knock counts and water injection
10-27-2007, 10:14 AM
#1
Instructor
Thread Starter
Join Date: Oct 2004
Location: Oxfordshire, UK
Posts: 240
Likes: 0
Received 0 Likes
on
0 Posts
knock counts and water injection
I had posted this in answer to another topic, and then thought perhaps it could do with its own.
I am currently injecting water only at the exit of the intercooler, and using the curtis counter i have been logging the knock on pin 15 of the klr.
I have a 0.9 mm jet that I have measured as flowing 300ml/cc in one minute, and use the smt6 to control a purpose made inector solenoid made by ERL. flow starts at about 3200 rpm in all cases
flowing no water i get:
10psi = 0 knock events
13psi = 1 knock events
16psi = 10 knock events
19psi = 17 knock events
flowing about 120ml per minute i get:
13psi = 0 knock events
16psi = 6 knock events
19psi = 9 knock events
flowing about 170ml per minute i get:
13psi = 1 knock events
16psi = 6 knock events
19psi = 10 knock events
flowing about 220ml per minute i get:
13psi = 1 knock events
16psi = 6 knock events
19psi = 9 knock events
all done in third gear, EGT is always around 1500F with water (at 13, 16 and 19 psi) and on euro 99 octane, with is the same as your US 93 octane
It clearly does show a decrease in knock, when I have more time i will try with more water. But at some point there must be max amount you can inject before EGT goes down, along with power.
Does anyone else have any figures I can compare?
With or without water injection, there are a number of other people using the curtis counter and i would be interested in what numbers they are getting at various boost levels.
Thanks in advance.
edited to add: engine is stock 2.5, running a vitesse maf, to4e 50 trim turbo with 8 hotside and EBC
I am currently injecting water only at the exit of the intercooler, and using the curtis counter i have been logging the knock on pin 15 of the klr.
I have a 0.9 mm jet that I have measured as flowing 300ml/cc in one minute, and use the smt6 to control a purpose made inector solenoid made by ERL. flow starts at about 3200 rpm in all cases
flowing no water i get:
10psi = 0 knock events
13psi = 1 knock events
16psi = 10 knock events
19psi = 17 knock events
flowing about 120ml per minute i get:
13psi = 0 knock events
16psi = 6 knock events
19psi = 9 knock events
flowing about 170ml per minute i get:
13psi = 1 knock events
16psi = 6 knock events
19psi = 10 knock events
flowing about 220ml per minute i get:
13psi = 1 knock events
16psi = 6 knock events
19psi = 9 knock events
all done in third gear, EGT is always around 1500F with water (at 13, 16 and 19 psi) and on euro 99 octane, with is the same as your US 93 octane
It clearly does show a decrease in knock, when I have more time i will try with more water. But at some point there must be max amount you can inject before EGT goes down, along with power.
Does anyone else have any figures I can compare?
With or without water injection, there are a number of other people using the curtis counter and i would be interested in what numbers they are getting at various boost levels.
Thanks in advance.
edited to add: engine is stock 2.5, running a vitesse maf, to4e 50 trim turbo with 8 hotside and EBC
Last edited by ross255; 10-27-2007 at 11:20 AM.
10-27-2007, 01:58 PM
#2
Rennlist Member
Judging by your information, I'd be questioning the integrity of the knock counter.
At 170ml/min, that should be plenty of water to totally suppress knock at those boost levels.
Also, shouldn't the stock knock sensor cut back timing, if there was knock?
The other thing is; how do you know that the smt6/solenoid combo is actually flowing those lesser amounts compared to the full 300ml/min? Did you flow test those amounts and also observe the atomization coming out of the injector. Sometimes the pulsing of the solenoid reduces the atomization of the water, dramatically reducing its effectiveness.
At 170ml/min, that should be plenty of water to totally suppress knock at those boost levels.
Also, shouldn't the stock knock sensor cut back timing, if there was knock?
The other thing is; how do you know that the smt6/solenoid combo is actually flowing those lesser amounts compared to the full 300ml/min? Did you flow test those amounts and also observe the atomization coming out of the injector. Sometimes the pulsing of the solenoid reduces the atomization of the water, dramatically reducing its effectiveness.
10-27-2007, 02:06 PM
#3
Rennlist Member
Ross I asked a question in the other thread that you posted those numbers on, but I'll ask here too. What is the scale of those knock numbers? Is it a universal scale and therefore what are the numbers out of. 0-20, 0-500 etc ?? So what is good, bad, ugly here? So this is just taking readings from the stock KLR?
10-27-2007, 04:21 PM
#4
Instructor
Thread Starter
Join Date: Oct 2004
Location: Oxfordshire, UK
Posts: 240
Likes: 0
Received 0 Likes
on
0 Posts
turbotommy, the counter only counts outputs from the klr, so I can only tell you how many signals the klr is sending out, to the best i can possibly determine they are accurate.
And yes the KLR would cut back timing if there was too much knock. As of yet I have not found out what this level is, however there was a picture in this post
https://rennlist.com/forums/showpost...4&postcount=43
and the purple lines from 0 to 5 are the signals from the klr, at this level it cut back timing, problem is i cant remember the time scale for how long the sample was(it was a full throttle run in one gear if i remember right), however you can still see that there is a lot of events happening there.
For my amount of knocks it does not appear to be enough pull timing at all.
Yes the solenoid still flows the levels indicated, I know what you mean about the pulsing slowing down flow, but it is designed for this and works as fast as a fuel injector, ts rated at up to 250hz.
I also would have thought 170ml would have been enough before going to 220ml! I will have to try 300ml. If this has no effect then I can only thinkit is down to poor cylinder distribution, and will have to go port injection.
333pg333, they are the number of signals sent from pin 15 (the knock ye/no output) of the KLR, so its what the klr has decided is knock.
I am hoping others will chime in with numbers so I can compare.
And yes the KLR would cut back timing if there was too much knock. As of yet I have not found out what this level is, however there was a picture in this post
https://rennlist.com/forums/showpost...4&postcount=43
and the purple lines from 0 to 5 are the signals from the klr, at this level it cut back timing, problem is i cant remember the time scale for how long the sample was(it was a full throttle run in one gear if i remember right), however you can still see that there is a lot of events happening there.
For my amount of knocks it does not appear to be enough pull timing at all.
Yes the solenoid still flows the levels indicated, I know what you mean about the pulsing slowing down flow, but it is designed for this and works as fast as a fuel injector, ts rated at up to 250hz.
I also would have thought 170ml would have been enough before going to 220ml! I will have to try 300ml. If this has no effect then I can only thinkit is down to poor cylinder distribution, and will have to go port injection.
333pg333, they are the number of signals sent from pin 15 (the knock ye/no output) of the KLR, so its what the klr has decided is knock.
I am hoping others will chime in with numbers so I can compare.
Last edited by ross255; 10-27-2007 at 07:49 PM.
10-27-2007, 05:22 PM
#5
Instructor
Join Date: Oct 2005
Location: Boynton Beach, Florida
Posts: 140
Likes: 0
Received 0 Likes
on
0 Posts
I had posted this in answer to another topic, and then thought perhaps it could do with its own.
I am currently injecting water only at the exit of the intercooler, and using the curtis counter i have been logging the knock on pin 15 of the of the klr.
I have a 0.9 mm jet that I have measured as flowing 300ml/cc in one minute, and use the smt6 to control a purpose made inector solenoid made by ERL. flow starts at about 3200 rpm in all cases
I am currently injecting water only at the exit of the intercooler, and using the curtis counter i have been logging the knock on pin 15 of the of the klr.
I have a 0.9 mm jet that I have measured as flowing 300ml/cc in one minute, and use the smt6 to control a purpose made inector solenoid made by ERL. flow starts at about 3200 rpm in all cases
10-27-2007, 06:03 PM
#7
turbotommy, the counter only counts outputs from the klr, so I can only tell you how many signals the klr is sending out, to the best i can possibly determine they are accurate.
And yes the KLR would cut back timing if there was too much knock. As of yet I have not found out what this level is, however there was a picture in this post
https://rennlist.com/forums/showpost...4&postcount=43
and the purple lines from 0 to 5 are the signals from the klr, at this level it cut back timing, problem is i cant remember the time scale for how long the sample was, however you can still see that there is a lot of events happening there.
For my amount of knocks it does not appear to be enough pull timing at all.
Yes the solenoid still flows the levels indicated, I know what you mean about the pulsing slowing down flow, but it is designed for this and works as fast as a fuel injector, ts rated at up to 250hz.
I also would have thought 170ml would have been enough before going to 220ml! I will have to try 300ml. If this has no effect then I can only thinkit is down to poor cylinder distribution, and will have to go port injection.
333pg333, they are the number of signals sent from pin 15 (the knock ye/no output) of the KLR, so its what the klr has decided is knock.
I am hoping others will chime in with numbers so I can compare.
And yes the KLR would cut back timing if there was too much knock. As of yet I have not found out what this level is, however there was a picture in this post
https://rennlist.com/forums/showpost...4&postcount=43
and the purple lines from 0 to 5 are the signals from the klr, at this level it cut back timing, problem is i cant remember the time scale for how long the sample was, however you can still see that there is a lot of events happening there.
For my amount of knocks it does not appear to be enough pull timing at all.
Yes the solenoid still flows the levels indicated, I know what you mean about the pulsing slowing down flow, but it is designed for this and works as fast as a fuel injector, ts rated at up to 250hz.
I also would have thought 170ml would have been enough before going to 220ml! I will have to try 300ml. If this has no effect then I can only thinkit is down to poor cylinder distribution, and will have to go port injection.
333pg333, they are the number of signals sent from pin 15 (the knock ye/no output) of the KLR, so its what the klr has decided is knock.
I am hoping others will chime in with numbers so I can compare.
Especially #4.
Trending Topics
10-27-2007, 09:02 PM
#8
Rennlist Member
Yes I realize that you are running off the KLR but you still need to know what the scale and time is for those numbers don't you? In other words what you have may be fine or it could be just short of catastrophic but how would you or anyone know? There has to be some sort of reference. Where you had no water and 17 counts at 19psi, how do you know this is actually bad? From what I read we do get some knock happening but not sure what's in the acceptable range.
Sorry to ask apparently newbie questions but it does seem kind of important.
Sorry to ask apparently newbie questions but it does seem kind of important.
10-27-2007, 09:41 PM
#10
Rennlist Member
Scott, do you know the scale and count relevance? You'd have to assume it's at WOT to get the psi levels he's discussing.
10-27-2007, 10:03 PM
#11
Patrick, I think his counter is a digital counter that just counts the knock events from KLR 15 so it doesn't have a time scale. Ross said all doen in 3rd gear so I'm assuming in each case he got that number of knocks for one run to red line but I guess he can confirm that.
As for relevance, I'd rather see no knock! If you look at Transaxle's post with the graphs, he went from heavy knock to zero knock when using a higher octane fuel.
I'm sure Ross's water injection would be more effective at reducing knock if it were a water/methanol mix, but that has its own issues.
As for relevance, I'd rather see no knock! If you look at Transaxle's post with the graphs, he went from heavy knock to zero knock when using a higher octane fuel.
I'm sure Ross's water injection would be more effective at reducing knock if it were a water/methanol mix, but that has its own issues.
10-27-2007, 11:25 PM
#12
Rennlist Member
Yes you can sort of go around in circles in tuning these cars. Octane, psi, c/r, spark, head design etc then introduce a spray system which can involve it's own issues. Ahh the beauty of performance tuning especially turbo cars...
10-27-2007, 11:27 PM
#13
Rennlist Member
my knock counter was picking up more knocks at takeoff from a dead stop than at full boost- and i dont think i got more then one or two then. I was up to about 500 in a month and almost all -20 at a time or so- was from taking off from a stop.
10-27-2007, 11:35 PM
#14
Addict
Rennlist Member
Rennlist Member
Join Date: Feb 2002
Posts: 5,384
Likes: 0
Received 0 Likes
on
0 Posts
Lets take into account that the knock count is the KLR attempt to detect knock from background noise. I would like to see a chart of knock count vs timing to determine when the KLR acts on the knock count.
10-28-2007, 03:33 AM
#15
Rennlist Member
Also a good point that Geneqco raised to me was, if the KLR is designed to pick up whatever noises the factory set, what happens when we change things in the car? Bore size, pistons, rods, crank mods, capacity, sleeves, rod bolts, headgaskets the list goes on. If we change some, any, or all of these components isn't this changing the sonics in the engine and therefore what is the KLR now picking up or indeed missing?
This seems like a good question to me.
This seems like a good question to me.