That was on an Evo IX, right? Could be less timing on the MIVEC too, no?

l8r)

 

Says 8MR on the dyno sheets.

 

Duh ... nevermind.

l8r)

 

Andy, what if the pull that had both maps had too much timing in the mid range and was knocking but didn't have a lower timing map to drop to?

As soon as you put the low octane map back to stock the midrange just takes a crap. Maybe it was finally able to fall back into the low octane map? (technically it was doing it the first pull too, but since both maps where the same, no change.

The too much timing in the midrange theory might make sense since the peak HP didn't change much.

If you had timing logs it could tell us things like this.

 

Another thought... if there is an octane value that interpolates between low and high octane maps (like a long term ignition trim if you like) then if we equalise the maps then it has no effect. However, that doesn't mean that the "short term ignition trim" (to coin a phrase) couldn't pull timing, whether this is 90/255ths of the knock sum or not...

I'm thinking that some logs we see where there is a persistently high knock sum are due to high and low maps being the same or too similar? In a properly calibrated setup the knock sum should be closed loop controlled by the octane value trimming the timing.

So I'm wondering if we just kill the long term learning by equalising the maps?

Phantom knock or real knock, the effect on torque would be similar?

Just some ideas from playing with different ECUs on different platforms in some detail, and being a newbie at reading up on the DSM ECUs.

 

Real knock occurs when timing is advanced too far from MBT. An engine does not make a smooth or clean horsepower or torque curve when it is knocking. I've tuned many engines using this technique (cars that never came without knock sensors), and it is an agreed method amongst many tuners and efi manufacturers I know.

It's likely that the ECU thinks its knocking, but really isn't. If it were really knocking the plugs would show evidence of it, parts would get hurt, and the people running this sort of torque in Norcal would have broken engines.

This sort of goes back to another conversation you had with Reggie (top whp in norcal on pump) you warned him his tune was too hot... but his car is still fine.

Back when we first started talking about how much a BPU EVO should make on pump my position was actually that 300 whp on the stock turbo isn't safe. Then you did it and ran your car relatively trouble free, and now cars with near 300 are showing up all the time.

I use the dyno to log boost, afr, and i check and pull plugs periodically on the dyno, as well as after road testing.

As for my timing logs, I prefer to keep that to myself, as while I will freely share ideas and discuss topics... the timing profile I use is generated after investing our companies time and money doing testing on the dyno... and using my own personal car as a guinea pig to determine what's safe and what's not. Sorry, hope you understand for me as much fun as this is... It still is a paycheck.

 

Guess we have to take your word for it on this one since you aren't posting logs.

Do you do a WOT run and shut the car off and let the car coast down and check the plugs right then?

Ok? I'm not warning anyone right now. Mearly bringing up an idea/theory....

Pot meet keetle? (See above)

So, is your standing on this, that even though it looks like the car is knocking and reverting to the low octane map in the 2nd graph, you think that the car "thinks" it's knock but really isn't? So we are back at square one where everyone is accusing you of thinking that the stock knock sensor is not to be followed/trusted?

 

EFI,

Could you describe what knock would look like on a spark plug? What should I look for?

 

Little metallic specks on the porcelin. To make sure the motor doesnt clean the plug, you really should do a WOT run, shut the car off at the end of the pull and let it coast down. Hopefully on the dyno since you won't have power brakes.

 

I don't do it after every run, but yes that is one technique to do it. It can get even more technical than that however. On extremely high hp cars 500+ i find where peak tq is and stop sampling at that rpm, and turn off the engine right at that point, to check plugs.

On the street it is impractical to just pull over immediately after going WOT. Also, if your car is tuned correctly and not making the plugs black from being rich, you can get a good idea of how the engine is running even if you've driven for some time.

Detonation actually causes transfer to aluminum on to the porcelain, its not a powder that can be wiped off. it is bonded on there strong.

It depends on how much difference there is in torque. 30 wtq is a huge margin away from MBT. If there was only a difference in 5 or 10 wtq I would be more inclined to believe that is real knock.

How I decide on how I tune the car depends on the client.

If he is paranoid, and wants his engine to last forever then there's just no way his engine can be tuned to 300 wtq or 300 whp with the basic bolt ons we are talking about.

What is completely safe worry free is 275-280 Mustang whp and wtq conservative tune.

Average tune I consider to be 280-290 Mustang whp and wtq

A fully optimized tune is 290-300+ whp and wtq.

The car whose dyno charts that belongs to did leave the dyno with a seperate low det ignition map, but it is significantly altered from stock. It is my close friend's car, and I have been tuning his various vehicles for the past 6 years, he has given me the car to drive for a week to moniter its performance.

 

When a car is just mildly dentonating, a silverish/grey ash covers the white part of the porcelain.

as it gets more severe the texture of the porcelain resembles an egg shell with the accompanying silverish/grey ash.

Severe detonation the ground strap of the plug will be missing, and at that point I would suspect other engine damage.

Under normal combustion, the porcelain will be clean with no deposits. It will be completely smooth with a slight brown tinge.

 

EFI,

I understand if you don't want to show your timing logs. That's understandable.

How about showing knock logs from EvoScan at least. That would help me a lot to understand what is going on for both runs.


Also, I agree with some of the points that RazorLab made. It is hard detecting knock by looking at the plugs unless the car is immediately shut off after the detonation events and the events have to be bad enough to actually deposit the aluminum on the porcelain. If you see that, it is pretty bad knock, wouldn't you agree?

I'm concerned with not only bad knock, but any at all. I have tuned cars that have knock quite a bit, but the plugs never showed anything, perhaps because they were cleaned off by the subsequent combustion cycles or perhaps because the knock just wasn't bad enough to deposit the metal.

If you get a chance I would like to see the EvoScan logs of your pulls...you could of course hide the timing values.


Thanks,
Eric

 

It is amazing to me that you make a incorrect assumption and then you construct a test to support your position which has no relationship to the assumption you are trying to prove.

The words "leads me to believe" are words that a professional should not use in trying to explain how something works. At least not something as easy to work out as a stock evo ecu anyway.

When I am at the "leads me to believe" level of understanding that is when I just keep my mouth shut until i have verified and tested.

 

I do not have EVOscan, I was considering purchasing it, but after hearing bad reviews on it from our local board I'm not sure anymore.

 

I set the map identical and the numbers are consistently over 290 wtq 290 whp.

I set the low octane map to stock leave the the high octane map alone and it falls 30 wtq.

I set the map identical once again and the numbers are consitently over 290 wtq 290 whp.

It is quite obvious to everyone here but you that leaving the maps identical severely impairs the ability of the stock ECU to pull timing in the presence of what it believes to have detected as knock.

We are not talking about 5 or 10 wtq of noise 30 wtq is a significant change.

I for one did not write the code in the Mitsubishi ECU, and I know my stock ECU isn't made by DYNOFLASH LLC. You might have bought a techtom earlier than I have, but your theories are just your beliefs too.

Do you have the source code to the Mitsubishi ECU? Let's stop pretending what you know to be fact, if you don't.

You don't even say what you think is actually happening. You don't want to share and hide with statements like for reasons I don't want to reveal. How do you explain the phenomenon I showed on the dyno? I tested it, verified it, and quantified it on an $80,000 piece of equipment.