IW STi Forum banner

1 - 20 of 55 Posts

·
Registered
Joined
·
34 Posts
Discussion Starter #1
Has anyone purchased a cooling kit for Cylinder 4 chamber? Seems to be a clever idea as it offers a solution to a known engine issue due to higher detonation risk in #4 chamber, and the install seems straightforward too. I am looking for feedback from early adopters.

I followed several online discussions where people seem to agree that it might be beneficial, but there is an ongoing discussion as to what should be the inside diameter of the cooling hose.


If you have the Dom's kit, what is the inside diameter of the plug that screws into the left head?

If you assembled your own cooling kit, what inside diameter do you have? How/why did you choose it?

Any quantifiable benefits that you could present? tuning impact, etc.

------------------------------------------------
Background info
http://shop.getadomtune.com/categories/cooling.html



Cylinder 4 Chamber Cooling System GR/VA/08+ LGT Fitment

Description
Industry first DOHC EJ Cylinder 4 Combustion Chamber Cooling kit increases coolant flow in the LH head to improve reliability by reducing detonation causing hot spots, reducing EGT delta and improving overall cooling system performance.

This is for all 08+ Impreza and Legacy which require a longer hose than pre 08 cars.

Each kit contains:

1x Coolant pipe assembled with a stainless steel T, high grade silicon hose, Stainless Pinch clamps, Custom machined and anodized 6061 fitting.
 

·
Premium Member
Joined
·
5,389 Posts
See YamahaSHOS build thread

Question: isnt the knock sensor located at cylnder #4 becsuse it runs hot snd is ghe first to detonate? Something seems wrong with that idea as even my old GD has roughness for each cylinder. What is roughness? But, if one sensor, then getting that cylinder to run cooler will help prevent detonation in that cylinder, but make it more likely on all the others?
 

·
Registered
Joined
·
352 Posts
I also am curious as to actual numbers...and mheyman, it wouldn't cause the others to heat up more as more fluid would be traveling throughout the block thereby allowing the radiator to cool off al lcylinders...if its overheating then theres another cooling system issue. IIRC number 4 is the only one that doesn't have complete flow through it like the other cylinders. I don't know much about thermodynamics or plumbing but i feel that a hose the same size as the heater hose would be sufficient to get fluid moving around to appropriately cool the cylinder...
 

·
Registered
Joined
·
594 Posts
I got DOM’s cylinder 4 cooling kit, and while I do not have actual EGT data I can confirm what Dom said that the coolant tempt is down by average 5 Deg C.

Another interesting thing that I felt is my STI revs nicer ever since I added this cooling mod. I don’t have an explanation as to why, but it ‘could’ be due to the now equal EGT resuling better exhaust pulse entering my twin scroll turbo.

Based on the figure posted by Dom, if I convert into Deg C the EGT for all 4 cylinders goes as below:


The before and after figure are of course taken on different day and time so the numbers are not relevant. What is interesting is how the EGT across all 4 cylinders are basically equal to each other. Again these are not my car’s figure and I do not have a mean to take measurement unless I install EGT sensors.
 

·
Registered
Joined
·
118 Posts
See YamahaSHOS build thread

Question: isnt the knock sensor located at cylnder #4 becsuse it runs hot snd is ghe first to detonate? Something seems wrong with that idea as even my old GD has roughness for each cylinder. What is roughness? But, if one sensor, then getting that cylinder to run cooler will help prevent detonation in that cylinder, but make it more likely on all the others?
To answer that, it's likely good to know how the Subaru ECU determines knock. I.e.: how does it know there's a P0301 and not P0302 or 303 or 304? Easy. The crank sensor. The crank sprocket has teeth that trip the sensor. There's a section that doesn't have teeth that tell the ECU where the crank is in the rotation and it basically keeps track which cylinder it fired first and continues to go through the firing order. So when the knock sensor picks up knock, it knows which cylinder it happened in simply from from the timing of it all.

The knock sensor very well is placed over cyl 4 due to its ability to knock due to heat. Knock is caused by the air/fuel igniting before the spark plug has that chance. Retarding the timing is one way to reduce/avoid knock. Knock is inevitable because the ECU is constantly advancing until it senses knock and then retarding a little until it find that fine line where it can give you the most timing without a major knock event, but it's still trying to give you the most so it's still tryna advance just a smidge before the knock sensor says, "nawp, too much" and calms the ignition system down a bit.

Reducing the likelihood of pre-ignition due to heat on cyl 4 will simply 'level the playing field' if you will. If I remember reading this correctly, supposedly Dom didn't make it so it would be less prone to knock than the other cylinders, but simply reduced it by a fair margin. I think it is still more so prone to knock than the other 3, just not near as badly or as soon as it used to.

Sent from my SM-G955W using Tapatalk
 

·
Super Moderator
Joined
·
8,427 Posts
On a proper tune, the ECU is running the timing maps... Not constantly advancing the timing. When no significant knock is detected, it runs the base map + (knock advance map * rough multiplier) + compensations. Additionally, I recommend removing per-cylinder timing compensations.

Pre-ignition is not normally the issue, but knock is.
 

·
Premium Member
Joined
·
5,389 Posts
Thankyou !!! All makes sense to me.
 

·
Registered
Joined
·
118 Posts
On a proper tune, the ECU is running the timing maps... Not constantly advancing the timing. When no significant knock is detected, it runs the base map + (knock advance map * rough multiplier) + compensations. Additionally, I recommend removing per-cylinder timing compensations.



Pre-ignition is not normally the issue, but knock is.
That is correct. It'll run the max allowed timing as long as there is no major knock, so it advances to the max allowed. If knock is an issue, then it will advance-knock-retard timing. We may see it as a consistent negative number on our Accessport, but that's because this process happens very quickly and the ECU has found the sweet spot. I didn't specify that to try and keep it simple.

Also, pre-ignition is not an issue only if a major knock event doesn't result. Where there is knock, there is pre-ignition. Extreme heat (like cylinder 4 can experience) can cause pre-ignition, and therefore causes cylinder 4 to be more so prone to knock.

Sent from my SM-G955W using Tapatalk
 

·
Super Moderator
Joined
·
8,427 Posts
It's not an advance, it's what the map calls for. Keeping it simple is it running the load/RPM maps that are input.

Pre-ignition is hardly the failure with cylinder 4 due to the fact they crack, not melt. Pre-ignition is also something the knock sensor will not have a chance at correcting.

Knock =/= pre-ignition!
 

·
Registered
Joined
·
118 Posts
Right. I was combining detonation and pre-ignition....my bad.

Pre-ignition still makes some interesting noises, though.

Either way, timing is manipulated as one way to control knock from either event.

Sent from my SM-G955W using Tapatalk
 

·
Registered
Joined
·
39 Posts
Just to clarify for anyone that comes across this thread...
Pre Ignition= a combustion event prior to the spark plug firing, can be caused by hot spots in the cylinder or the incorrect heat range plug.
Knock or detonation= A secondary ignition event after the spark plug fires that causes competing flame fronts.

The danger of both is an unwanted cyl pressure increases and the resulting heat generated.

Both are bad, Pre ignition is generally uncontrollable by the ECU's knock logic unless the knock from the previous combustion cycle is pushing the cyl temps over the top.
 

·
Registered
Joined
·
34 Posts
Discussion Starter #13
I got DOM’s cylinder 4 cooling kit, and while I do not have actual EGT data I can confirm what Dom said that the coolant tempt is down by average 5 Deg C.

Another interesting thing that I felt is my STI revs nicer ever since I added this cooling mod ...
Do you know what is the inside diameter in the Dom's kit?
 

·
Registered
Joined
·
594 Posts
You’re right 25mm is too big.

That being said, the internal diameter of the hose must be the same size of the oem heater hose, because the T joint is is fitted into it.

 

·
Registered
Joined
·
118 Posts
We are fixing to find out the ID of the "special" metal fitting that threads into the head. As someone mentioned on another forum, that is the secret sauce of this whole kit and apparently very top secret.

I imagine OP reposted on this forum to seek the answers a lot of people are wondering because comparisons were made and a lot of chaos ensued (that I too kinda got involved in, I won't lie).....and that's all I'll say to avoid another forum war.

Sent from my SM-G955W using Tapatalk
 

·
Registered
Joined
·
594 Posts
Oh, you mean over at NASIOC? ;D

Hahaha.

Why not just get one? I know it is pricy for what seemingly just a hose and a plug but I appreciate Dom's effort in doing the R&D.

Ok I better stop.
 
1 - 20 of 55 Posts
Top