High part throttle KR, load <1, boost <1-2psi, AFRs high 14s

[BAT-man]

Setup:
100K original engine gen2 ms3, K04, TR8 FMIC, 3.5" TIP, cattless 3" DP+full 3" exoss, Autotech HPFP, 180psi compression on all 4 cyls, self-tuned, re-tuning for this setup using my old maps from a very similar setup.

Driving around fine-tuning part throttle MAFcal, I am seeing some very odd high KR, from 1 to 8.
The MAFcal is almost done, I am in the 3-4% LTFTs in either direction, so it's not like i'm seeing tons of un/over-metered air (that I know of).

Circumstances:
During the 2k-3k rpm range, while gently rolling onto the throttle and going from -12in/hg to say -6in/hg or almost 0 vac, so not really boosting (1-2psi by mistake but you will see it's unrelated).

Basically a gentle 3rd gear/4th gear accel, no spikes in AFRs (holding in the high 14s as i'm cruising), no spikes in load (mostly under 1, mostly in the .5-.8 range).
KR will jump to 8ish and then drop off as rpms rise, while my pedal position is constant (trying to ensure i'm not actively accelerating during datalogging this).
It's very consistent in terms of load, pedal position, AFRs, etc (I am attaching two separate logs and pics from them, KR is brown, load is green).

Never happened to my old gen2, and I can't think of a logical reason this should happen, there is no KR when I log WOT, but this part throttle light acceleration KR is kind of crazy. Is this the pre-detonation stuff they used to say is the death of all DISIs? Is it injector seals (
Any tips or pointers are much appreciated.

EDIT: I have recently given up on fixing a P0101 CEL which has to do with the MAF signal. I have both cleaned and replaced the MAF, and wiggling the harness doesn't seem to have an effect on airflow when monitored on the AP. Not sure if that helps.

Also, even though I cleaned the crap out of the EGR system (from intake mani to the pipe to the block attachment that the coolant runs through), and replaced the solenoid, the EGR CEL P0401 will come on every other week or so, after being reset (so more than a dozen drive cycles). I know EGR supposedly helps with lower cruising combustion temps, so it sort of makes me wonder about that too.

 

[SharksInSpace]

Part throttle KR is a non-issue. It's supposedly something to do with ECU logic attempting to eek out better gas mileage.

 

[BAT-man]

hmmm.... i mean you can really feel it retarding the timing hard, car will stumble a bit til it gets down to 1 or 0 KR. i just never experienced or seen this in like 100k mi of logging and tuning my other car. it feels like something is off.

 

[SharksInSpace]

Hmm if you can actually feel it pulling power then maybe that is something. I've never felt it do that but I know KR at cruise in my juan used to hit up to the 7s. I haven't logged to see what it does at cruise in my pu, and can't live monitor since Torque doesn't have that PID and I don't want to be trying to look at a laptop screen while driving (VT tuned not AP).

 

[The_big_dill]

I feel your frustration BAT-man, this is something I have been living with on my gen1.

Everyone says its normal, but a lot of people report different degrees of timing retard during cruise (some 4 KR, some 8). I am very interested to hear about this especially because you had a gen 2 you self tuned that didn't experience this problem, and now you do.

The reason its annoying to me, is when I am cruising and I want to step on it for power, I have to let off the throttle, let KR go to 0, and then punch it...

There is a decay control in ATR that you can use to kill KR quicker, but care needs to be taken not to reduce sensitivity too much.

 

[JohnnyTightlips]

I feel your frustration BAT-man, this is something I have been living with on my gen1.

Everyone says its normal, but a lot of people report different degrees of timing retard during cruise (some 4 KR, some 8). I am very interested to hear about this especially because you had a gen 2 you self tuned that didn't experience this problem, and now you do.

The reason its annoying to me, is when I am cruising and I want to step on it for power, I have to let off the throttle, let KR go to 0, and then punch it...

There is a decay control in ATR that you can use to kill KR quicker, but care needs to be taken not to reduce sensitivity too much.

Got a log?

 

[BAT-man]

Yup, it's the excel file in the OP, I converted from CSV so I can have the tables in there.... and now I realize you are prolly asking The_Big_Dill for a log, so nm.

 

[Sho]

Looks very similar to the kr both my current ms6 and my old gen 1 had at part throttle. Goes away instantly if you lift or jab the throttle in a bit more. Otherwise it'll hold steady or decay if you keep constant throttle. Could basically get it to read 5-8kr at part throttle on command if I wanted to based on pedal position and which gear I was in

My solution, made the knock sensor inactive below 3k rpm lol.

 

[BAT-man]

bah. Sho, i totally get where you're coming from. i know that the gen2 supposedly got rid of either the sensitivity or algorithm to sense knock under 2.5-3k rpm that was present in gen1/ms6, so i just wonder how it's even possible on this 2012 model, and if it's
a) something stupid, like a very specific resonance that happens in my engine/bolt-ons during 1.5k-2.5k rpm under certain loads, or
b) an actual issue like previous owner using VTA and super sooting/carboning the combustion chamber/valves so PT knock/pre-detonation actually happens under the given conditions, or the MAF/MAP being off afterall, or something else.

i'll keep adding to the OP, as I do this mod or that, and eventually gett o cleanign the valves and replacing the injector seals, to let you folks know what (if anything) made this go away. in the meantime, again, any input or ideas are very welcome, and i appreciate the ones above.

 

[Sho]

Are the injector seals orginal? And do you ever get a popcorn popping type noise at higher rpm under load?

 

[BAT-man]

Yes they are original Mazda injectors seals, 110k miles old, and no, no popping noise as far as I can tell. My exhaust is pretty quiet, so I can usually even hear knock when it happens, but not a popping noise under WOT. Now that I'm thinking you probably meant not quite WOT but load ranging between 1-2, I'll try to replicate it and listen for one.

 

[Sho]

Eh youd probably hear it at wot as well. But I've had a few locals with severely blown/damaged oem injector seals and it'll sound like popcorn popping in the microwave under load above 5k usually and plenty of associated kr to go along with it

 

[SharksInSpace]

I can try to grab some cruising part throttle logs in my pu this week if you think it might help to have a comparison.

 

[The_big_dill]

Got a log?

I scaled APP 1/10 and RPM 1/1000 so its readable. Starting with cruising KR (22%), climb to around 70% APP as knock slowly decays. Let off the gas, KR=0, punch it, KR still 0.
NOTE: Although I only get to ~70% APP, the behaviour is the same at 100%. You can feel this in the power band.

 

[BAT-man]

Huh... yours stay even when you boost a little, and decays, mine goes away as soon as I start boosting, otherwise it's very similar.

 

[The_big_dill]

Huh... yours stay even when you boost a little, and decays, mine goes away as soon as I start boosting, otherwise it's very similar.

Correct. If I kept the pedal planted while it was decaying KR, it would continue at that decay until it hit 0. You can definitely feel this in the power.

I did reduce the decay rate quite a bit in ATR, but am not 100% sure of the consequences if I set the number too low... I imagine worse fuel economy and not enough KR when the engine needs it.

 

[BAT-man]

My decay is 300 right now, but I'm going to make it 150, just to see where the real KR is isolated more.
The way I understand it, if it's true knock that it senses, it will add KR, but if it stops sensing knock it will keep taking it away (or adding timing in other words) quicker with lower decay rates.

Still unsure what the Knock Multiplier table does (and why it's 2.0 everywhere except a few rpm columns), and whether the Knock Offset table just reduces any given KR by the offset amount or plays a multiplication role somewhere. No descriptions from COBB or any tuners, besides "leave as is."

 

[The_big_dill]

My decay is 300 right now, but I'm going to make it 150, just to see where the real KR is isolated more.
The way I understand it, if it's true knock that it senses, it will add KR, but if it stops sensing knock it will keep taking it away (or adding timing in other words) quicker with lower decay rates.

Still unsure what the Knock Multiplier table does (and why it's 2.0 everywhere except a few rpm columns), and whether the Knock Offset table just reduces any given KR by the offset amount or plays a multiplication role somewhere. No descriptions from COBB or any tuners, besides "leave as is."

With a lower decay, I found that the system is not as consistent when cruising KR is present. Meaning it would normally stay steady at 8, but with a lower decay it would hover around 4-8. The only problem with that is fuel economy, from my understanding.

I do believe that there should be a database of what all the tables do, even if COBB didn't provide it. I am rather new to the platform, so I am not sure who may bring info on these unidentified tables?

 

[BAT-man]

Presumably professional/long standing self-tuners. But they are hard to get a hold of. It's fairly frustrating that COBB doesn't even attempt to explain some of the tables. I would take it as ignorance on their part, but it's most likely just lack of interest in the platform, outside of supporting tuners that pay them. Separate topic, didn't meant to get on a soapbox.

 

[BAT-man]

Still working on isolating the KR. Even with the lower decay, it's still hard to pinpoint. However I did open up my Max KR to 15 (was 8 before) and low and behold, my highest KR values are now up to 15. So I will try to reduce the timing map areas where I see this happening as those are relatively low load and low RPM areas, offer no performance gains, and are where I am exceeding the OEM values. Perhaps it's my tune, but I never had this type of KR show up in my other car and if anything the FMIC makes the current setup run cooler than the stock TMIC. More to come...