ECMlink Trying to get rid of knock causing NEGATIVE timing, 2.4 4G64 DOHC Spyder + 14B turbo

[Kryndon]

Hey folks.

I'm trying to tune this 2G spyder with a 2.4 block and 2.0 Sonata DOHC head + stock cam gears, 1G black CAS, balance shafts, 2G knock sensor, 1G 14B turbo with a 16G wastegate actuator, 2.5" FMIC setup. ECMlink V3, Speed Density bundle, 2.5" exhaust system, screamer pipe.

So far this car is NOT behaving remotely close to other 2.0 4G63s that I've tuned. The turbo spools ridiculously fast and starts knocking from very early on. IATs are around 95F which is pretty fair for current weather. Running RON95 fuel with no additives. I've installed an alleged walbro 255 pump, rewired it so it gets a nice 13.4-13.8 volts at the pump, and R35 GTR injectors which are high Z and 550cc flow at 43 psi, adjustable regulator.

The mechanical timing is spot on, so is the CAS, 5 degrees verified with timing light. Both balance shafts are still in and working.

Simply put, the only way I could minimize but not completely eliminate the knock is by running VERY low timing to begin with. AS low as 3-4 degrees in some areas. And even then, as soon as it starts knocking it pulls another few degrees and often leaves me at -1 or even -2 degrees which at this point makes it a jet engine and I assume EGTs are going through the roof.

I've got my fueling dead on, 11.5 AFR. I don't feel comfortable running less than that. I'm starting to think this might be entirely phantom knock or a very noisy engine, as this 2.4 does sound like a real 'thumper' compared to a regular 2.0. Also I installed an OEM knock sensor from a 2G DSM. I would ASSUME a 2.4 block is thicker and would resonate at different frequencies than a 2.0, so there is a chance the 2.0 sensor is picking up false vibrations since the ECU doesn't know it's controlling a 2.4. But ECMlink is limited in that regard and nothing can be done. Also the way I ran the wiring for the knock sensor is using the purge control solenoid valve signal wire and moving it to pin 78 for knock input. Then ran a shorter ground wire from the knock sensor connector to the GM MAP sensor ground wire, since they all terminate at the ECU sensor ground.

I've checked everything, I really don't know what else to check or try. It's a street car so race gas, ethanol or meth injections are out of the question, and the possiblity of a bigger turbo right now is not possible. At best I could try to find a 14B actuator and see if it will hold less boost since right now it peaks at 15 psi, runs amazingly good but like I said the timing is ridiculously low.

I would appreciate any suggestions. Also for the record, the current timing maps are not finalized obviously and I've been building it up but if I try to load a stock 2G timing map, it just detonates as soon as you start even lightly accelerate, so I don't think running HIGHER timing would help, I already tried it's even worse.

Attached are 2 logs of 3rd gear pulls, those were the lucky logs where knock was minimal but you can see the timing being very low.

 

[biglady112]

You need to be no leaner than 11.1-11.2:1. You would be highly benefitted moving to 10.5:1. You are not losing any power but, are gaining much more cooling with the extra fuel. Whoever taught you 11.5:1 is wrong.

It does not matter what the timing map says. Read your spark plugs. They do not lie. Having very low timing is not as bad as you think it is if you are on shitty pump gas.

Move the knock sensor to the bottom of the block if you have to and shield the hole of the sensor by filling it with silicone/RTV before you screw the sensor in. And then, start reading your spark plugs.

Your timing mark needs to be at the 45* mark on the bend of the strap. And it is okay to be on the high side of the strap towards the tip. The closer you get the mark to or into the base circle, the more trouble you are going to get into. This may require you to buy quite a few spark plugs to get this right. One spark plug, one pull. As little data on the plug will give you the real story as to what is going on in there.

It is okay to check on cylinder at a time to help save on plugs. With a 2.4, a 14b and shitty pump fuel, you are likely seeing real detonation as that turbo is way too small and is in boost if you even sneeze at it.

And quite the opposite, a 2.4L is much thinner than a 2.0L, again leading to being a noisier engine. They are the same casting with one having a bigger and thinner bore. The 2.4L is just 6mm taller. We have had major problems with head gaskets on 2.4’s at big power levels. I would never use one for any reason. 2.0L all the time, every time.

 

[Kryndon]

You need to be no leaner than 11.1-11.2:1. You would be highly benefitted moving to 10.5:1. You are not losing any power but, are gaining much more cooling with the extra fuel. Whoever taught you 11.5:1 is wrong.

It does not matter what the timing map says. Read your spark plugs. They do not lie. Having very low timing is not as bad as you think it is if you are on shitty pump gas.

Move the knock sensor to the bottom of the block if you have to and shield the hole of the sensor by filling it with silicone/RTV before you screw the sensor in. And then, start reading your spark plugs.

Your timing mark needs to be at the 45* mark on the bend of the strap. And it is okay to be on the high side of the strap towards the tip. The closer you get the mark to or into the base circle, the more trouble you are going to get into. This may require you to buy quite a few spark plugs to get this right. One spark plug, one pull. As little data on the plug will give you the real story as to what is going on in there.

It is okay to check on cylinder at a time to help save on plugs. With a 2.4, a 14b and shitty pump fuel, you are likely seeing real detonation as that turbo is way too small and is in boost if you even sneeze at it.

And quite the opposite, a 2.4L is much thinner than a 2.0L, again leading to being a noisier engine. They are the same casting with one having a bigger and thinner bore. The 2.4L is just 6mm taller. We have had major problems with head gaskets on 2.4’s at big power levels. I would never use one for any reason. 2.0L all the time, every time.

Appreciate the insight.

It's not that I was 'taught' to aim for specific values. I just prefer to use them as basic guides when building up a map from scratch and if the engine likes it, I tune for it. Of course sometimes you just gotta give it what it needs. In this case you might be right because I did notice if I richened it up a lot more, down to 10.4-10.5, the extra fuel did help cool the charge leading to less knock. But that in its own produces more boost sooner which puts me in a cycle of taking out even more timing from earlier on. Our European RON95 pump gas is equivalent roughly to 91 in the States. I rather tune it on that fuel and know the car can be fueled up anywhere with the shittiest fuel and still be safe, rather than rely on race fuels or injections just to make a few more horsepower.

I have actually been taking the plugs out in to inspect. In fact I regapped them (BCPR6ES) to 0.027" and they did not have the distinct traces of knock on the electrodes or on the tip.

Never heard of filling the hole with silicone. I understand what the aim is, to fill the remaining air cavity so it doesn't resonate excessively but have you personally had success with this mod? Also I have attached a picture of a 2.4 block from that side. I currently have the 2G knock sensor mounted in hole 1. Moving it lower down might make it less sensitive, or at least allow us to ignore insignificant knock or phantom knock. Which of all 4 locations have you tried with good results?

You must be logged in to view this image or video.

 

[biglady112]

Appreciate the insight.

It's not that I was 'taught' to aim for specific values. I just prefer to use them as basic guides when building up a map from scratch and if the engine likes it, I tune for it. Of course sometimes you just gotta give it what it needs. In this case you might be right because I did notice if I richened it up a lot more, down to 10.4-10.5, the extra fuel did help cool the charge leading to less knock. But that in its own produces more boost sooner which puts me in a cycle of taking out even more timing from earlier on. Our European RON95 pump gas is equivalent roughly to 91 in the States. I rather tune it on that fuel and know the car can be fueled up anywhere with the shittiest fuel and still be safe, rather than rely on race fuels or injections just to make a few more horsepower.

I have actually been taking the plugs out in to inspect. In fact I regapped them (BCPR6ES) to 0.027" and they did not have the distinct traces of knock on the electrodes or on the tip.

Never heard of filling the hole with silicone. I understand what the aim is, to fill the remaining air cavity so it doesn't resonate excessively but have you personally had success with this mod? Also I have attached a picture of a 2.4 block from that side. I currently have the 2G knock sensor mounted in hole 1. Moving it lower down might make it less sensitive, or at least allow us to ignore insignificant knock or phantom knock. Which of all 4 locations have you tried with good results?

You must be logged in to view this image or video.

As low as the harness can reach.

I am not talking about looking at and inspecting for a look. I am talking about where the timing mark falls on the strap after a run. Not multiple runs. I also would never run .027” gap. I set all of them up in the .010-.016” gap because everyone is too cheap to put a real ignition on their car and I run out of ignition quickly when tuning them.

I didn’t tell you about silicone for fun. I have been doing it on stock ecu cars with shitty worthless knock sensors for over 20 years.

If you are tuning you global air/fuel ratio to a standard number, you are doing it wrong. Each engine and combination is different. Tune for what makes it happy. I run E85 customer cars where your car is on pump gas. Put fuel in it and stop wasting your time.

Running it richer is not going to noticeably make that 14b spool faster. You have way more cubic inches than turbo. It is a poor combination and you need to adjust accordingly.

You need to shield/move the knock sensor, add fuel and do some research on how to read timing on a spark plug. It sounds like you are making this harder than it needs to be with a car that only has a 14b on it.

 

[Kryndon]

As low as the harness can reach.

I am not talking about looking at and inspecting for a look. I am talking about where the timing mark falls on the strap after a run. Not multiple runs. I also would never run .027” gap. I set all of them up in the .010-.016” gap because everyone is too cheap to put a real ignition on their car and I run out of ignition quickly when tuning them.

I didn’t tell you about silicone for fun. I have been doing it on stock ecu cars with shitty worthless knock sensors for over 20 years.

If you are tuning you global air/fuel ratio to a standard number, you are doing it wrong. Each engine and combination is different. Tune for what makes it happy. I run E85 customer cars where your car is on pump gas. Put fuel in it and stop wasting your time.

Running it richer is not going to noticeably make that 14b spool faster. You have way more cubic inches than turbo. It is a poor combination and you need to adjust accordingly.

You need to shield/move the knock sensor, add fuel and do some research on how to read timing on a spark plug. It sounds like you are making this harder than it needs to be with a car that only has a 14b on it.

Thank you so much for the suggestions!

Today was a very productive day and I can confidently say the knock situation was taken under control with a combination of fixes.

Firstly I installed new BCPR7ES plugs, 1 step colder than the previous. Gapped them to 0.020" which lead to some occasional misfires on idle but that's not a deal breaker considering we don't get spark blowout now during boost and the electrode doesn't get as hot.

I also installed a true 14b wastegate actuator since I was using one from a 16G that opens at 11 psi, the 14b now opens at 7 psi and peaks at 12, which is perfect for this motor.

And I richened up the AFRs throughout the map, blending many of the cells much more nicely and target AFRs are 11.2 during boost. It helped a ton!

I also blended in more the timing map because I hadn't spent too much time on it. Now, during a 3rd gear pull from 2000 to 7000, timing starts at around 8* advance, plateaus to about 5* at peak load, then rises back up to 9*. This is WAY better than the -3* I was getting before, due to knock. Yeah the car isn't who knows how fast but what matters is it's now very happy and safe. Getting about 26-27lbs of flow.

Interestingly, I checked the old plugs and they were all identical and looked.... very good. The center porcelain was flat white and only the base was slightly black in the rich area. The elctrode tip was marked very nicely at the 45 degree. So actually based on the old plugs it would appear the motor was running good fueling and timing, but ecmlink and knock were saying otherwise, even though my but was definitely feeling good power.

Still, Id rather be safe and sacrifice some HP than to risk it. I do appreciate all your tips, Biglady112!