I'm having a hard time finding the cause of my spark retard issue. its a 2500hd 6.0, tsp cam and heads (216/220) and a S475 turbo. Im getting knock at 1600-2000 rpm from 80-100 kpa. Same thing in the 3000-3500 rpm range on occasion. I have tried removing spark(2-4 degrees). I also tried desensitizing the knock sensors. I do get a lean spike (about 1 point afr) when i roll into it then it knocks. My p59 ecm doesn't have the transient fuel tables to adjust the lean spike. Any advice appreciated. VE table is within 1-2% commanded.

 

Took a look at your log. Just have a few questions I suppose. Are you using a wide band? I didn't see the input on the PID list. I didn't see dynamic air either, which is something else I would add to that list, especially if you have knock.

Is it bad enough to hear it? It could be worse than just 8 considering that's as much as it can pull outside of PE mode on your Max Knock Table (non-PE). I also noticed you have your boost enrichment table set to 1.0. 1.28 across that table would put you under 11.5:1 AFR commanded, but you'll have to build that into your VE.

Your timing doesn't seem too aggressive, what kind of fuel are you running?

Does the knock get better when you pull timing out?

I would look at everything mechanically and check for clearances in things like exhaust, engine mounts, and so forth. Knock sensors detect 6400 hz or so noises, so anything metal on metal can set them off. Noisy valve trains have been known to set them off too.

If everything checks out clearance wise, as in nothing is making metal on metal noises, then I would add a tank of race gas and tune the sensors while keeping the timing conservative.
It could also be the lean spike to blame perhaps, and that is something I would look into tuning out. Increasing your min transient to something higher than .015 might help. A smoother VE table would probably help too.

And another thing you could do is delay your EOIT a few degrees, that tends to richen it up a little too. Don't take that too far though, go a few degrees at a time (adding .05 to .10 to normal table) and see if your spikes go away.

That's a few things you can try. Sorry it was a lot of things, but with these things it's not always a simple answer.

 

Forgot to ask, how much boost are you pushing into this thing? Do you know your dynamic compression ratio? Might be as simple as that come to think. Limit for 93 octane is about 8.61:1 DCR.

Calculating boost compression:

((DCR or SCR / 14.7) + 1) * boost PSI = compression ratio.

Example:

7.5 / 14.7 = 0.51 + 1 = 1.51 * 8 PSI boost = 12:1 compression ratio.

Normally that might cause problems on pump gas, but I may be wrong. 7.5 DCR would be kind of high though for a boost motor. Definitely worth researching it to see what you can get away with. NA is so much easier lol. It's either above 8.61 or it's not. Mine is at 8.2 and stays at 8.2. I also have a 12:1 Static ratio, but that is irrelevant. DCR is all that matters. Can't compress anything until the intake closes.

 

Seems like I saw that ve table somewhere else. If you turn on your stft's, you can see how far off the fueling is where you have knock.. Your cylinder volume is for a smaller engine, should be 0.70819 for a 6 litre.

 

Thanks chopper, I will give that a try. The pushrods were a little bit long. It has 1 1/4 turns of preload. I looked and nothing is loose that could be hitting. To answer the questions. Yes on the wideband hooked to a/c pressure sw. 7psi max. 93 pump gas. Cyl volume is lower so I see 100kpa at .60 g/Cyl instead of .86g/cyl. Timing is at 21-22 degrees at 100kpa. Bpost enrichment was causing a lean spike because it has a slight delay to kick in. Instead I added 7% fuel over 100kpa. It's at 11.5-11.6 afr in boost. This helped with the lean spike.

 

Okay, so after looking for this for some time, I was able to find the max Effective Compression Ratio (ECR) for aluminum heads on pump gas. Dynamic Compression is measured with a transducer, so more theoretical than anything, as it is based on MAX VE. I do know there are calculators for it, but let's just use ECR for now, which is very similar, still based on valve being closed, to TDC.

It would be pretty cool if you could give us your numbers for this, and if you don't know, then I suppose we'll need to know the stroke, bore, and IVC event from your cam card. Will also need to know how much boost you're running too. Here is a better formula for figuring out Effective Compression under boost. The one above is meant for use with Static Compression Ratio.

sqrt((boost+14.7)/14.7) * CR = ECR

sqrt = square root

boost = psi of boost

CR = static compression ratio of the motor

ECR = effective compression ratio

Example - Sqrt((14PSI+14.7) / 14.7 * 8:1 = 11.1:1 ECR <<< leaves room for more boost lol.

Anything above 12.0:1 and you're going to have issues on pump gas with detonation. Watching your log, most of your knock was happening well above 80 MAP at low RPM, some as high as 100 and still under 2000, causing me to think you've got too much compression for the fuel. Obviously the lean spike is not helping you with that. Maybe more fuel would help, but I'd be looking at your boost settings first. Seems a bit early to be into boost, at 40% TPS and 1800 RPM.

So really, before messing with the knock sensors again, make sure you have a Wide Band hooked up, and are logging it into the scanner, and turn the boost controller down a few notches until you get the fueling figured out. Not trying to be a dick, just saying that's exactly what I would do in your position right about now.

 

I was typing that before I saw your reply. Do you have a way to get the WB input in the scanner? I am aware of the A/C input trick, but now I'm assuming (again lol) that you don't have HPT Pro huh?

 

Okay, I added the math to my scanner. What formula are you using? I'm working on a way to turn this into a percent, lol.

Starting with this, adding commanded and some other math to figure it out... Taking me some serious thought given the lack of commands lol.

([7101.10] / .5) + 9.8

 

do you have burst knock disabled?

 

Okay here is the formula I came up with for this input:

100 * ((([7101.10] / .5) + 9.8) - [50121]) / [50121]

7101.10 = AC Voltage
50121 = Commanded AFR

9.8 represents any error between your gauge and what it reads in the scanner. If that's not accurate, just change that number until it matches your gauge. It's the offset pretty much. To find it, note your voltage at any particular reading, multiply it by 2 and add the difference to what it says on the gauge. That would be this value. I don't know if you've done this yet.

According to this, you're WAY lean brother. This is by PERCENT, by the way.


Obviously my number for your offset might be wrong, but only you can correct that. I can't exactly see your gauge lol. I want to say normally, you add 10 to the offset, but I saw on another forum 9.8 in use, so I went with that. Hope this is helpful at least. Not sure if you had this set up or not. But now those numbers can be applied directly to your VE table via copy and paste special now if not.

 

Yes, it's off doug. Sorry for the answer but I've been pretty intimate with this particular tune. Don't ask me why. lol.

 

He could just plug his wideband into a serial port or a usb port with a usb to serial adapter.

 

But that's too easy, and less fun lol. I'm just going off what he told me brother. I might be assuming a few things, but when I heard that it had me a little worried honestly lol. Haven't seen that for a LONG time, unless it's for something like trans or fuel pressure, etc. Been a while since I've heard of a WB being used for that. Still works though so hey...

 

It's a stock 6.0 truck motor with ported heads and a cam. 9.5 CR. It is making 1.5 psi at 2200 rpm. 7 psi max. Burst knock is disabled. I am logging that. Your math is off. My wideband operates from .5v to 4.5v. Either way my VE table is within 2% in open loop.

 

I was pretty clear with not knowing the offset. Only you will know that because you have the gauge. I don't know if you have it set up that way in the scanner, but I took some time to figure out the math for you to make it easier to use as a tuning tool. Up to you though. If your lean spike is saying 16:1 with commanded being 12.5, that's 22% off. Not 3.5. Much easier to let the scanner do this. I've never had to deal with it though personally since I have the Pro version with analog inputs. Just trying to help you with your problem.

 

Sorry, I wasn't trying to come off as a di*k. The lean spikes are off 1 -1.5 points. ((7101.10)×2.375+7.3125)-(50101)

 

At 7 PSI, in theory, you should pull approximately 25% of your spark out, at that pressure. Does it respond to spark changes? Does this reduce the knock? Of course this starts with good AFR, to rule out any lean conditions as the cause. That should be your focus.

If I were you, I would pick up a PRO unit and ditch the A/C input altogether. Then you can add your WB input to the scanner and get an accurate number. With a 5k boosted setup like yours, a few hundred bucks is like insurance to make sure you don't do any real damage. Or as mentioned earlier, add the input through a serial USB input.

That's the best advice I have for you on this. If you still want run the A/C input, play with the math, get that formula to match your AFR displayed on the gauge, in percent. I was able to find a lot of info on this from threads on this same forum from 2006. However, you simply don't hear about people doing that anymore because the scanner has come a long way since then, and you can now just add it from the gauge properly with the math already done for you.

As always, completely up to you.

 

Alright, so I used your math to make something useful for you, turning your raw AFR into a % error from commanded. Try this formula:

100 * ((([7101.10] * 2.375) + 7.3125) - [50121]) / [50121]

Then I went ahead and made a chart on the scanner, to see what the AFR is doing before your knock occurs. On your tip in throttle, your AFR gets lean. The first spot being 17% lean as the knock comes in.

Here is a snip from the log. Note the "AC WB" line on the bottom, that is the CURRENT AFR in percent, so it indicates a lean transition. If you look at the table above though, you actually look pretty decent on the VE error, overall. That doesn't exactly matter since you are obviously getting lean spikes into throttle. Look into transient fuel here. It's not getting fuel in fast enough.

 

Have to ask, is your injector data from the manufacturer? And it looks like you're using a fuel pressure regulator on the rail. If you're not then the flowrate should have a slope, increasing as the MAP increases. That lean spike is definitely causing problems as I look through your log, and happens EVERY time right before the knock shows up or just as it starts.

Also, looking at your log, your ECT got up to 225*. This isn't helping you either, and I'm positive the knock added to it. Looking at that alone tells me this is real knock, as it's adding heat as expected with detonation.

 

I don't know why I didn't look at this earlier, but your spark is very high before the knock comes in. You have 30* of spark commanded before the first knock area. This looks like you have some serious work to do on your spark tables. Simple as that, but fixing your lean spikes also remain other problems to correct, but not nearly as important as the spark. Reduce spark considerably. 30* is an insane amount when pushing boost. That's a LOT of added pressure in the cylinder. Lighting it early only increases it, as the piston is trying to compress a fireball. I'd start with pulling at LEAST 10* out, maybe more in the MID range. Don't push more than about 12* total at WOT. Get this sorted out, and you should be able to revisit it later and push for power, on a dyno, or on the track looking for Trap Speed. It's hard to say if 10* will even be enough since you have max retard set at 8*, and predictably, the PCM is pulling all of it. Set that back to 15* for max allowed spark retard while you're doing this. Then pull ALL of that out if it registers in the scanner, and then some, usually 2-3 degrees more as a safety. Good luck, hope this was helpful.