I haven't posted much on my progress with the 3900 because I'm not making any. Recently I swapped out the forged motor after a couple of careless mistakes damaged it. I currently have the stock 3900 reinstalled.

I need some advanced ideas as I've checked most if not all of the most obvious potential causes.

Problem:

Before the engine swap and after a large quantity of oil was ingested by the engine, I encountered a problem with intermittent unexplained detonation under boost levels and typical hard launches where I did not have it before. I thought the turbo may have been contaminating the intake charge with oil but that does not appear to be the case after inspection.

The problem occurs under boost pressure usually around 3-4 psi and up and although I let off the throttle it continues to pepper deceleration with timing retard. For example, an initial 3 deg retard at 3200 rpm and 4 psi will show timing retard for as much as the entire Kpa range for that rpm plus some in a few of the surrounding cells in the datalog although no throttle is being applied.

This problem exists across two different motors which previously had no trouble with boost, the second of which was compression tested and showed 238 psi on all 6 cylinders and has less than 8K miles on it. It's also important to note that this current motor ran fine previously under the following conditions:

212 deg intake air temps
7 psi of boost SUSTAINED!
89 octane fuel, detonation free except for a few small areas due to needed timing change but nothing like what I'm currently experiencing.

Currently:
I tried colder plugs in the built motor to no avail.
Premium fuel makes little to no difference.
Retarding the timing does not seem to help although it shouldn't be necessary due to previous trouble free performance under boost with the same spark advance.
Current inlet temps are peaking at 175 deg but only at 7 psi which is cooler than the previous trouble free performance.
Water injection is not helping! and wasn't present on this motor when it ran 212 deg intake air temps.

My suspicions:

Glitch in the code mask during a parameter change.
ECM problem.
Ignition Module.
Coil pack.
Spark plug gap, long shot but I'm considering it because I don't remember at what point I changed the gap from .030 to the stock .040 although I don't believe it's the cause.
Never heard of a crank position sensor causing this kind of problem.

Both engines have all of the above items in common as well as the other sensors.

I hate to throw parts at it but I'm tired of the current state of things, the car is a blast and runs strong up to the trouble spot but I just can't risk anymore kicks from the motor due to detonation.

Most of this timing retard occured after releasing the throttle.


SAME TUNE only with ~25 cooler engine temperature and cooler intake charge since it was early in the morning as opposed to 6 pm for the above log. This log is insane with detonation.

[This message has been edited by Joseph Upson (edited 08-29-2011).]

I don't want to sound ignorant, but what the hell. Has the cam timing map been checked out? I'm afraid I don't have much internals experience with the 3900 to this point.

First off.. not running 93 is just crazy.

Secondly, I dont know what ECU you are running.

third: You never said what your commanded timing advance is.

fourth:

Is it possible it's erroneous? Can you hear it knock? I've read that a loud exhaust can cause a false reading. I've also read that the knock sensor must be from the same engine as the memcal.

What's your AFR?

Have you checked to ensure that the vacuum line is still hooked to the FPR?

This is a complete shot in the dark.

The fact that it has done this on several different engines tells me that it's not mechanical, but related to the PCM or a peripheral. (Duh.)
The 3900 has VCT. Right?
Does your tune retain that? Is it possible that the cam is staying too advanced, and not retarding as the revs come up?

When you make a change your tune, do you save the CAL file as a new file, and document the changes and results?
I've probably got about 30 (mostly useless) 4.9 bin files, with all changes documented. (Well, mostly)

[This message has been edited by Raydar (edited 08-29-2011).]

Have you tried seeing if there is any change with race gas? If no change I would say its false KR. Could there be some kind of safe mode going on, killing timing to "save" the motor/trans?

------------------
Turbo 3800 E85 5spd spec5
11.53@126.7

[This message has been edited by Justinbart (edited 08-30-2011).]

Important points guys:

Running 93 octane on any engine that does not need it is crazy. No need to get beside the point with those concerns as I've established that the motor ran fine under more intense conditions 212 deg inlet temps vs. current 175 deg and on 89 octane. Also I mentioned fuel grade has no effect although the high compression motor did get premium and don't forget water/meth injection also has not helped since the issue surfaced.

Raydar is on track with identifying that I pointed out the same problem exists across two motors, both of which didn't have the problem before.

I left out details like cam position and ecm because those items are the same or nearly the same across both motors, or at least performed without a problem early on and therefore need not be considered except for a possible bad ec.

Race gas is out of the question as fuel is not the problem given that pump gas has been proven to be fine for both.

I'm trying to find the cause of the problem. Remember, the engine ran fine at one point so technicalities related to engine specifics (VVT, cam position, fuel grade) can be ignored as the problem would have been present from the start if they were responsible.

Thanks for the input though. I think I need a narrow focus here.

I'm thinking ignition system since high load seems to be the trigger point.

[This message has been edited by Joseph Upson (edited 08-30-2011).]

quote Originally posted by Justinbart: Have you tried seeing if there is any change with race gas? If no change I would say its false KR. Could there be some kind of safe mode going on, killing timing to "save" the motor/trans?

Might as well stop wasting words.....

quote Originally posted by Joseph Upson: The knock retard is not false by any means, it was vaguely placed but I mentioned I can't risk anymore bucks from the engine. In other words an occasional backfire along with a jolt can be felt and heard from the motor under these conditions.

quote Originally posted by MstangsBware: Might as well stop wasting words.....

I agree

Actually, my NA '7730 swap is doing the same thing... I'm running 9.8:1 on iron heads and the 3400 roller cam block and 93 octane but lower octane didn't change anything. I'm pretty sure it's a wiring issue. The engine ran great with the stock Fiero computer. :/

With me, a datalog shows my fuel pump voltage is 1.2v and occassionally every 6 ticks jumps to 13.x volts and my motor does a buck. Other sensors seems to be off as well. Part of me thinks I need to move the MAT sensor so it's not over my cross-over pipe. It feels like the engine runs lean and occassionally burps rich.

Note, my gap is at .050. I believe .060 is what's recommended for DIS which I'm running but the higher compression made me lower it to .050. It ran like crap when they were at .035...I'm using NGK UR5s...

[This message has been edited by lou_dias (edited 08-30-2011).]

quote Originally posted by lou_dias: Actually, my NA '7730 swap is doing the same thing... I'm running 9.8:1 on iron heads and the 3400 roller cam block and 93 octane but lower octane didn't change anything. I'm pretty sure it's a wiring issue. The engine ran great with the stock Fiero computer. :/ With me, a datalog shows my fuel pump voltage is 1.2v and occassionally every 6 ticks jumps to 13.x volts and my motor does a buck. Other sensors seems to be off as well. Part of me thinks I need to move the MAT sensor so it's not over my cross-over pipe. It feels like the engine runs lean and occassionally burps rich. Note, my gap is at .050. I believe .060 is what's recommended for DIS which I'm running but the higher compression made me lower it to .050. It ran like crap when they were at .035...I'm using NGK UR5s...

Okay, but this info doesn't help me and you should be running spark gap characteristic of the 3.4L iron head motor. My fuel pressure is steady according to my dash fuel pressure gauge.
Hope you get it ironed out though.

Hmm

quote the car is a blast and runs strong up to the trouble spot but I just can't risk anymore kicks from the motor due to detonation.

Kicks from the motor? KR is nothing more than a mild amount of pinging... All a knock sensor does is pick up that ping noise and transmits that information to the PCM, where the PCM uses the crank sensor to figure out if that noise should be there.

In no situation would any symptom other than some pinging noise be noticed here unless you are running more than 50 degrees of spark advance.

quote I've established that the motor ran fine under more intense conditions 212 deg inlet temps vs. current 175 deg and on 89 octane.

Let me first start by saying, dumb..

Ok now thats out of the way... You are not going to be doing much lowering temps by 25 degrees... you might make a fractional percent more horsepower from density, but you are not doing anything else here.

Secondly, 89 octane will cause A TON of KR to show up on any random GM car out there... GM designs stuff to run on 91+ octane, and relies on the knock sensor to dial things back to optimal levels. It is impossible to say that you are doing anything "correct" when trying to run 89 octane on any setup with boost on it.

quote The knock retard is not false by any means

Considering you have no idea what "knocking" even is, i dont see how you can say this.

quote but this info doesn't help me and you should be running spark gap characteristic of the 3.4L iron head motor

Uhh, wut?

The simple fact of spark plugs is centered around "does it spark?". If you can spark, it sparks, if you cant, it doesnt... Gap on boosted cars can not be left stock typically, Its a general rule of thumb to take spark gap down by 50% over stock at 15psi of boost, and 75% of stock at anything over 25psi.

All of this plug talk has nothing to do with "real KR", or whatever you think it is.

Hook a scope up to the crank sensor and do a time record during the fail event. Just for shits and grins, have you made sure that the knock sensor isn't picking up on something external like mount or accessory bangs or other transient mechanical noises?

quote Originally posted by 86FieroSEv6: Hook a scope up to the crank sensor and do a time record during the fail event. Just for shits and grins, have you made sure that the knock sensor isn't picking up on something external like mount or accessory bangs or other transient mechanical noises?

Sorry, this was already explained. There is no false knock. Try again.

------------------
Turbo 3800 E85 5spd spec5
11.53@126.7

So maybe the whole knock/timing thing is a red herring and we're not seeing the trees through the forest. I'd still like to see that CKP trace. And yes, I have seen bogus crank signals do this albeit not on the 3900. Found a distorted trigger wheel on a high value 5.7 Z28 on a couple of occasions. It really raised hell with driveability. I just haven't seen enough 3900's come through the shop to have regular sex with them yet.

For those of you who are actually trying to help by focusing on my concern rather than win an opinion pole on specifics I've already excluded as potential causes (like you know who above), I appreciate it.

Some of you have probably tried to crank a lawn mower or motorcycle and experienced the kick towards the opposite direction that results from a cylinder firing too early and therefore understand the "buck" I referred to underload that is immediately followed by timing retard is indicative of and proof that the two are related.

I typically try to avoid this kind of nonsense but sometimes it's tough to look away, so I'll ignore the comments from dark dodo that's beside the focus of the problem. I've destroyed enough motors to know what I'm talking about as it relates to my build.

I appreciate the HELP, but save the nonsense rhetoric for someone who cares.

quote Originally posted by Joseph Upson: My fuel pressure is steady according to my dash fuel pressure gauge.

Steady at what psi? I don't know what car your 3900 came from, but looking at the specs for a 2009 Buick Lucerne with a 3900, the spec is 56-62 psi w/key on, engine off. Maybe you're psi is a bit low causing lean conditions? Maybe it needs to be even higher than stock specs with the added boost? I mean, those specs are for a non-boosted engine. Maybe the stock injectors can't flow enough?

quote Originally posted by Joseph Upson: For those of you who are actually trying to help by focusing on my concern rather than win an opinion pole on specifics I've already excluded as potential causes (like you know who above), I appreciate it. Some of you have probably tried to crank a lawn mower or motorcycle and experienced the kick towards the opposite direction that results from a cylinder firing too early and therefore understand the "buck" I referred to underload that is immediately followed by timing retard is indicative of and proof that the two are related. I typically try to avoid this kind of nonsense but sometimes it's tough to look away, so I'll ignore the comments from dark dodo that's beside the focus of the problem. I've destroyed enough motors to know what I'm talking about as it relates to my build. I appreciate the HELP, but save the nonsense rhetoric for someone who cares.

As I tried telling you before. I have this same problem. I tell you think so that you can rule out that it has anything to do with your turbo. The other thing about my swap is that I had the knock sensor disabled in the code to see if that's what was retarding it beyond belief for that to happen and the problem persists. What I did find is that the coil pack needs to be grounded...and grounded well. I also found that the problem only seems to happen after the motor has warmed up. Cold it has alot of power. Can you send me a datalog from Tuner Pro RT so I can see if it looks like mine?

I have seen several 7730 modified bins do some real crazy things, like show so much knock that the knock counts just count up to 255 faster than 1 per second. Swap the bin with a stock one (ostrich emulator) and the issue instantly went away. So you might try switching back to a bone stock bin and start the tuning process over again.

On my 7730 swap (V8), it once started to act up after running great for more than a year. After a length of time (usually about 20 minutes of driving), it would start to randomly buck and kept getting progressively worse... ended up being a faulty knock sensor.

On my L31 (V8) in my truck, I have experienced random bucking that ended up being a faulty crank sensor.

quote Originally posted by 1fatcat: Steady at what psi? I don't know what car your 3900 came from, but looking at the specs for a 2009 Buick Lucerne with a 3900, the spec is 56-62 psi w/key on, engine off. Maybe you're psi is a bit low causing lean conditions? Maybe it needs to be even higher than stock specs with the added boost? I mean, those specs are for a non-boosted engine. Maybe the stock injectors can't flow enough?

Regardless, lean spots would show up in the tune and that's not the case. Remember, the exact same platform ran fine with boost and suddenly this problem occured so looking in this area is not likely the cause. Also this motor is not running an OE PCM so the OE specs don't fully apply. My electric in car fuel pressure gauge holds steady at the same psi from the beginning. The injectors are 60 lb/hr at 43 psi but that info isn't necessary as I keep saying; The engine ran fine under the current conditions and with it's current parts at one point.

Thanks for the input.

So far code59 is behaving much better. There was some timing retard but I haven't encountered the kicking/bucking sensation from the engine yet. I reviewed the log and noted nearly all of the small timing retard points off boost occured mainly during decel and when the AFR went very lean momentarily so that can be tuned out. The timing retard under boost might be the result of a needed adjustment to the spark table, AFR was about 9:1 at that point. The tables in code59 are very different resolution wise from those in TGP code.

I've never had a problem with running stock plug gaps with boost, but that's a function of your ignition system's strength as air density and widened gaps increase resistance. In this case I'm also dealing with relatively high compression at 9.8 and since I can't recall if the problem was present before or after I returned the gap to stock, I'm going to bring it back down to see what happens.

[This message has been edited by Joseph Upson (edited 09-01-2011).]

quote The timing retard under boost might be the result of a needed adjustment to the spark table, AFR was about 9:1 at that point.

Wooooooooooooooooooooooowwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww

Did master tuner akimoto and aaron have a kid together?

quote Originally posted by Hudini: Is it possible it's erroneous? Can you hear it knock? I've read that a loud exhaust can cause a false reading. I've also read that the knock sensor must be from the same engine as the memcal.

quote Originally posted by 86FieroSEv6: Hook a scope up to the crank sensor and do a time record during the fail event. Just for shits and grins, have you made sure that the knock sensor isn't picking up on something external like mount or accessory bangs or other transient mechanical noises?

At this point you two may be right. Since switching back to code59 I have not experienced the bucking sensation under load I mentioned but then I haven't pushed the engine past 5-6 psi yet. I went from 100% water to ~50/50 water alcohol in the injection tank and 93 octane plus pulled a couple of degrees timing and still the exact same result but without the bucking and occasional pop from the exhaust. I returned the plug gap to .030 yesterday where I initially had it also. The timing retard occured in pretty much the exact same location which is highly suspect of a problem part.

The tables in code59 are scaled differently than those in the Turbo Grand Prix code so it's not easy and in some cases it's impossible to match values across the spark and fuel tables not to mention the effects of tables they don't share in common activating. You would have to experience all that happened in a very short period of time to understand why problem solving the matter is so difficult as immediately prior to the incident several things happened/changed that could have been responsible.

Okay, so I didn't mention that I may have bumped the already slightly distorted knock sensor putting the turbo back on before the problem occured, and maybe I did ignore the fact that I had some difficulty removing it from the engine during the swap where I thought the top half may have spun a little and thought to myself "I need to change this thing".

Hope a replacement fixes the matter as the consistency of the error is uncharacteristic of a mechanical problem and the fact that fuel grade and alky injection has had no effect further supports this.

I appreciate those of you who have taken a constructive approach to helping resolve this problem.

[This message has been edited by Joseph Upson (edited 09-02-2011).]

Solid lifters?

Stock roller lifters

Disturbed the knock sensor wire at the sensor today accidently by hand before starting the car while trying to locate the new style 3900 sensor on the engine to conduct an OBD I compatibility experiment and got a code 43 Knock sensor /ESC error.

I have a new sensor to install so hopefully this will cure the problem, as I mentioned previously I did bump the sensor and wire while reinstalling the turbo just before the problem surfaced but never thought anything of it since it appeared to be harmless contact.

Hopefully the late model sensor will work properly in place of the stock sensor also as it is located higher up near the engine deck and attaches to the block like a banjo fitting.

IIRC you are using a turbo, running 7 psi of boost with high compression and you expect to run succcessfully, detonation free using 89 octane gasoline. I say this with all respect for your project-its just not going to happen unless you are prepared to re-engineer the combustion chamber shape and piston top design.
My solution is get that compression ratio down to 8.5:1 and use 93 octane gasoline. Even so with 7-10 psi of boost you will be running an effective compression ratio of 12.5:1- 13.5:1.
I know a guy a few years back who built a 3.4L engine and ran 10.5:1 compression with 8 psi of boost. He would not listen to the advise that everyone was giving him but even running premium gas with conservative timing, he blew a hole in a piston on the first run.
You may not like the answer in this post and I am certainly not trying to be rude, but thats the way that I see it.

------------------
" THE BLACK PARALYZER" -87GT 3800SC Series III engine, ZZP Intercooler, 3.4" Pulley, N* TB, LS1 MAF, Flotech Exhaust Autolite 104's Custom CAI 4T65eHD w. custom axles, HP Tuners VCM Suite.
"THE COLUSSUS"
87GT - ALL OUT 3.4L Turbocharged engine, Garrett Hybrid Turbo, MSD ign., modified TH125H
" ON THE LOOSE WITHOUT THE JUICE "

quote Originally posted by Dennis LaGrua: IIRC you are using a turbo, running 7 psi of boost with high compression and you expect to run succcessfully, detonation free using 89 octane gasoline. I say this with all respect for your project-its just not going to happen unless you are prepared to re-engineer the combustion chamber shape and piston top design. My solution is get that compression ratio down to 8.5:1 and use 93 octane gasoline. Even so with 7-10 psi of boost you will be running an effective compression ratio of 12.5:1- 13.5:1. I know a guy a few years back who built a 3.4L engine and ran 10.5:1 compression with 8 psi of boost. He would not listen to the advise that everyone was giving him but even running premium gas with conservative timing, he blew a hole in a piston on the first run. You may not like the answer in this post and I am certainly not trying to be rude, but thats the way that I see it.

Very tactfully put and I appreciate and welcome it. You'd have to read my first post to see that much of what you have touched on has already been proven a non issue. The high compression engine is not in the car at the moment due to an incident unrelated to octane choice leading to its removal.

This is the stock motor except for the cam, that I have already proven at the beginning of this swap runs just fine with 7 psi, 89 octane and 212 deg inlet temps. It has a revised combustion/quench combo compared to the old Turbo Grand Prix 3.1 and comes from the factory with 9.8:1 compression and an 87 octane fuel recommendation per GM's drivetrain site. 8.5:1 on an iron head motor is fine but on the aluminum head engines it's a move back in time and there is a littany of aluminum head motors on the various associated forums running their stock 9+:1 compression ratio with more than 10 psi and kicking butt all the way down into the low 12s and high 11s qtr.

In addition to effective compression, lets not forget dynamic compression directly affected by camshaft duration and timing. I have some VVT function capability that I've already calculated will drop my dynamic compression by ~.5 compression points when the cam is retarded for high end performance, in addition to the increased duration above stock I had ground into it. You can increase duration to offset high compression to an extent.

Note that the problem was the sensor all along and that throwing 93 octane, alcohol and reduced timing at it didn't fix the problem which leads to the most important aspect of situations like this; finding the problem and fixing it, rather than applying a band aid, lower compression/ higher octane for example, when you know it isn't necessary despite popular oppinion.

i have been running 10:1 compression and 10 to 12 psi for the past 3 years with no problems .first it was on a bone stock 2.2 ecotec with a 16G turbo and now i run a 20G turbo on a 2.2 with wiseco pistons and eagle rods . a lot of people on the ecotec forums have broken pistons running turbos , but almost all of them decided to save money by not using water meth injection .i use both water meth (snow progressive boost sensing) and an air to water intercooler .my IAT never goes more than 10 degrees above ambient .i run premium gas , never tried it with anything else .when i ran my car before i got the intercooler , i had to put the IAT sensor on the intake pipe before the turbo because the high charge pipe temps would cause a drastic reduction in timing advance .the computer cant read the cooling effect of water meth .with my intercooler on i move the sensor back to where it was supposed to go .early on in my turbo setup i had a lot of problems that i struggled with and spent a lot of money on that turned out to be something simple and cheap - i had a bad injector harness .changed the harness , problems solved .

As for that Knock Sensor, make sure you use correct mounting torque. I have found several that were overtorqued, and not by very much either, that would create ghost signals as the engine expanded coming up to operating temp. To reiterate, mounting torque for a knock sensor is crucial!

quote Originally posted by 86FieroSEv6: As for that Knock Sensor, make sure you use correct mounting torque. I have found several that were overtorqued, and not by very much either, that would create ghost signals as the engine expanded coming up to operating temp. To reiterate, mounting torque for a knock sensor is crucial!

Will check into that, I'm just glad to find that the majority of what was logged is false timing retard as opposed to real.

So far the spark table is clean and free of timing retard.