phantom knock??????

[tboostndsm]

Hey guys an gals,
Some one told me that the problem with my car is PHANTOM KNOCK;I was wondering what that phrase means and if it is any different from just KNOCK.Your help is appreciated very much.

 

[thegreatms]

Phantom knock is what people call knock when they don't know what is causing there knock reading. People get frustrated when their knock count jumps from 0 to 43 instantly or when they get knock at part-throttle.

Usually they think it is not being caused by preignition, but it might be. Basically pk is knock without a clear cause. It could be:
real knock from lean condition or to much heat
real knock from rich condition
bad knock sensor
bad wiring or grounding
any kind of noise or rattle that triggers the knock sensor
any engine sound

as you can see there are lots of theories and possible solutions

I suggest that you rule all the above before going to the "other" solutions. Those solutions are ways of tricking your ecu to not see the knock that the sensor is saying is there. This could be dangerous if the knock is real and could cause engine damage

 

[BrokenTsi]

all phantom knock can be fixed.

I had a bad case of "phantom knock" at part throttle. If i even thought about touching the gas pedal, my logger read 43 counts. Under normal conditions, there shouldn't be any reason for any knock at part throttle on a well tuned car. Turned out my lifters were ticking really bad because of a bad bshaft bearing causing low oil pressure. After tearing it apart, the bshaft grenaded inside causing a whole mess and now I need a new engine.

What's weird is, its been like that for 2 years and the car runs fine, just pk every now and then.

From what I've read and people I have talked to, PK can be fixed by doing an alternator re-wire (just search it, you'll find it), a new knock sensor, new oil change with engine restore or lucas, and some others I forgot.

 

[AL92]

I think a better word than phantom knock should be random knock .
In my car I get it off idle every once in awhile.I have changed out my knock sensor but the phantom knock is still there occasionally.One pretty much foolproof method of getting rid of it is to get a custom chip with phantom knock programming to remove it or dsmlink or other standalone but this of course is expensive solution.

I personally suspect most phantom knock off idle and of course off boost is caused by the lifters.I notice in my car when engine is very cold and it gets cold here in sask the knock reads high until car warms up.Then it goes away and you can hear my lifters.As the car warms up the knock at idle goes away and you then get the odd knock off idle.I have no knock when under boost at current boost setting of around 15psi.

I have a set of brand new lifters to install and hope that should fix that.That thing about the balance shaft bearing makes some sense too.Also I had a balancer pulley that fell apart..something to really watch for on any higher mileage dsm and maybe that was also causing some knock.It was making the timing marks on my car highly inaccurate as the marks had slipped around on the balancer.Too much timing can cause excessive knock.Check your timing and balancer.To check balancer put engine at tdc check timing gear marks and if they are dead on and your balancer mark don't line up like it should your balancer is falling apart.Word to the wise.


Also installing the new lifters can be done pretty easily there are some faqs on it.I have bene waiting to maybe do my valve springs and valve seals at same time as well as installing my new cams.My phantom knock seems to be a bit less lately not sure why.
Using mobile high mileage synth latley maybe it has helped.I also tried Mopar combustion chamber cleaner once and used restore once.I don't like that restore stuff it looks way too thick and dont pour well at all so not likely I will use it again in my car.Some guys swear by it.

 

[GPTourer]

I was under the impression that there is no such thing as phantom knock. Knock is knock period, you fix it with tuning and that's that. Phantom knock is just some sort of other engine noise that is fooling your knock sensor into counting it. Finding the source of that noise should be priority one because obviously something is seriously wrong with your engine - or you're just like the STi guys who run EMS and tune to the ragged edge, remove the knock sensor anyway and hope for the best because they just have such noisy valve trains.

 

[AL92]

Knock happening under boost is much more dangerous than off boost off idle knock the type that is usually called phantom.You can quickly see if knock on boost is real by trying race gas,or xylene or toluene mix in about 1 gallon xlyene to three gallons gas.This raises octane and no way you can get real knock at 10psi on that type of octane level.

knock off idle is not a big danger.The engine has safety mechanisms and it greatly retards timing when it sees knock.This kills power but not a chance an 7 something 1g compression engine is going to detonate itself to death off idle and off boost!!

The phantom knock likely is real engine noise..and most likely its the lifters.The old design sucks the new replacements are much better.Phantom knock off idle is annoying.It makes the car really really gutless when trying to leave the lights.You can lose a race with a diesel rabbit when its doing it.But it dont' do it every time and thats what makes it phantom.Again think phantom is wrong choice of word as it implies its not real.

 

[GPTourer]

Again think phantom is wrong choice of word as it implies its not real.

Hey, I believe in ghosts, so I think its a good word.
Phantom as in, now you see it now you don't. To me, a car could knock repetedly on too much boost, with a heatsoaked IC on a hot day on 91 octance gas everytime and that would not be phantom. But a infrequent lifter tick might come and go causing the ECU to reatard the timing, but then again it isn't "real" knock. I guess that's why people who are very confident in their tuning might forgo a sensor entirely because they realize there's going to be too much noise and interference for a knock sensor to actually do its job right. I'm thinking the Mitsu engineers allowed for lifter tick in the OEM sensor, so it would just be a matter of getting them back to OEM spec to keep timing from being pulled off the line.

 

[UPGRADED2G]

This is some really useful info.....I learn something new everyday. I guess i have to push my turbo upgrade about ANOTHER month back. My lifters sound really bad and like most people with a 2g, I have nothing to read knock from. I should be getting a logger pretty soon and i guess if i log and see that timming is being taken away under boost....then maybe i have some knock going on.

 

[UPGRADED2G]

Sorry for the slightly-off topic question. but.......

Im guessing faulty/worn lifters many create some knock at idle, however, what is the impact on oil pressure.

after starting my car again after a long drive(30mins-1hour), it sometimes struggles to stay started and eventually shuts off because the oil pressure does not build up fast enough. Can i blame this on my lifters or on the oil pump. It only happens when trying to start her up when she is already warm after being driven for a while.

 

[Turblown]

PK can be fixed by doing an alternator re-wire (just search it, you'll find it

I just searched for this because I've never heard of it before... didn't find anything. Are you refering to a Fuel Pump Rewire? If you are talking about actually rewiring the Alternator to fix PK, I would be very interested in some more info. Any links to a VFAQ on the subject? Thanks!

-Turblown

 

[1stGenRocks]

i always thought of phantom knock as off boost knock. usually its from the lifters. one solution is to get a chip from dsmchips.com which has code that only lets the ecu retard timing from knock if the engine is in open loop mode (non idling). this seems like a should get rid of problems from the lifters ticking and causing the computer to pull timing.

 

[jetdriven]

I had the worst case of phantom knock as anybody. 43 counts at 10" of vacuum no matter what, new lifters, new head, tuning, etc. everything. Even a custom EPROM.

Above 3,000 RPM the car ran fine obviously thats not a solution driving in town.

I finally bought a J&S safeguard. Disabled the factory knock control. fixed it

 

[AL92]

I had bad phantom knock until a few days ago.I have new lifters but didn't install them yet.I did new knock sensor and most of the other ideas with no help.
I put in custom chip in my new eprom ecu with phantom knock elimination code.Its all gone now.I can leave the lights and never see any knock fact for first time in long time saw like 1 count of know today running my 50 trim at 14psi with stock intercooler.
And don't even have my water/alco injection installed yet.

Not sure where you get this J and R safeguard or how much it is.Eprom ecus are pretty pricey but they do work if you get the phantom knock elimination code or use dsmlink pretty sure you can dial it out with that too.

 

[86MCSS]

I recently got mad phantom knock. Its my lifters. I need to buy new ones. But even in idle if i hit the throttle the knock counts jumps up to 15 or higher depending on the RPM. Higher RPM higher knock. Under WOT it automatically jumps to 43 im losing all my power from no timming and it sucks.

 

[BrokenTsi]

I just searched for this because I've never heard of it before... didn't find anything. Are you refering to a Fuel Pump Rewire? If you are talking about actually rewiring the Alternator to fix PK, I would be very interested in some more info. Any links to a VFAQ on the subject? Thanks!

-Turblown

Yeah, I know a few people who did it and for some reason it worked. They just added a bigger gauge wire from the alternator to the fuse box, then from the fuse box to the battery terminal. I guess it provided more voltage to the knock sensor thus giving it better readings at lower rpm's or something weird like that. I'll search for the thread, he posted a "how'to" here on tuners.

 

[Turblown]

Yeah, I know a few people who did it and for some reason it worked. They just added a bigger gauge wire from the alternator to the fuse box, then from the fuse box to the battery terminal. I guess it provided more voltage to the knock sensor thus giving it better readings at lower rpm's or something weird like that. I'll search for the thread, he posted a "how'to" here on tuners.

I'd really like to see how this works... Please post a link ASAP! Thanks!

-Turblown

 

[dsmchips]

I had the worst case of phantom knock as anybody. 43 counts at 10" of vacuum no matter what, new lifters, new head, tuning, etc. everything. Even a custom EPROM.

Ouch! I thought the PK chip fixed it.
Anyway, PK that I have seen is a random knock, usually very high knock count, like 43, coming out of nowhere. My 2nd GVR-4, #880, has it. You will be cruising, steady state, about 25% throttle, and BAM! All of a sudden you see the knock sum climb to 43 for no reason. But, in the case of #880, I know why: the PO took it to the dealership to have all the gaskets replaced, and they left gasket scrapings in the oil galleries. 2 weeks later, the engine seized, and the dealership said "not our fault". (I found the gasket material later when I tore it down). Anyway, I'm SURE #880 has ring and bearing damage from oil starvation, since it was enough to spin/seize the rear balance shaft. But, in the meantime, the PK code has made the car VERY driveable, and I rarely see any knock. I am tearing it down next week for a complete rebuild anyway, but I just wanted to emphasize that oftentimes you don't know the history of the car, how it was abused, if it was run out of oil, etc, and the PK could be REAL knock, from a rod bearing, or the pistons slapping.
I can't tell you how many people have emailed me and said "I found my PK: I threw a rod".

 

[GPTourer]

I have nothing else of consequence to add to this thread. But I just want to say that it kicks ass the way you refer to your car as "#880"

 

[jetdriven]

I suspected this in my 93 AWD. I bought the car with a bad motor in it and I swapped in a 150K mile 6-bolt engine from a 92 GST.

After all the other fixes I replaced the rod bearings. Even though the old bearings plastiguaged at the tight end of the limits and only had normal witness marks I put in some new Celvite 77's.

I have also had the head off the car and the cylinder walls look good.

Mind you this short block has a new head and 160K on the bottom end. It has 150-150-147-145 compression and uses a half quart of oil every 3,000 miles. most of that is leaks I'd bet. Piston slap usually occurs right at the point of "zero loading" i.e. the transition from drive to coast, on my car about 12" of vacuum. mine would get 43 counts continuously.

So far the J&S safeguard has cured this but I have only driven the car a few miles in the month since I bought it. my next step is a wideband 02 as I dont think you can truly tune the car without one. There might be a lean spot somewhere enough to cause actual knock and the factory 02 sensor wont respond enough or accurately to catch it. But the knock hysteresis will continue.

 

[2GeNTSi]

Ah I'm starting to think this is what I've been looking for.

My lifters haven't been a problem lately, I only notice them on cold mornings so I doubt it is that. However my injectors tick as well, sometimes I think it's my lifters but if I go out and check and they are coming straight from the injectors (can feel them pulsing, one more then the rest)...typically blipping the throttle would stop it.

Is it possible for a shop to diagnose the problem as knock? I don't know anyone with DSMLink or any other ECU with the knock reading capability. There's a shop around here called Newlogics, supposedly all former Mitsubishi mechanics with DSMs so I think they might be credible.