Knock question

[Bmxr152]

Ok ive been having some problems with knock for a while now. I was only able to boost about 10psi before alcohol injection without knock, and no with the alcohol injection i can boost about 15 or so psi. I was wondering what could be causing this knock, mabie a bad tune? Well if the knock goes away with race gas or alky injection, would it be my fuel system? I do have a new knock sensor too.

 

[JayRolla]

Ok ive been having some problems with knock for a while now. I was only able to boost about 10psi before alcohol injection without knock, and no with the alcohol injection i can boost about 15 or so psi. I was wondering what could be causing this knock, mabie a bad tune? Well if the knock goes away with race gas or alky injection, would it be my fuel system? I do have a new knock sensor too.

I looked at your profile and it does not mention a fuel pump. If you dont have one that is your problem most likely.

 

[AWDfrank]

I'm having the same problem. 10 psi and I get 6 or 7 counts of knock and at 6500 rpm it is 15-18. Does it make sense that at night I get less knock, seems right to me because of the temperature. Disregard my mods, all I have right now is a downpipe, all of that go's in this weekend Not rying to hijack the thread but we have the same problem.

 

[Bmxr152]

I do have a fuel pump sorry...updating

 

[Bmxr152]

I also noticed that i have alot of partial throttle knock. Since i have the dsmlink it show on my stock boost guage counts of knock at crusing speed at partial throttle the knock will shoot up and as soon as i let my foot out of it and put it back in it goes away. I know its most likely phantom knock, but could that phantom knock also be casuing my knock at WOT at only 15psi?

 

[GST with PSI]

I have heard of much higher boost numbers with minimal mods and water injection systems. I am guessing that you mean a water and alcohol injection system when you say alcohol injection.

I'm sure you know, but your target water to fuel ratio may be anywhere from 10%-25% water to fuel. The hardest part of water injection is finding what percentage is right for your application. In order to do that, you have to find how much water you need to eliminate knock at your given rpm and boost pressure. This required percentage will also change with engine conditions and likely never be a constant. After you determine your percentage you're in the clear. I would suggest concentrating at WOT and tuning your injection system to prevent knock safely at that point. Without a variable controller that creates a map to match you fuel curve, you will never be able to optimize water injection rates to consistently minimize knock throughout the RPM range. In the lower RPM ranges many people use a 2 stage system to minimize knock at partial throttle, and concentrate on tuning their water injection systems to eliminate knock at WOT with a second water injector. Keep in mind timing advance will also have an effect on knock even when coupled with water injection.

With a water injection system, you can run very high boost with almost no mods. Start tuning your injection system to make some power by preventing knock. Keep in mind when you ECU sees knock it retards timing and power loss occurs. To prevent this you need to prevent knock. I would not replace your knock sensor, nor would I worry about knock that isn't derived under a load such as at partial throttle cruising or deceleration. Furthermore, your stock fuel pump accompanied by a rewire will easily deliver enough to fuel to support 15 psi. The goal here is not to deliver more fuel, but to allow what is delivered to be burned in a controlled manor without detonation.

see these websites to help determine your water to fuel mix:

http://www.enginerunup.com/shop.php...witch-stage-2-kit-1-30-psi-of-boost-/p_7.html

http://www.rbracing-rsr.com/waterinjection.html

hope this helps....

 

[kenamond]

I have heard of much higher boost numbers with minimal mods and water injection systems. I am guessing that you mean a water and alcohol injection system when you say alcohol injection.

I'm sure you know, but your target water to fuel ratio may be anywhere from 10%-25% water to fuel. The hardest part of water injection is finding what percentage is right for your application. In order to do that, you have to find how much water you need to eliminate knock at your given rpm and boost pressure. This required percentage will also change with engine conditions and likely never be a constant. After you determine your percentage you're in the clear. I would suggest concentrating at WOT and tuning your injection system to prevent knock safely at that point. Without a variable controller that creates a map to match you fuel curve, you will never be able to optimize water injection rates to consistently minimize knock throughout the RPM range. In the lower RPM ranges many people use a 2 stage system to minimize knock at partial throttle, and concentrate on tuning their water injection systems to eliminate knock at WOT with a second water injector. Keep in mind timing advance will also have an effect on knock even when coupled with water injection.

With a water injection system, you can run very high boost with almost no mods. Start tuning your injection system to make some power by preventing knock. Keep in mind when you ECU sees knock it retards timing and power loss occurs. To prevent this you need to prevent knock. I would not replace your knock sensor, nor would I worry about knock that isn't derived under a load such as at partial throttle cruising or deceleration. Furthermore, your stock fuel pump accompanied by a rewire will easily deliver enough to fuel to support 15 psi. The goal here is not to deliver more fuel, but to allow what is delivered to be burned in a controlled manor without detonation.

see these websites to help determine your water to fuel mix:

http://www.enginerunup.com/shop.php...witch-stage-2-kit-1-30-psi-of-boost-/p_7.html

http://www.rbracing-rsr.com/waterinjection.html

hope this helps....

He's knocking at 10psi. You shouldn't need water/meth injection at 10psi to eliminate knock. I have the same problem. I can't monitor knock count, but I'm seeing timing curves that indicate knock at 9psi through 15psi (can't go lower due to WGA, can't go higher due to stock FP). I'm wondering if it could be lifter tick registering as knock. I wanna do 3g lifters, but that's on a long list of things with $ next to them.

Do you other guys with this problem (Bmxr152 and AWDfrank) have lifter tick?

 

[GST with PSI]

He's knocking at 10psi. You shouldn't need water/meth injection at 10psi to eliminate knock. I have the same problem. I can't monitor knock count, but I'm seeing timing curves that indicate knock at 9psi through 15psi (can't go lower due to WGA, can't go higher due to stock FP). I'm wondering if it could be lifter tick registering as knock. I wanna do 3g lifters, but that's on a long list of things with $ next to them.

Do you other guys with this problem (Bmxr152 and AWDfrank) have lifter tick?

What do you mean. It is not unheard of nor is it uncommon to get knock at 10 PSI. 15-18 counts is a bit more than I would expect but does not surprise me, and I don't understand why it does you. The car obviously can not pull enough timing to prevent the knock counts occurring, which means to me some outside factor must be introduced: ie higher octane fuel, water injection, foot off the throttle etc. A lean condition due to lack of fuel at 10 PSI seams unlikely with a properly working problem free fuel system. There isn't a rule that states when a detonation deterrent is needed, it depends on application. It doesn't make any sense to say at a certain PSI you shouldn't need water injection and at a certain PSI you would. A million factors can influence when detonation occurs and needs to be prevented.

With a water injection kit already installed I was simply suggesting a more optimal tune to reduce knock. Hell, for all I know the tune could already be optimal and be a separate problem all together. Either way something needs to be done to prevent knock weather it be water injection, timing or fuel related.

As for the phantom knock I doubt it would be caused by lifter tick. If a lifter were making enough noise to produce phantom knock, it would be constant linear curve. As RPM increases so would the frequency of the sound therefore producing knock counts which grow higher with engine speed. If this were the case you would be able to reproduce this symptom in say for instance neutral, by increase the engine speed until knock occurs; as load on the engine wouldn't effect the lifter tick, and I would therefore test the vehicle stationary rather than driving in order to rule out actual knock.

 

[kenamond]

As for the phantom knock I doubt it would be caused by lifter tick. If a lifter were making enough noise to produce phantom knock, it would be constant linear curve. As RPM increases so would the frequency of the sound therefore producing knock counts which grow higher with engine speed. If this were the case you would be able to reproduce this symptom in say for instance neutral, by increase the engine speed until knock occurs; as load on the engine wouldn't effect the lifter tick, and I would therefore test the vehicle stationary rather than driving in order to rule out actual knock.

Kinda hard when you don't have the ability to monitor knock count (I have a stock 2g ECU). If you could tell me what a "neutral pull" timing curve should look like and if I could get more than 10 samples during this "neutral pull", then maybe your speculation about tick-induced phantom knock would be useful, but atleast one of those aren't going to happen. I see timing curve flat spots and dips in 3rd gear pulls at 9psi. Nitromethane octane booster didn't affect it at all. Knock sensor is not oozing goo. I posted on this, and everything that was suggested didn't fix the problem.

And lifter tick would be louder at higher rpm, because the metal parts are smacking into each other at higher velocity as rpm increases. It wouldn't be linear. Maybe you're speculating, but I'm looking for facts and real world experience.

Do you think that a stock 2g setup in perfect condition should knock at 9psi boost on a 70ºF day with 95ºF intake temps? I'm not saying that mine or the other two guys' cars are in perfect condition, but do you think that would be normal? I don't. The norm for folks on this forum is no knock at 9psi. I was just trying to see if there was something common between us that might shed light on my problem.

If you can propose ideas why I'd be knocking at 9psi with nitromethane octane booster and rewired pump (like it matters, I was knocking at 4k rpm), then I'd appreciate it.

 

[AWDfrank]

^^Lifter tick would be harder to hear at WOT even though it is louder, it is not more audible. It will be loudest at idle.

 

[oldman]

Possibly boost leaks -> over spining turbo -> higher IAT -> heatsoak.

 

[Bmxr152]

Does anyone know a link for the best DIY boost leak tester?

 

[gsxtasy]

http://www.vfaq.com/mods/ICtester.html

 

[GST with PSI]

Kinda hard when you don't have the ability to monitor knock count (I have a stock 2g ECU). If you could tell me what a "neutral pull" timing curve should look like and if I could get more than 10 samples during this "neutral pull", then maybe your speculation about tick-induced phantom knock would be useful, but atleast one of those aren't going to happen. I see timing curve flat spots and dips in 3rd gear pulls at 9psi. Nitromethane octane booster didn't affect it at all. Knock sensor is not oozing goo. I posted on this, and everything that was suggested didn't fix the problem.

And lifter tick would be louder at higher rpm, because the metal parts are smacking into each other at higher velocity as rpm increases. It wouldn't be linear. Maybe you're speculating, but I'm looking for facts and real world experience.

Do you think that a stock 2g setup in perfect condition should knock at 9psi boost on a 70ºF day with 95ºF intake temps? I'm not saying that mine or the other two guys' cars are in perfect condition, but do you think that would be normal? I don't. The norm for folks on this forum is no knock at 9psi. I was just trying to see if there was something common between us that might shed light on my problem.

If you can propose ideas why I'd be knocking at 9psi with nitromethane octane booster and rewired pump (like it matters, I was knocking at 4k rpm), then I'd appreciate it.

I think you are misunderstanding. Nobody said anything about a neutral pull? In layman's terms rev the car up in neutral and see if you get knock, or in your case see if the ECU pulls timing. Even without the car under a load, a lifter tick loud enough to simulate knock may show up and pull timing in neutral (that means out of gear sitting still). Even if didn't pull timing, you would still be able to audibly hear the tick and determine using some common sense weather or not it is loud enough to cause knock (meaning it would have to be very abnormally loud). I can't think of a simpler test to run to find your "phantom lifter tick knock".

I also believe I said in the last post:
If a lifter were making enough noise to produce phantom knock, it would be constant linear curve. As RPM increases so would the frequency of the sound therefore producing knock counts which grow higher with engine speed.

And you said:
And lifter tick would be louder at higher rpm, because the metal parts are smacking into each other at higher velocity as rpm increases. It wouldn't be linear.

I not sure who you are disagreeing with but it's not me. I already said lifter tick would get louder with RPM? Also just to clarify linear and a linear curve are not the same thing. I stated the lifter tick would look like a linear curve, aka get louder with engine speed and be directly proportional.....

^^Lifter tick would be harder to hear at WOT even though it is louder, it is not more audible. It will be loudest at idle.
AWDfrank I agree...

And yes I do think it is possible that a stock 2g setup in perfect condition would knock at 9psi boost on a 70ºF day with 95ºF intake temps. My 1g will knock at 10 psi under the same conditions. Keep in mind also I'm not running a T25 or a smaller side mount intercooler (both of which a stock 2g is equip with compared to the stock 1g), that will significantly raise intake temperatures at the manifold. Now would 15 counts of knock be likely, I think not. Also just to clarify I'm talking about knock at or near WOT under a fairly heavy load, not partial throttle. Knock at partial throttle is a whole different story, and one I would worry about considerably less if at all.

I'd also be happy to give you some suggestions why you are knocking at 9psi with nitromethane octane booster and rewired pump at 4k rpm. Post some information that could narrow down the problem, perhaps a log, and possibly myself or someone who knows far more than I could help.

because you can't see your knock this scale could be used to make an educated guess...
• 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 possibility of seeing less than 4deg total advance at the maximum knock sum of 43.

more questions with that read here: http://www.3si.org/plhelp/index.html

at 9 counts of knock the ECU is just getting off it's ass. Don't know why your so worried. A car pulling 8-10 degrees timing I would be worried about. 3 degrees, not so much. I'm not sure what your case is but if it's in the first category you have a serious problem at only 9 PSI. Post a log.

 

[Bmxr152]

I was also notiing when after i turn the car off and let it sit for a couple hours and come back the fuel pressure drops all the way to zero, but when i start it back up it goes back to normal fuel pessure. I just installed the walbro 255. Is this normal?

 

[xveganxcowboyx]

Do you other guys with this problem (Bmxr152 and AWDfrank) have lifter tick?

I have this problem as well, but have little to no problems with lifter tick (only a little bit with regular oil, which I rarely use). I'm not sure where the problem lies, but I'm convinced it's a phantom knock issue VS real knock.

 

[GST with PSI]

On some vehicles the fuel system does hold some fuel pressure, however others do not it just depends on the manufacturer. Fuel pressure held for a length of time with the vehicle off is refereed to as residual fuel pressure. Rarely is there a specification for residual fuel pressure. To my knowledge there isn't one for DSM's. Because our cars don't hold residual fuel pressure, Mitsubishi solves this problem by activating to fuel pump on initial key up. Cycle the key to the run. Don't start the car. When the key is cycled to the run position fuel pump will kick on for about 3 seconds or so. This builds initial fuel pressure before the engine is actually started. Try this procedure to see if your fuel pressure comes up a bit before the car is started.


.........

 

[oldman]

I was also notiing when after i turn the car off and let it sit for a couple hours and come back the fuel pressure drops all the way to zero, but when i start it back up it goes back to normal fuel pessure. I just installed the walbro 255. Is this normal?

As long as it doesn't drop to zero immediately, it is normal with the Aremotive AFPR. If it drops to zero immediately, go back and check your fuel pump installation and make sure the o-ring isn't twisted or ripped. More importantly, make sure you taflon both the vacuum fitting and adjustment rod on the AFPR.

On some vehicles the fuel system does hold some fuel pressure, however others do not it just depends on the manufacturer. Fuel pressure held for a length of time with the vehicle off is refereed to as residual fuel pressure. Rarely is there a specification for residual fuel pressure. To my knowledge there isn't one for DSM's. Because our cars don't hold residual fuel pressure, Mitsubishi solves this problem by activating to fuel pump on initial key up. Cycle the key to the run. Don't start the car. When the key is cycled to the run position fuel pump will kick on for about 3 seconds or so. This builds initial fuel pressure before the engine is actually started. Try this procedure to see if your fuel pressure comes up a bit before the car is started.

The stock FPR does in fact hold pressure, it is the AFPR that causes pressure drop. Also, if you're refering to "on" when you say "run" position, our fuel pump doesn't activate until you crank the car.

 

[kenamond]

I can't monitor knock count. I have a 2g stock ECU. I have to rely on a timing curve and infer knock from timing flatspots and dips after boost hits. Reving the engine in neutral won't tell me anything, because I'll get about 5 samples before redline, and I have no idea what the timing curve should look like even if I could get enough samples.

Below is the thread I started long ago about this problem.

http://www.dsmtuners.com/forums/showthread.php?t=235111

Bruce, I had no boost leaks at the time.

Compression at the time was high if anything, and seafoam didn't drop it. Could still be hot spots from carbon buildup, but I figure if I had a good bit of deposits, the compression should've changed after the seafoam treament.

My lifter tick wasn't very bad at the time; it would tick a bit until it warmed up and/or oil pressure went up.

Rewiring the FP didn't help, but my O2v looked fine, so I don't think that had anything to do with anything. My 14psi pulls leaned out past 6k rpm, so that gave me some confidence that the O2 sensor was working.

One of the wisement (steve?) was a bit perplexed at my advance. It seemed way too high to him. My best explanation is that I'm at about 8000ft elevation for these pulls, so the mass airflow is about 80% at the same boost and rpm as someone at sea level. This would put you at a more advanced location on the timing map. If I had the raw timing map data, I could see if shifting the airflow up 20% would put me at a more typical timing advance. But I don't have that data.

NoX nitromethane octane booster didn't help which I found interesting. If I was close to knocking for real, half a bottle of NoX in half a tank should have helped, but it didn't seem to affect my timing curves.

I haven't logged any pulls since that thread. The O2 sensor has been replaced since, but my fuel economy (or lack thereof) over the past couple weeks makes me think I need another boost leak test. But at the time of my investigation, I had no boost leaks. It held 20psi on a warm motor and would leak down 1psi every 3+ seconds.

I'll probably do a couple of pulls in the next week or so to see what the new O2 spits out so I can compare to older logs. The weather is about the same now (70ºF), so it would be a fair comparison.

 

[GST with PSI]

The stock FPR does in fact hold pressure, it is the AFPR that causes pressure drop. Also, if you're refering to "on" when you say "run" position, our fuel pump doesn't activate until you crank the car.[/QUOTE]

The fuel pump actually does activate briefly before the car is started. Check the voltage at the pump test port on the firewall after cycling the key from off to on/run. Or, you could stick your ear near the pump unit at the tank and actually hear it briefly. It cycles on for 2-3 seconds.

If you could explain how a fuel pump that allows an extent of external leakage can hold residual fuel pressure, and also how the stock fuel pressure regulator aids in this process?

thanks..

 

[steve]

The fuel pump actually does activate briefly before the car is started. Check the voltage at the pump test port on the firewall after cycling the key from off to on/run. Or, you could stick your ear near the pump unit at the tank and actually hear it briefly. It cycles on for 2-3 seconds.

I wonder if this isn't a rare '90 abnormality or some flakey CAS's out there pulsing on power up. None of the ECU code I've looked at had anything that would "prime" the fuel pump on power up. I know that none of the DSM's I've worked on ever turn their fuel pump on before you engage the starter and I know that my car never has in the 15+ years I've oned it.

I haven't torn the early 90 GST E705 ECU code apart so perhaps it does unlike all the rest.

Since I have the code I guess I can test that possibility.

Steve

 

[xveganxcowboyx]

I'm almost positive I can hear my fuel pump start and run for a few seconds on the run position. I've never cared much, but I've always heard it.

 

[AWDfrank]

Well I am putting in a fuel pump and 560 cc injectors this week, I'll be using a safc2 to tune it. I hope my knock is fuel related and this will take care of it. Today I floored iT all through 2nd with little knock, later I floored it in 4th and had 16 counts of knock..?

It is so intermittent, sometimes I will get no knock, and sometimes while cruising I get 5-10 counts. If this is real knock and not PK, could I just have some type of engine damage? Unlikely I think but I don't know.


10 psi btw.

 

[romeen]

NoX nitromethane octane booster didn't help which I found interesting. If I was close to knocking for real, half a bottle of NoX in half a tank should have helped, but it didn't seem to affect my timing curves.

Mack, refer to post #18 in this current thread.

http://www.dsmtuners.com/forums/showthread.php?t=257611

Unless I made a mistake in that post it may explain why the octane booster didn't yield the results that you were hoping for. On that note have you tried actual race gas or using toluene to increase octane? With the octane boost your left guessing how much final octane you actually have. I also suspect some misleading marketing hype going on with those. The race gas has a known and reliable value. With the toluene you can easily mix it to achieve a predictable target octane level. Just a thought.