General knock under any load

[sweet97]

The screw on the mas will cause the engine to run richer if it is turned in restricting unmetered air. The honeycombs are often removed which can cause problems but especially the metered one. The EGR specifically prevents partial throttle knock but I will leave that research and decision to you. Mine is going back on!
Do you have the stock "accordian" intake or a hard pipe?
What did the new knock sensor cost? Thanks, Mark

 

[ucancallmedebo]

hard pipe,from road race i think, just checked mass. all stock with silicon over the screw, really crappy homeade adapter for the filter, will order new from ebay. while i was at it i made a pressure tester out of one of my old 2G soft pipes with valvestem and plug. leaks everywhere but the intercooler piping. one of the tees from the y from the intake to the blowoff which goes to the wastegate had a hole right in the middle of it,like somekind of a bleeder. put solid line there. also turbo elbow leaking pretty bad fixed this to. only leak left is small one at the shaft which comes out of the throttle body,where the spring is wraped around.any ideas?before fixing leaks 20-30 psi lasted about 5 sec in intake, now it lasts about 30+.. lots of air wasescaping though that bleeder tee no doubt giving my wastegate a laggy signal and poor boost control. i had fast boost response before though? i have not driven it yet but these leaks are after turbo. moral of the story pressure test first. Anyway the knock sensor was 91.00 at mitsu and took 1 week to get,ripoff. search on these forums because i know one can be had for 65 somewhere. i had already paid for mine when i found this out.looking at my logs it seem my knock is related to boost pressure as there is a lag between TPS and knock. this lag has to be boost, hope fixing these leaks helps the problem dont leaks after turbo cause during vaccume- lean condition , during boost- rich condition. during vacume it would suck in extra air past mass and during boost push out extra air which is already metered. my o2's have gotta be right enless something is lowering my 02 voltage , i already disconected my blue wire mod from afc. helped .01-.02. ? am i thinking right on the lean/rich boost/vaccum theory?

 

[sweet97]

If you had a leak after the turbo you would run rich as the air would have been metered and then lost but the fuel for that air would still be put into the engine. Your TB shaft is leaking. It has plastic bushings. I've neve done the job though the TB comes off quick. i saw a pic once on leak testing. was it on vfaq? www.roadraceengineering.com has some info on 02 voltage and what is considered low and yours is. they might have some fixes. I run in the .90+ range and my motor likes a 10.5:1 or 11.0:1 a/f for it not to knock. Mark

 

[Joesmoke]

Sweet97,
This may not give you guidelines for "tooning" but will give you some real world examples of real knock and false knock.

Ok here goes. My thoughts on this are based on some common sense, 3 yrs research and real driving experimentation. Here are the key points about knock:

1. When I start to talk about PK, I always try and back it up with "if your engine is healthy" Heres a checklist:
-Good fuel system and enough fuel
-Good spark
-None or very minimal boost leaks
-None or very little carbon on cc or on piston top
-No sharp edges in cc (chunks of piston broken, valves etc) Smooth CC
-Reasonable boost for altitude and ambient temps
-Reasonable timing adv.

So your engine is healthy as per the list above. What is going to cause real knock at high levels, at all kinds of throttle positions? Almost nothing. (If i missed an example of a healthy engine above, the basics still stand. Healthy engine) Another note. I added water injection and still had high, random knock counts. BS on logger counts after that!

2. Please keep in mind that real knock values will climb and fall slower then false knock values. Pegging an automatic 43 count is very indicative of false knock. This is not only true on our cars, but other makes and models. Widely accepted as fact. Also, most engines will experience some detonation and preignition that is not destructive (non audible).

3. When mitsu designed the knock value, what does it mean? Is a 43 count of real knock even enough to cause massive failure? What is there scale? The jury is still out on this, but keep in mind, I have heard of real detonation values of 12 holing pistons. What do you think 43 would do in real life?

4. Mitsu (and most other manufactures) build in high levels of safety to avoid warranty claims. Cars run rich from the factory, etc. For those guys like me still on stock injectors, keep this in mind.

I have tested with a grounded and disconnected knock sensor, and what do you know, the car did not melt down! Now, I run meth injection, and that allowed me to go pretty far, but even when I had several split second audible detonation events, (injection pumped was failing!) no harm done. I was able to back off the throttle in time. I have not yet examined the plugs for damage from those events, but I bet they wont show heavy damage.

Even when not injecting, I could achieve above factory performance settings with no sound or sight of detonation. This testing proved my theories above on the knock sensor circuit in my car. Even at a 43 count of knock (previous to pulling the sensor), I never heard audible pinging! And I have a 3" exhaust with no resonator that is pretty loud on wot. What does that testing mean to you?

As a big note to add to the above testing, ambient temps were always low. 30-50 deg f. This has a huge impact on knock sensitivity. If the ambient temps were greater, I would have to change something. Less timing and or boost, and more methanol, but I believe the tests would still be relative.

I dont advocate yanking your knock sensor and going balls out. I advocate a better knock sensor circuit, like the J&S safeguard (if you have satisfied yourself that your engine is healthy). I also say for a person to use their own judgement as to the state of their vehicle, rather than soley relying on inet discussion boards. This and other boards have their place, which is good, but nothing beats getting your hands dirty and just plain common sense.

Now, all the kiddys who read once that "buschur blew his engine with no knock sensor" and that Im an idiot can begin their flames!

 

[sweet97]

OK Joe. I will replace my knock sensor since it's 12 years old. Followed you until the part with 43 counts of knock and no audible knock while having a loud exhaust. I do not have hearing that would allow me to hear knock and not with a loud exhaust! I believe you are trying to say that even short run at a full knock count might not cause any physical damage due to the safety net installed by mitsu? I don't have a huge knock issue but need to know all I can to eliminate as much as possible.
I've backed my timing a couple of degrees and am hooking my EGR back up because I lke the idea of runing cooler cylinder temps and there is no performance advantage in having it not functioning. A bigger fuel pump is going in also al with an AFPR. If you could elaborate on the one part that confused me....well thanks for all! Mark

 

[Joesmoke]

Sorry sweet, I thought you had more knock problems. Post is still valid though for the title of this thread.

So, just want to say again that my last post is a buyer beware, your mileage will vary! My setup is not your setup etc.

1. Yes, replace that old knock sensor. It lives in a harsh environment.

1a. If you dont feel that you could hear detonation, then all you have to go on is my experiment, others stories, common sense, etc. You could never varify for yourself that your knock count is accompanied by a audible detonation.

2. I had a multiple, frequent, up to 43 counts of knock. Not once did I hear detonation (pinging) over the vehicle noises. Even on a stock setup. This was happening over a period of months. Once again, not all detonation is audible, but not all levels of detonation are damaging either. Individual call as to what your comfortable with.

When I increased vehicle performance (increased timing and boost, decreased fuel) and my methanol injection system was not operating at full capacity, I could hear detonation. (pinging). To compensate until repairs could be made, I increased the fuel again and the detonation stopped. I did not however, alter the base timing or boost levels.

3.If you have satisfied yourself that you indeed have a healthy engine, then what I want you to ask yourself is "Is that a true knock count or a false knock count?" If it is false, its not the knock (or lack of it!) that is hard on the engine, its the huge timing retard the ecu applies. This generates a lot of heat, and your back to square one. A sustained high knock count (false) being damaging, especially on long sustained runs.

As far the "safety net", there is no real data showing that there is or is not one present, and if there is, how far does it extend? Is each car different? You would have to be comfortable with that.

I believe however, that there is one present and in the case of my car, extends farther then the general populous believes.

 

[sweet97]

Ihave fullthrottle's Scanmaster which allows me to see the knock count as it is occurring. I have to admit I have had the car 1 season and have not changed the plugs or wires. Me bad! I have the plugs and guess I will go with Magnacores. Just realized I need to install the Walbro and I have an AFPR to go with it.
I understand what you are saying about kock count and it's relatioship to actual counts of knock. As I understand it 7 counts is the line under which nothing really happens. After that timing can get pulled. I will be happy with a count of 10 and under. Though my situationis not as badas the post originator who seems to have constant knock having causes an "cures" listed is important and often lackng. As I said often guys say TUNE, TUNE. OK! Help out with some solid suggestions and parameters! thanks, Mark

 

[ucancallmedebo]

anyone have any insight on if it is possable that my low 02 reading is correct, could something be lowering the voltage. 02 has no relation to knock but is is a good indication of lean/rich , and mine says lean. I beleive as it is a good working new sensor.

 

[ucancallmedebo]

just took the peice around the block. backed timing to 0-1 btdc. most boost leaks fixed,afc at +5 hi across and +4 low across. at same mbc setting i am getting 9-10psi because of fixing the leaky tee fitting. FTL 96.8% FTM 100.00 FTH 104.6%, looks ok.

idle

rpm 750 injp 2.81-2.55ms 0 knck 02- .66 .44 .03 .03 .05 .05 .05 .15 .31 .52 .62 .52 .09 .19 .72 .76 .56 .19 .09 .03 .68 .74 .60 .05 .13 .60 .35 .11 .66 .72 .07 .76
this look normal?

First gear

0 knock, 25-30% th, 750-3800 no knock

second gear
rpm inlp knk 02 tps tim
2200 5.63 0 .07 40.2 28
2400 5.63 0 .76 41.0 29
2450 5.63 0 .74 41 29
2550 5.63 0 .72 41 28
2687 6.14 0 .05 41 29
2900 6.39 4 .58 41.4 27
2937 6.39 3 .58 41.4 27
3125 7.93 8 .21 51.1 27
3281 8.95 16 .82 54.2 24
3406 9.72 20 .83 54.2 20
3562 9.98 26 .83 54.2 19
3750 10.49 25 .83 53.5 17
3906 10.75 25 .85 53.1 17
4093 11.77 29 .83 64.8 17
4312 12.54 29 .83 68.7 15
4687 13.31 32 .85 57.4 14
4875 12.79 32 .85 54.9 14
5000 13.05 31 .85 49.6 14
5156 13.31 35 .85 49.2 14
5281 13.05 34 .85 48.8 14
5468 15.10 38 .83 67.1 15
5750 14.59 37 .83 65.6 14
5875 14.59 37 .83 63.2 15
5906 9.47 37 .83 48.8 16
5531 2.30 27 .85 10.9 16
4343 1.79 0 .83 10.9 45

while cruising in 5th throttle roll on at 2200+- from 20% to 25-35% produces knock in the 40+ rance while i noticed another time 5th at 2300 with th up to 99.6% produce 0 knock. my 02's look better with lower boost +.02. phantom knock + real knock maybe.


under heavy throttle in 3rd exhaust temp would go over 1600+ at 10psi thats what keeps me out of heavy throttle,also the 40+ counts of knock didn't help.

 

[pboglio]

ucancallmedebo,

Your timing curve for the 50% throttle is not right. Anything below 38% or so is around 38* of timing or so for cruising conditions, anything above 50% throttle is basically WOT timing or around 21* MAX. Something on your SAFC is ####ing up the ECU from getting an
accurate picture of the airflow. The computer thinks there is MUCH less airflow then there is. It still thinkgs your cruising around at light throttle, when in fact 50% throttle is bringing the turbo onto boost and bringing in a LOT of airflow. Your 02 looks O.K. for the given TPS values, I am running .86 volts @ WOT on 20 psi and have little to no knock.

NO amount of fuel is going to keep you from knocking with that timing advance curve you have. On the contrary, there is a bell curve for A/F ratios that gives the least knock. Adding a LOT of fuel or running lean both make your car knock. A fuel pressure reading would also help people figure out what is going. You should be seeing around ~45 psi fuel pressure at 10 psi of boost.

 

[ucancallmedebo]

my karmen hz on the afc at 10psi was about 1500,is about 32 deg here. when i first got the car boost was at 18-19psi and karmen was about 2000 at about 60 deg outside temp. these looked close to what they should be. is there anyway to check the afc or 2g mass. it runs and drives very good exept for high knock ,ish timeing. would switching to gm maf help? remember i have 2g mass on 1g which causes computer to see about 21%less air which is compensated by 550's. a lot of people run this setup without problems even thought though the ecu is seeing 20% less air. a chip would solve this prob but would involve get eeprom and socked with chip.

 

[ucancallmedebo]

i was just thinking but since the computer is 20% behind on my airflow, it knows what th% and rpm its supposed to see boost at. so with all the porting i have with the b16g and its ability to move more air at say 50%th at a set RPM i am still 20% behind as far as the ecu is concerned with air"super high timing". so if over 50%th the computer is looking for boost i gotta get it to 70%th before its gonna register to start giving big fuel.basically the 20%less air the ecu is seeing makes more of a diff at 50%thottle than at 100% becuse this is where it transitions.form vac to boost and the computer hasn't yet making it run lean..so by driving cautiously in light throttle could i in theory be making the problem leaner at lighter throttle.

 

[pboglio]

Well,

You even mention NO knock from 25-30% throttle. WHY? Because the ECU normally runs the car around 38* timing in this range, the motor can easily tolerate it and its good for getting a complete burn. Once you step over 50% throttle you should be seeing 21* timing or so, your showing 27* and thats right where your knock counts start going sky high. This is what I tried to explain to you. I have over 200 datalogs, I'm right now looking at Mitsubishi's timing advance curve tables (3 of them) and I know EXACTLY what the factory expects the ECU to do as far as timing is concerned. You have TOO much timing advance at anything above 40%-50%, right where you get high knock counts. Once you start knocking at part throttle you don't get the timing back unless you back out of the throttle, even if you had NO knock at WOT. This is why I tell stubborn guys to put back their EGR's, even though it gets shut off at WOT. WHY, because part throttle knock pulls timing even at WOT. EGR helps DEFEAT part throttle knock, aside from reducing emissions.

You lose 1* of timing for every 3 counts of knock. At 4100 rpm, 69% throttle, your own datalogs show 29 counts of knock and 17* timing, the computer pulled out 9* which means you would have had 26* of timing. This is insane timing at only 60% throttle. I've never even seen a race car run this much timing, you actually LOSE power adding too much timing, even with NO knock.

I'm telling you something has screwed up your ECU's idea of what the airflow is coming into the engine. Your looking in the wrong place to fix your problem.

 

[sweet97]

DEBO help a new guy. If the 2g mas flows 21% more air why would the ECU "see" 21% less, I assume metered air?
Does a logger show the a/f ratio? The Scanmaster from full throttle shows all of the info a logger wouldshow plus more at actual time while it is occurring and I would highly recommend one if you have the $250. Mark

 

[ucancallmedebo]

thanks 4 the reply. sound like u know what u are talking about. u think the 2g mass is making me run to much timing, the 20% less air the ecu is seeing is changing my timing curve a lot. the only way to change this my getting 450's and moving afc to +20across. balencing out 20% lean with the 2g mass or getting a chip to compensate. do u think i could pull enough timing on the CAS to compensate for this or is the computer gonna do what it whats according to how much air it is thinking it is getting.

 

[ucancallmedebo]

say at idle the mass flow for example only on a 1g mass in a 1g is 100hz , a 1g with a 2g mass would only show 79hz. the computer sees less air and runs timing and everything else assosiated with less air even though it is getting more. the 550 flow aprox 20% more fuel than the 450's. the comp thinks it still has 450's so it balences out the A/F mixture but the ecu still sees 20 less air.

 

[sweet97]

I guess you would have 20% less TPS if the mas flows 20% more huh? If you need a set of 450's I can help. Mark
PS: We were posting at the same time but I understand.
I use the MAF-T as a blow thru and love it!

 

[ucancallmedebo]

tps dosn't change but i was just saying that because the ecu is 20% behind in air it sees that if u could get 20% more tps% "at the same airflow"maybe this would change its curve. this is getting complicated

 

[ucancallmedebo]

my mmcd software show 02 in voltz. no a/f ratio. mine looks lean compared to everyone elses.

 

[sweet97]

Well I get the idea so I'll back off and watch. Perhaps PBOGLIO can steer you in the correct direction. mark

 

[ucancallmedebo]

what he is saying makes sense but i see a lot of people with this setup that runs fine with the advanced timing assosiated with 2Gmass. and thats a lot of knock for only 10psi. i am hoping with the 255wineboro it will help the prob.

 

[ucancallmedebo]

about the EGR if i had one laying around i would put it back in,but the car came like this along with all these other problems.

 

[sweet97]

I went right to the maf-t so i know nothing about the afc's. There must be plenty of guys with an lying around that would help you. I will have a spare but not for a week or 2 until I test one of the 2 I have as like I said I am making mine operational as my car came with it rendred inoperable and I have read enough to convince me it's worth running for better gas mileage, cooler cylinder temps & lower emissions.
Stupid quetion but can you test the safc to see if it is operating correctly? Maybe makle an adjustment that would show on your logger at idle so you could see if the result is the anticipated one. Higher 02 voltage if you richen the idle? Mark
PS: Excuse my lack of knowledge if I ask ignorant questions!

 

[pboglio]

what he is saying makes sense but i see a lot of people with this setup that runs fine with the advanced timing assosiated with 2Gmass. and thats a lot of knock for only 10psi. i am hoping with the 255wineboro it will help the prob.

Reconnect your EGR if its disconnected, that would help prevent some knock with whats going on. The ECU relies HEAVILY on a properly working MAF. TPS is primarily used for fuel enrichment on throttle tip in, then as a BACKUP along with the CRANKANGLE sensor for when your MAF fails. The SAFC just uses low, mid, high TPS signals to then ALTER the airsignal going back into the ECU. Forget TPS, think AIRFLOW.

I'm not trying to be harsh here. You have to get a baseline for what is a CORRECT airflow signal under ALL driving conditions. THEN you can plop in your 2g MAF, 550 cc injectors, SAFC, and 255 fuel pump and know what you have. You have already lied to the ecu with the different MAF, different injectors, altered the already screwed up airsignal with yet another device that screws up the screwed up airsignal, and then want to add a fuel pump that will blast so much fuel at part throttle and idle. You can't possibly know where you stand with all this hacking going on, no way.

The ECU uses grams/rev from the MAF sensor. Anything below 1.5 g/rev you get high timing, 1.5-2.1 g/rev you get the medium timing table. Anything over 2.1 g/rev you get below 21* timing, according to 2G tables. The 1G run a little more cruise timing along with WOT timing. The ECU doesn't care less what the TPS says, its all about airflow. It doesn't really matter what other people have on their car. YOUR car has a timing curve that is punishing your motor, I have both a 1g and 2g with datalogs for both. Only thing saving your motor is timing retard, and just barely by the looks of it. Don't take it the wrong way, I'm trying to help you out. You have 3 variables (2g MAF, 550 injectors, SAFC) that are killing you. My recommendation, pull every one of those damned things off, then get a base line using your datalogger, then see what your timing curve looks like then. I bet you see little knock and a more reasonable timing curve.

Next, reinstall the 2g MAF & SAFC without the big injectors and trouble shoot from there.
Then finally add back in the 550 cc injectors when you have a feel for whats going on.

 

[ucancallmedebo]

sounds like a good idea,but i dont have any of the old parts to swap back to. would an eprom chip programed for 2Gmass and 550 injecters make a good baseline. this should bring my timing back to factory spec. i could take off the afc as this would involve splicing 1 wire which is cut "airflow wire" the rest i could leave as they only tap into other wires. this would take the ability of the safc to change the airflow signal away. besides timing i cant get my 02 voltage to go past .85 at any setting, including +15 across on afc. before the rewire and disconnection of the blue wire mod from the afc i would only get .82-.80 volts. i thinnk my problem could be weak fuel pump aside along with the high timing. oh,and do u think i could make base timing -5 or 10 with CAS. if i did this and a run showed improvement than i would know it is the timing 4 sure. thanks again for the input oh checked my plugs again an they still look very good.