Audible knock with new A/W intercooler

[A.J.]

So I dyno tuned my SRT-4 Caravan with my Toyota Supra intercooler, bumped up the timing, gained some horsepower and torque and didn't hear any knock. Drove it, everything was fine. I didn't bring my scan tool with me because I didn't plan on tuning so I don't know if the computer was seeing knock or not.

Now all I did was swap out my Toyota Supra i.c. with 7-90* bends and 12' of 2" i.c. plumbing for an air to water i.c. with about 3' of 2.5" plumbing and now I can hear knock. It happens sometimes around 5 psi of boost but always between 10-15 psi of boost. My a/f ratios are still in the 11's (low 11's with the Supra i.c. and high 11's with the a/w i.c.) and with my new before/after i.c. temp gauge has the intake temp about 120*-130* on a 95ish degree day (up to 220*-230* going into the i.c.). Even in the morning when the ambient air temp is in the 70's my intake temps are in the 80's I'm still getting knock.

On a side note, I have my check engine light flash with knock. My light is always flickering, even when the scan tool is not registering knock. A turbo Volvo friend of mine came up with a good suggestion. Since I'm using a knock sensor and a computer that was meant to sense knock through a cast aluminum intake manifold, through a gasket, through a cast aluminum head, that's where my knock sensor on my 2.4 ought to be. There is a spot in between #2 and #3 intake runners where I can drill, tap, grind flat, and mount my knock sensor. I realize that this isn't going to help my audible knock problem but it might help keep the light from flickering or registering knock when there is none.

[cordes]

Are you pulling the boost signal from the manifold? It seems pretty obvious that you're more efficient now. Perhaps try to up the fuel to the previous AFR and see if the knock goes away?

Also, that's a pretty darn good delta T with the new intercooler.

[A.J.]

Are you pulling the boost signal from the manifold? It seems pretty obvious that you're more efficient now. Perhaps try to up the fuel to the previous AFR and see if the knock goes away?

I was pulling the boost gauge signal from behind the throttle body but moved it to the manifold that supplies the MAP and FPR. I haven't driven it since. It's parked in the back of my shop so I can pull the manifold for my knock sensor relocation so it's out of the way so I can still get work done during the day.

[cordes]

I was pulling the boost gauge signal from behind the throttle body but moved it to the manifold that supplies the MAP and FPR. I haven't driven it since. It's parked in the back of my shop so I can pull the manifold for my knock sensor relocation so it's out of the way so I can still get work done during the day.

Excellent so you know that there isn't anything funny going on with the boost readings before and after the cooler.

Have you read any of bucar's posts regarding knock control at all? I'm not sure that moving it around will do you too much. I would think that you would get the best feedback from the factory location for that motor. Have you tried logging the noise that you're getting to adjust the table?

[A.J.]

Have you read any of bucar's posts regarding knock control at all? I'm not sure that moving it around will do you too much. I would think that you would get the best feedback from the factory location for that motor. Have you tried logging the noise that you're getting to adjust the table?

I have, but it just makes sense to mount it in the intake manifold the way the 2.2/2.5 has it. I'm thinking in the center of the cast iron block is just to sensitive.

Logging that noise how, through the scan tool or a listening device? I have a Bosch knock sensor that I put a headphone jack on to listen to knock. My 65 lb per hr injectors are really loud and that's mostly what it picks up. I had a stereo that had a headphone jack in it but I put that back in my Dakota. I need to get either a dedicated speaker or set of head phones to listen. Zin awhile back posted up a listening device that has an alligator clip that you can mount anywhere, you don't have to bolt it to the engine the way I have it now. That's something else I need to look into.

One other thing, I just discovered the spread sheet calculated timing table in MPTune through a thread that Simon started asking a question. I need to find an SRT-4 timing table (I think a found a Megasquirt one but don't know how good it is) and compare mine to it. Also, I can data log with my Snap-on scan tool. I'm thinking about asking a good customer of mine to borrow his turbo PT Cruiser and do some data logging and possibly make my own timing table that way.

It's just frustrating to install a major improvement over what I had and get a poor result. All I wanted was to get this on, do some dyno-tuning to get my new torque curve so I can get a proper torque converter made, and now I can't because I have this knock issue. I don't want to blow $100 an hour on the dyno if can figure it out on my own and save some someone else's time and my money.

[wheming]

Did a battery reset do anything?
Based on the summery of your change, your plumbing setup is night and day less restrictive than the old with the 2" pipes.
But I would think if anything, that would manifest itself as a lean condition.

[cordes]

Log the knock table through the RPM range and that'll give you the noise the motor makes. The SDAC 18 tech session was great for info like this. IIRC They suggested to log it with virtually no load just going down the road, and then add 10% to that value to make your threshold.

[A.J.]

Did a battery reset do anything?

No.

Log the knock table through the RPM range and that'll give you the noise the motor makes. The SDAC 18 tech session was great for info like this. IIRC They suggested to log it with virtually no load just going down the road, and then add 10% to that value to make your threshold.

I'm not sure what you mean there or how to do that.

I'll give this new knock sensor placement a try. If it doesn't work I can always put it back in the other hole.

[cordes]

The sensor is always picking up some sort of noise. There is the curve to the knock table which dictates when the computer actually pulls timing due to knock. If you record the knock value at 500RPM increments you can see how much noise there is all the time. Then you can use those values to make a proper knock table curve for your specific motor.

[Force Fed Mopar]

The KnockVoltageFromRpm table is what you want to log. The stock setting is very conservative.

I am pulling a good bit of timing on my 2.5 w/ the AWIC setup. However, it likes it, definitely pulls harder with less timing in boost. I'm down to 1* total timing above ~16 psi in the mid-range rpms. I have tried several ways of raising the timing, and it just doesn't want it, always pulls better on less.

[A.J.]

The KnockVoltageFromRpm table is what you want to log. The stock setting is very conservative.

I am pulling a good bit of timing on my 2.5 w/ the AWIC setup. However, it likes it, definitely pulls harder with less timing in boost. I'm down to 1* total timing above ~16 psi in the mid-range rpms. I have tried several ways of raising the timing, and it just doesn't want it, always pulls better on less.

As far as knock goes, the only PIDS I can log with my Snap-on scan tool is Knock Voltage, Cylinder #1 retard, Cylinder #2 retard, Cylinder #3 retard, and Cylinder #4 retard. Is there some other method to log what you are talking about?

That's pretty amazing that you have to pull that much timing out switching to an A/W intercooler. Should I start by pulling timing from the RPM spark advance table or the Part/Full MAP tables? Also is there a way to add points in MPTune? When I was at the dyno, the dyno operator wanted me to add/pull timing at certain points but there were no points there to adjust so I had to move them over.

[cordes]

log the knock volts at given RPMs throughout the range. You can add points in MPtune. I think you right click on the line to do it? It may have changed since I've made a cal though.

[Force Fed Mopar]

Use MPScan to log with, you can log any table you want with it. To add points in MPTune, click on a point, then right-click and it will give you an option to either delete that point or add another. To add, count how many points are before the place you want to add one, then type in the next number. IE if there are 4 points in front of the spot you want to change it at, type in 5.

I would use the MAP tables to pull timing mostly. Although I have been playing with the rpm table a little to try and make it less knock prone when it spools, which on my engine is probably pretty close to the torque peak. I have seen tunes on other vehicles where they richened it up slightly around the torque peak to achieve the same goal. Also I've seen where some people get better results on slightly richer AFR's all round then what is normally accepted as optimum. The instance that comes to mind is the Evo guys, who I have heard like to run around 10.8. However, that may be at higher boost levels.

So long story short, don't be afraid to play with things around a bit.

[A.J.]

Use MPScan to log with, you can log any table you want with it.

I'm going to have to do some reading/research then. I haven't messed with MPScan yet. Thanks for the help.

[A.J.]

So I opened up MPTune to see about adding points and it doesn't give me the option to insert points, only delete.



???????????????????????????????????

Also I road tested my van today with the relocated knock sensor. My light comes on a lot less than it did when it was mounted in the block. I haven't compared it to the scan tool readings but when it did come on I could hear knock. So now I'm more apt to believing it.

[zin]

Sounds like the block is a "higher fedelity" location, but that the code to distinguish between noise and knock isn't "smart" enough to work effectively in it's stock location ... or the knock voltage curve needs to be completely re-mapped.

Mike

[tryingbe]

So I opened up MPTune to see about adding points and it doesn't give me the option to insert points, only delete.

You have to add the points before you compile the file. Once the file is in its bin form, you will not be able to add points.

[Force Fed Mopar]

You have to edit the .tpl file.

[A.J.]

Ugh, you guys are killing me. So if I want to add a point to an already modified cal, I have to add points to the .tpl file, compile it, then compare my old cal to it and move everything where it's suppose to be? That's a lot of work.

[A.J.]

Okay, I want to start with part throttle timing from MAP. Here is my scale. I'm knocking bad between 10-15 psi of boost. Point #5 is 10 psi and point #6 is 15 psi. Do I go up or down to pull some timing. I'm still new at this.