

 

I really cant stand it when someone who I would consider a flash newbie starts to draw conclusions and makes inccorect statments of opinion and casts them as firm FACTS.

I am really not interested in debating these matters with you as I feel it would only serve to further enhance your now limited understanding of how eaxctly the stock ecu works.

I have been tuning evos on a daily basis since the first evo was delivered here on the east coast now over 4 years. Thus far there have been ZERO reported tuning related engine failures out of the hundreds and hundreds and hundreds of customers I have tuned one way or the other. Some of my customers have gone over 100,000 miles on my tunes and have particpiated in every form of sanctioned racing.

Oddly, the customers I have tuned are consistanly the fastest out of any reflash tuned evos in the US and enjoy reliability and durability as good or better than any other tuner - or stock for that matter.

I am not here to detract any other tuner but to rather let my work speak for itself.

I find it very regratable when other tuners try and advance their notariety by attacking my work.

It is even more bizzare when the assertions being conveyed are 100% incorrect.

 

yea, i need to learn how to read hex natively, so i wouldn't even need a ecuflash software right?

 

P.S. Al, do you remember what happened when David Buschur pulled the knock sensor in his RWD Talon? LOL