Mitsubishi 3000GT & Dodge Stealth Forum banner
1 - 14 of 14 Posts

·
Registered
Joined
·
53 Posts
Discussion Starter · #1 ·
Hey. Sorry to bother you guys, but I'm running into a brick wall right now. I've just spent the last 2.5 hours on an abandoned road trying make a single pull without my knock sensor blowing up. Build specs: 3.0 forged engine, 20t turbos at 12 psi, maft gen2 with 3.0 maf zeroed out, 1220cc fic injectors, e85 and walbro 525. On my most recent pull, I logged 13 knocksum at 3900 rpm, 10.6 afr, 0 degrees timing. Ive also done other pulls where knocksum climbed up to 28 even when at single digits for timing. I don't understand. When I was running 550ccs and 15gs with my maft on 50/50 e85, the knock sensor would barely react at 18 psi 11.5 afr and 18 degrees timing, but now its going crazy at 10 psi when everything logs fine. I have verified base timing is at 6 degrees when grounding the brown connector. Any ideas? Thanks guys.

 

·
SDSU Alumnus
Joined
·
3,998 Posts
Look at your parameters' values leading up to it - not just highlighting the peak occurrence. You always review left to right.

The graphs are almost useless without any Y axes, for reference. You can clearly see registered knock immediately as you roll into WOT. Your AFR still hovers at stoich during this event. From there, it is a domino effect of crap conditions. 10.6 AFR is a consequence of interpolated fuel maps from knock. Your actual tune is most probably too lean. This is also evident from the rough plummeting of your ignition timing.
 

·
Registered
Joined
·
53 Posts
Discussion Starter · #4 ·
Did another couple dozen logs last night. I dumped in as much fuel as I could and reduced timing to where the car feels sluggish but it's still detecting knock. Dropping fuel down to mid 10 afr and timing under 10 degrees stopped it from knocking from 2000 to 3000 rpm. But it inevitably knocks from 3000-4000 rpm no matter what. Engine still sounds healthy. I swapped to heatrange 8 non-projected plugs gapped at .022 earlier this week and I don't feel any spark blow out. When I pulled the plenum off, I did notice that the lower intake has a oily tint to it and there is a bit of carbon build up on the valves, which is strange because it only has 5000 miles. On previous engine setup, the lower intake always looked clean and the valves looked brand new. I might try disconnecting the pcv from lower intake; maybe it's sucking in oil and causing all this knock.
 

·
SDSU Alumnus
Joined
·
3,998 Posts
You could have a blow-by issue which would degrade the octane, but that would need more investigation. It could even be noise being picked up from movement and contact of parts - just look at the random knock counts even at the beginning prior to WOT... red flag.

You also mentioned base idle timing, but have you verified that commanded timing matches actual timing...? This is often overlooked especially for 1G users with a flashable ECU, and I have seen cars be wildly advanced compared to commanded. VERY important.

These cars' knock sensor do not easily pick up "knock" on ethanol, so registering anything remotely dangerous is either too late/real or false/phantom.

Judging by your newest log, you still need to continue fine-tuning... look at where your marker is and when knock resumes picking up. It tries to advance from a supposed 8 degrees to 9-10 which is seemingly too much for your car. But given the modest load and airflow numbers, you have something else going on that needs addressed. "It is never just a tune"
 
  • Like
Reactions: rez604

·
Registered
Joined
·
53 Posts
Discussion Starter · #6 ·
Yeah you were totally right about a week and half not being enough time to tune. I'm going to cancel my dyno appointment. I checked with a timing light and the timing does match whats on EvoScan. on Just gonna need to do some diagnosis work and I'll eventually find the problem. Thanks Alan.
 

·
SDSU Alumnus
Joined
·
3,998 Posts
Good call and most welcome! No need to rush lol better to play it safe and keep the experience somewhat enjoyable and less aggravating. Keep us posted here with your findings. Hopefully these hurdles are the result of something small and silly. The important thing is to stay on it and follow through 👍

-sent from my Galaxy Note 9
 
  • Like
Reactions: jabarger

·
One fix at a time
Joined
·
8,127 Posts
I'm currently chasing some "knock" on E-85 and a 20T setup. It didn't really happen consistently until I changed over to the Ashspec SDT speed density add-on for Chrome. I'm pretty sure it is phantom/false but like Alan said by the time the engine picks up knock on ethanol your engine is toast since you can run way past MBT and not knock. For example, on a different translator MAF setup I was using at one time, it did not do a direct translation to the Mitsubishi Karman Hz. As a result, no matter what changes I made in Chrome for injector size, latency dead times, maf scaling, and so forth I was always only seeing a max of 130 ecu load not 200+ that I needed. Things were working, I was lazy, and I didn't check my ignition timing. At 23-25 psi I lifted the heads on a bone stock 2G motor. No idea why it didn't grenade but I guess I had a safe tune otherwise. I could feel the misfiring events happen which was actually coolant being pushed into the cylinders. It could have been resolved by a complete ignition timing map but I decided I understood enough about how the systems interacted with Chrome to go ahead and move on.

You definitely have some serious issues going on if you had to fatten up the AFR that much and reduce the timing to those low numbers. Check your PVC system since you only need an upgraded aluminum check valve between the lower intake and front valve cover nothing else. There's likely other mechanical stuff going on as well. You should be able to run the stock ignition timing map on Chrome with as much boost as you can throw at the 20Ts on E-85 and whatever your motor will handle. I aim for high 10s at WOT higher rpms depending on how good your intercooling is since I would much rather substitute fuel to quench/cool combustion temps than run leaner for literally 2-4% increase in power. You can run up to 12:1 WOT but that's playing with fire in my opinion if you have any tuning oopsies.

Jabarger, are those graphs that you made within evoscan? Or is that different plotting software? I don't think I had seen it before.
 

·
Registered
Joined
·
53 Posts
Discussion Starter · #9 ·
Yeah, I was surprised by the amount of knock as well. When I was running a 50/50 mix of 93 and e85 with 15gs, I think the only time I saw it hit over 20 knock was at 20+ psi and it bent the stock rod so much that the crank counterweight was grinding into the piston and the wrist pin was half exposed. Didn't spin a bearing though and the machine shop couldn't believe it didn't even scratch the bore. I now know exactly how much you can bend our rods before catastrophe.

I think the block is still solid. My buddy worked at a machine shop and did all the machine work to it and I watched him do most of it myself. She feels strong when on the stock timing map, at least until the knock sensor starts messing with timing and fuel. I'll probably do a compression test just to make sure. Definitely going to disable the PCV system for now.

I'm pretty new to evoscan. I tried to find a way to make it graph logs, but I don't even know how to be honest. I bought MegaLogViewerHD for $40 and I feel like it's 110% worth it.
 

·
One fix at a time
Joined
·
8,127 Posts
You probably have a lot more issues going on then if you were fighting that much knock before and more than just a little now. The knock counts I'm fighting are 1-2 counts. It isn't much but I should not be getting them in the first place. That above situation could happen on E-85 even with so few of counts at higher boost.
 

·
Registered
Joined
·
3,210 Posts
Try straight e85 or 120octane race fuel and see if it goes away. That's a real easy test to see if it is real knock or if you are knock limited on that build you got there. I wouldn't rely on factory knock sensor on a forged internal motor to begin with. It doesn't register at all sometimes, due to the different frequencies the metal gives off when it knocks. I know this since i knocked and my sensor did not pick it up, i only knew when i had to pull heads and i seen the pin holes by the fire ring. I switched to a flat response bosch knock sensor and then i could see it happening at 14psi on pump gas. Not sure if this is an option on you're current engine management system or not. If not then maybe look into seeing if the local dyno shop has a plex knock monitor system. It will use the same bosch sensor and it will pick it up. Possibly wasting you're time right now if they do. Or damaging you're engine doing these pulls.
 

·
Registered
Joined
·
53 Posts
Discussion Starter · #12 ·
You probably have a lot more issues going on then if you were fighting that much knock before and more than just a little now. The knock counts I'm fighting are 1-2 counts. It isn't much but I should not be getting them in the first place. That above situation could happen on E-85 even with so few of counts at higher boost.
It was running great on my old setup though. I think I was just running too much boost and my fuel pump couldn't keep up, causing the blow up. Everything about my new setup is different, but my knock sensor is the same as before. It's strange.

Try straight e85 or 120octane race fuel and see if it goes away. That's a real easy test to see if it is real knock or if you are knock limited on that build you got there. I wouldn't rely on factory knock sensor on a forged internal motor to begin with. It doesn't register at all sometimes, due to the different frequencies the metal gives off when it knocks. I know this since i knocked and my sensor did not pick it up, i only knew when i had to pull heads and i seen the pin holes by the fire ring. I switched to a flat response bosch knock sensor and then i could see it happening at 14psi on pump gas. Not sure if this is an option on you're current engine management system or not. If not then maybe look into seeing if the local dyno shop has a plex knock monitor system. It will use the same bosch sensor and it will pick it up. Possibly wasting you're time right now if they do. Or damaging you're engine doing these pulls.
I have been running straight e85 on my current setup. I have heard you shouldn't rely on the knock sensor on e85. I guess I could unplug it to keep the ecu from pulling timing, but I am a little bit afraid that one injector might be bad and causing one cylinder to lean out, so I don't think I should ignore it. I should probably just get my injectors flow tested. Thanks for the bosch knock sensor recommendation. Yeah it doesn't make me feel good logging a 28 knock count on multiple runs. AFR, boost, and timing are all super conservative and I'm not even going past 4000 rpm so hopefully it hasn't hurt the engine. I guess it could be something loose that's vibrating at a certain frequency as well.
 

·
Registered
Joined
·
3,210 Posts
It was running great on my old setup though. I think I was just running too much boost and my fuel pump couldn't keep up, causing the blow up. Everything about my new setup is different, but my knock sensor is the same as before. It's strange.


I have been running straight e85 on my current setup. I have heard you shouldn't rely on the knock sensor on e85. I guess I could unplug it to keep the ecu from pulling timing, but I am a little bit afraid that one injector might be bad and causing one cylinder to lean out, so I don't think I should ignore it. I should probably just get my injectors flow tested. Thanks for the bosch knock sensor recommendation. Yeah it doesn't make me feel good logging a 28 knock count on multiple runs. AFR, boost, and timing are all super conservative and I'm not even going past 4000 rpm so hopefully it hasn't hurt the engine. I guess it could be something loose that's vibrating at a certain frequency as well.
Oh yes i forgot about a injector leaning out...... This is adding up to be a very good possibility. I would 100% get the injectors tested, and also check spark plugs to see if its obvious. Good luck keep us posted. The factory knock sensor is designed to go off at a certain frequency, i doubt its anything loose, its either just fake or like you said a injector leaning out. Get them flow tested and pull plugs.
 

·
One fix at a time
Joined
·
8,127 Posts
It was running great on my old setup though. I think I was just running too much boost and my fuel pump couldn't keep up, causing the blow up. Everything about my new setup is different, but my knock sensor is the same as before. It's strange.


I have been running straight e85 on my current setup. I have heard you shouldn't rely on the knock sensor on e85. I guess I could unplug it to keep the ecu from pulling timing, but I am a little bit afraid that one injector might be bad and causing one cylinder to lean out, so I don't think I should ignore it. I should probably just get my injectors flow tested. Thanks for the bosch knock sensor recommendation. Yeah it doesn't make me feel good logging a 28 knock count on multiple runs. AFR, boost, and timing are all super conservative and I'm not even going past 4000 rpm so hopefully it hasn't hurt the engine. I guess it could be something loose that's vibrating at a certain frequency as well.
I just had multiple sets of stock injectors cleaned and flow tested. I used these guys since they were semi-local within an hour drive. Log into Facebook. They're through LSXKilr Custom Calibrations. https://lsxkilrcustomcalibrations.com/home-1. I would highly recommend them and you can't beat the price at $10 each. It's worth it just for peace of mind.
 
1 - 14 of 14 Posts
Top