Notices
ECU Flash

My evo ix knock sum ok?

Thread Tools
 
Search this Thread
 
Old Jul 16, 2006 | 10:00 AM
  #31  
nj1266's Avatar
Evolved Member
iTrader: (6)
 
Joined: Nov 2004
Posts: 3,254
Likes: 13
From: USA
Originally Posted by dudical26
Yea I really have no idea what is going on here. I just filled up on a new tank of BP 93 instead of the mobil 93 i normally get. I am gonna go put back my stock intake back on and put the boost to 19 like stock and see what happens
Let me give you my opinion. Your car is fine. The more I read about electrical knock detection the more I doubt the whole process. Slowly, I am begining to believe that the best way to detect knock is the old fashioned way: YOUR EARS. You can go get an old people hearing device and modify it to hear knock with your own ears. You can also create a knock detection can. Search google and you will find how to do this.

Another way is to learn how to read your spark plugs. I checked my spark plugs and there was no indication of knock on my car.

Another way is to get a good AFR meter and check your AFR logs. A serious knock event is usually associated with a dip in the AFR. It happened to me. I had a serious dip in the AFR and then I checked my timing log ans saw serious timing being pulled at the same rpm that the AFR dipped. Check this link:

https://www.evolutionm.net/forums/sh...6&postcount=23

You can see how my AFR dropped and at the same time the timing got SERIOUSLY pulled. Timing was pulled from 7 to 2. Now that is reall knock pulling real timing.

Last edited by nj1266; Jul 16, 2006 at 10:07 AM.
Reply
Old Jul 16, 2006 | 10:48 AM
  #32  
DynoFlash's Avatar
Account Disabled
iTrader: (91)
 
Joined: Sep 2003
Posts: 16,850
Likes: 0
From: 2003 Evo VIII - Silver
Originally Posted by nj1266
Let me give you my opinion. Your car is fine. The more I read about electrical knock detection the more I doubt the whole process. Slowly, I am begining to believe that the best way to detect knock is the old fashioned way: YOUR EARS. You can go get an old people hearing device and modify it to hear knock with your own ears. You can also create a knock detection can. Search google and you will find how to do this.

Another way is to learn how to read your spark plugs. I checked my spark plugs and there was no indication of knock on my car.

Another way is to get a good AFR meter and check your AFR logs. A serious knock event is usually associated with a dip in the AFR. It happened to me. I had a serious dip in the AFR and then I checked my timing log ans saw serious timing being pulled at the same rpm that the AFR dipped. Check this link:

https://www.evolutionm.net/forums/sh...6&postcount=23

You can see how my AFR dropped and at the same time the timing got SERIOUSLY pulled. Timing was pulled from 7 to 2. Now that is reall knock pulling real timing.

I concur with your observations 100%

The way I like to obeserve knock is with either my ears or by watching what I scan directly off the knock sensor

The problem of course is the IX ecu is very senstive to any sounds in the engine it percieves as "abnormal" whioch can simply mean added engine noise over stock

Its a balacing act to set up the ecu to make decent power levels while keeping it knock free - - AND - - making the ecu happy with the level of engine noise and not pulling any timing

I have seen cars reporting "KNOCK COUNTS" which are not knocking at all

I have seen cars loudly knock that show zero knock count

and I have seen cars with no knock and no knock count pull timing like mad

As tuners we must balance the BENEFITS of the stock knock control system vs. the minor inconvienience of having timing being pulled on occassion

I would prefer to keep the stock knock control system 100% active and protect the engine as much as posisble than risk engine damage by trying to alter the knock threshold buffers

The main thing to keep in mind is that IX evos are very sensitive and the ecu acts very quickly in regard to any knock sounds the knock sensor picks up
Reply
Old Jul 16, 2006 | 10:50 AM
  #33  
DynoFlash's Avatar
Account Disabled
iTrader: (91)
 
Joined: Sep 2003
Posts: 16,850
Likes: 0
From: 2003 Evo VIII - Silver
Originally Posted by nj1266
If you look at the first log, the pulled timing coincides with peak torque. This could be the results of two things. It could be that there is knock and the ECU is pulling timing. But it could also be that the ECU pulls timing by itself at peak torque. Below is a log from my EVO w/o any tune and engine management. Notice how the ECU pulls timing at peak torque.

RPM, Timing, O2
3125.0, 20.0, 0.88
3273.0, 13.0, 0.92
3438.0, 7.0, 0.92
3660.0, 4.0, 0.94
3938.0, 3.0, 0.94
4180.0, 7.0, 0.94
4430.0, 7.0, 0.94
4660.0, 7.0, 0.94
4898.0, 8.0, 0.94
5133.0, 9.0, 0.94
5355.0, 8.0, 0.96
5586.0, 9.0, 0.96
5813.0, 9.0, 0.96
6035.0, 11.0, 0.96
6270.0, 13.0, 0.96
6465.0, 15.0, 0.96
6660.0, 17.0, 0.96
6848.0, 19.0, 0.98
7055.0, 19.0, 0.98
7211.0, 21.0, 0.98
7391.0, 21.0, 0.98
Also - at peak TQ there are additional dirvetrain and engine HARMONICS which may be comming into play
Reply
Old Jul 16, 2006 | 10:59 AM
  #34  
DynoFlash's Avatar
Account Disabled
iTrader: (91)
 
Joined: Sep 2003
Posts: 16,850
Likes: 0
From: 2003 Evo VIII - Silver
Originally Posted by dudical26
Is there any possibility that Evoscan does not work properly on Evo IX's. I only ask because even with my bone stock evo I was getting knock counts. I even lowered my boost to 17Psi or less and I still reported knock.

Here is a log of stock map with an intake (added after the previous incident I just described) at 17 psi.
Code:
RPM	TPS	AirTemp	Knock TimingAdv
3750	100	96.548	0	13
3812.5	100	96.548	0	11
3875	100	96.548	0	9
3875	100	95.504	0	9
3906.25	100	95.504	0	8
4000	100	95.504	0	7
4031.25	100	95.504	0	6
4093.75	100	95.504	0	4
4125	100	94.46	0	3
4187.5	100	94.46	0	2
4281.25	100	94.46	0	3
4375	100	94.46	0	5
4468.75	100	94.46	0	5
4531.25	100	94.46	0	6
4625	100	94.46	0	5
4718.75	100	94.46	1	4
4781.25	100	94.46	1	5
4906.25	100	94.46	1	7
4968.75	100	94.46	1	7
5062.5	100	94.46	4	5
5156.25	100	94.46	4	7
5250	100	94.46	4	7
5312.5	100	93.416	4	7
5343.75	100	93.416	4	7
5406.25	100	93.416	4	8
5500	100	93.416	4	8
5562.5	100	93.416	4	8
5625	100	93.416	4	8
5656.25	100	93.416	3	9
5718.75	100	93.416	3	10
5781.25	100	93.416	3	10
5843.75	100	93.416	3	10
5906.25	100	93.416	3	10
6000	100	92.372	3	10
6062.5	100	92.372	3	10
6125	100	92.372	3	11
6187.5	100	92.372	3	11
6281.25	100	92.372	3	12
6343.75	100	92.372	3	12
6406.25	100	92.372	3	13
6468.75	100	92.372	3	13
6562.5	100	92.372	2	13
6656.25	100	92.372	2	13
6718.75	100	92.372	2	13
6750	100	92.372	2	14
6843.75	100	92.372	2	14
6906.25	71.37254902	92.372	2	14
6781.25	12.94117647	92.372	0	38


Keep in mind a VERY important point

Stock tuned cars normally do KNOCK and pull timing - this is 100% normal

The manner in which the ecu is designed is that the ecu will adjust itself to macth the fuel quality that is being used.

At some point when your octane rating is high enough you will just run 100% off the highest timing map with zero knock correction - assuming there is no harmonic or noise which is giving a flase knock indication

Back when I first experimented with Evo 8's in Californoia on 91 octane I was shocked to see that they knocked and pulled timing more with the stock tune making less power than they did with the Dyno Flash tunes.

By carefully working with fuel and timing it is possible to make decent power while keeping percieved knock at the ecu to a minimum and getting a very smooth timing curve
Reply
Old Jul 16, 2006 | 11:02 AM
  #35  
RazorLab's Avatar
EvoM Guru
20 Year Member
Liked
Loved
Community Favorite
iTrader: (8)
 
Joined: Aug 2003
Posts: 14,094
Likes: 1,092
From: Mid-Hudson, NY
Originally Posted by nj1266
If you look at the first log, the pulled timing coincides with peak torque. This could be the results of two things. It could be that there is knock and the ECU is pulling timing. But it could also be that the ECU pulls timing by itself at peak torque. Below is a log from my EVO w/o any tune and engine management. Notice how the ECU pulls timing at peak torque.

RPM, Timing, O2
3125.0, 20.0, 0.88
3273.0, 13.0, 0.92
3438.0, 7.0, 0.92
3660.0, 4.0, 0.94
3938.0, 3.0, 0.94
4180.0, 7.0, 0.94
4430.0, 7.0, 0.94
4660.0, 7.0, 0.94
4898.0, 8.0, 0.94
5133.0, 9.0, 0.94
5355.0, 8.0, 0.96
5586.0, 9.0, 0.96
5813.0, 9.0, 0.96
6035.0, 11.0, 0.96
6270.0, 13.0, 0.96
6465.0, 15.0, 0.96
6660.0, 17.0, 0.96
6848.0, 19.0, 0.98
7055.0, 19.0, 0.98
7211.0, 21.0, 0.98
7391.0, 21.0, 0.98
That actually looks like the factory timing curve at peak torque in the 240 load column. The car is moving into 200,220 and then peak trq 240 load column at 3500-4k (3degrees) then dropping into the 220,200 and then 180 load columns on up.
Reply
Old Jul 16, 2006 | 11:10 AM
  #36  
nj1266's Avatar
Evolved Member
iTrader: (6)
 
Joined: Nov 2004
Posts: 3,254
Likes: 13
From: USA
Originally Posted by DynoFlash
I concur with your observations 100%

The way I like to obeserve knock is with either my ears or by watching what I scan directly off the knock sensor

The problem of course is the IX ecu is very senstive to any sounds in the engine it percieves as "abnormal" whioch can simply mean added engine noise over stock

Its a balacing act to set up the ecu to make decent power levels while keeping it knock free - - AND - - making the ecu happy with the level of engine noise and not pulling any timing

I have seen cars reporting "KNOCK COUNTS" which are not knocking at all

I have seen cars loudly knock that show zero knock count

and I have seen cars with no knock and no knock count pull timing like mad

As tuners we must balance the BENEFITS of the stock knock control system vs. the minor inconvienience of having timing being pulled on occassion

I would prefer to keep the stock knock control system 100% active and protect the engine as much as posisble than risk engine damage by trying to alter the knock threshold buffers

The main thing to keep in mind is that IX evos are very sensitive and the ecu acts very quickly in regard to any knock sounds the knock sensor picks up
Very well said Al. I would not touch my knock sensor settings either. There are several steps that people can take to reduce the possibility of knock:

1. Run the highest octane gas that you can afford and that is available. I am currently mixing 91 octane and 100 octane to get 93 octane. The AFR has never looked better. I used to get spikes and dips in my AFR, but now I hardly see them. I will continue to mix gas to get 93 octane. After all the EVO was designed to run on 93 octane. My gas bill has increased $10 per tank, but it is well worth it.

2. Set your AFR in the low 11s. That will really help in the hot temps. Constantly log your AFR. Once or twice a week should do it. I do it once on my way to work in the early AM hours. It is cool in the AM hours here in CA. I also log with an Innovate unit w/o a laptop and that makes it much easier to log.

3. Log your timing along side the AFR and see if the car is pulling serious timing. IMO, 1 degree here and there is not a big deal. When the ECU pulls 3-4 degrees unexpectedly, then you have a knock event. There should be a dip in your AFR as well.

4. Do not run high boost on 93 octane gas. IMO, 22 psi is the max that you should run. If you want to run higher boost, then get some 100 octane gas and have fun boosting.
Reply
Old Jul 16, 2006 | 11:15 AM
  #37  
elhalisf's Avatar
Evolved Member
iTrader: (171)
 
Joined: Apr 2005
Posts: 1,216
Likes: 0
From: Chandler, AZ
Originally Posted by nj1266
What you are saying is for a DSM. Does it also apply for the EVO? Here is the quote from pocket logger web site DSM section:

Knock:

Range: 0 to 43
The knock sum scale is given to represent the signal from the knock sensor to the ECU. The ECU then calculates the amount of detonation present. The higher the knock sum the more the ECU retards timing to get rid of the knock.
With a knock sum of:
3 or below, the ECU will advance timing.
Between 3 and 7, the ECU leaves timing alone.
Higher than 7, the ECU will begin retarding timing. Higher knock = higher retard, with the possiblity of seeing less than 4deg total advance at the maximum knock sum of 43.
One important note to remember is the knock sensor can detect "false" knock and retard timing. This is most commonly caused from the noisy lifters but anything it can audibly pick up could cause this.
yes i know that. but like i said, it is unconfirmed. i have also seen VERY similar response in my ecu, so i believe that the dsm calculations are similar to the evo knock calculations
Reply
Old Jul 16, 2006 | 11:19 AM
  #38  
DynoFlash's Avatar
Account Disabled
iTrader: (91)
 
Joined: Sep 2003
Posts: 16,850
Likes: 0
From: 2003 Evo VIII - Silver
The funny thing about all of this conversation of course gentleman is that up until very recently all the dozens of super AFC gurus with cars hitting knock counts over 10 and pulling 5 - 10 degress of timing all the time
Reply
Old Jul 16, 2006 | 11:20 AM
  #39  
RazorLab's Avatar
EvoM Guru
20 Year Member
Liked
Loved
Community Favorite
iTrader: (8)
 
Joined: Aug 2003
Posts: 14,094
Likes: 1,092
From: Mid-Hudson, NY
Whoever is using an SAFC to tune and has 5-10 degrees of timing being pulled, isn't tuning it very well. Mine never pulled that much when I used one. Even on 91 octane.
Reply
Old Jul 16, 2006 | 11:26 AM
  #40  
dudical26's Avatar
Evolved Member
iTrader: (17)
 
Joined: Nov 2005
Posts: 2,544
Likes: 0
From: NNJ
I have a utec Tuner which has Wide band and it also has a knock sensor that you can log seperatly and also you can listen to it directly. When I get a chance I will hook this up and listen for any real knock.
Reply
Old Jul 16, 2006 | 11:47 AM
  #41  
5k6's Avatar
5k6
Thread Starter
Evolving Member
15 Year Member
iTrader: (19)
 
Joined: Apr 2006
Posts: 147
Likes: 0
From: IL
Ok here is a run I did today in the heat haha (its like 100 out) my boost is now at 20psi and I did not change it since last night (21psi)? I am still waiting for my gas to go before I refill up with a different brand of gas. Also my cont-g's were like 4.5 last night and they are 3.4 today.

Last edited by 5k6; Jul 16, 2006 at 11:49 AM.
Reply
Old Jul 16, 2006 | 12:05 PM
  #42  
nj1266's Avatar
Evolved Member
iTrader: (6)
 
Joined: Nov 2004
Posts: 3,254
Likes: 13
From: USA
Originally Posted by 5k6
Ok here is a run I did today in the heat haha (its like 100 out) my boost is now at 20psi and I did not change it since last night (21psi)? I am still waiting for my gas to go before I refill up with a different brand of gas. Also my cont-g's were like 4.5 last night and they are 3.4 today.
The reason why your boost is lower is the heat and less dense air that you get with heat. At night your boost goes up by 1-1.5 psi assuming it is cooler.

Your log looks good to me. you have 1 degree pull @ 4300, but then the ECU adds this 1 degree even though the knock sum stayed at 3. Go figure.

I would also like to see less timing at peak boost which is around 3800 rpm. But this is an EVO IX and I believe it can handle more advanced timing at peak boost.
Reply
Old Jul 16, 2006 | 12:36 PM
  #43  
4TUN8's Avatar
Evolved Member
iTrader: (18)
 
Joined: Jan 2005
Posts: 2,458
Likes: 1
I've seen in my own logs timing being pulled on 1 count of knock, numerous times.
Reply
Old Jul 16, 2006 | 12:51 PM
  #44  
djmikeymike's Avatar
Evolved Member
iTrader: (1)
 
Joined: Aug 2004
Posts: 935
Likes: 0
From: eastern pa
^^^ me too
Reply
Old Jul 16, 2006 | 01:08 PM
  #45  
RazorLab's Avatar
EvoM Guru
20 Year Member
Liked
Loved
Community Favorite
iTrader: (8)
 
Joined: Aug 2003
Posts: 14,094
Likes: 1,092
From: Mid-Hudson, NY
Honestly, I'm not sure how much faith I have in the EvoScan knock reading. I've seen it showing 10 counts of knock and no timing being pulled. Knock voltage from ECU+ log showing no big problem.
Reply



All times are GMT -7. The time now is 11:40 AM.