View Full Version : My preemptive tuning problems thread
q2418130103p
09-25-2011, 12:32 AM
I just got my 550cc injectors installed this morning, planning on tuning with my (reflashed) neo tomorrow. I am anticipating loads of problems, so I figured I would just start my 'tuning troubles thread' now and get some of the baselines out of the way.
PTE 550cc injectors
hotwired supra pump
afpr (with gauge)
apexi neo (reflashed)
13t's
fmic (not a very good one, but it is still an fmic)
fresh copper plugs
WB o2 (which I can log in HHH)
Based on the team3s safc guide, I set my low throttle position to 30% and high throttle position to 85%. For initial corrections I did -30% for high throttle points and -35% for low throttle points.
So far I have only started the car, and let it idle (warm). This went fine, but I had to drop the correction to around -25% for the 1000rpm low throttle point to get the afr correct. The low fuel trim never dropped to 100%, but I am guessing I have to wait longer than a couple minutes to see that move.
So any suggestions, or 'plan of action' ideas considering my parts, would be appreciated, but I think 550cc injectors with the safc or neo have been discussed an awful lot. So I figure I will post an update after things go horribly wrong tomorrow. I am hoping that having a wideband tied into my logging data will make this a snap...
q2418130103p
09-25-2011, 04:54 PM
Alrighty, it has begun.
First concern: The team3s tuning guide (Team3S- Dodge Stealth Mitsubishi 3000GT Owners Pages (http://team3s.org/FAQ-AFCtuning.htm)) attempts to get you to maintain 100% low fuel trim when tuning the low throttle points. This sounds good to me.. except that it seems to take forever for fuel trims to adjust, and the guide give the indication that it should be quick (this is a inference). In fact, I never actually see the trims change, they only seem to change after I turn the car off and on again.
Second concern: I am at -18% reduction at idle. The o2s cycle at around -25% or so, but I kept reducing to try and get the fuel trim to 100%. This seems like its way too little of a reduction for 550s. I am aware that some people have seen PTE injectors flow less then they are rated at. Perhaps I am not in the middle of the ECU's correction band, I assume this is why team3s has people shooting for 100% trim.
RealMcCoy
09-25-2011, 06:31 PM
You need to step back and wrap your brain around what the all the trims mean, and what the low and high throttle points represent...
The fuel trims (low, med, high,) react to FTO2. FTO2 is a short term reaction to O2 sensor changes. The low, med, high, designations are for for airflow. They do not represent RPM or throttle opening. Your total trim is going to be whichever trim cell you're in + FTO2. so if you're driving down the road in the med cell, FTmed at 95%, FTO2 averaging 105%, your total trim is 100%.
If your state runs E10 as standard fuel, knock about 4% off your expected correction right off the top.
The fastest way to get a quick set, is to reset your computer, get all the trims parked on 100 (that's where they default on reset) then tune from FTO2 before the trims have a chance to move around.
You'll also find that the load tables are off in the lower rpms due to your airflow correction, and you'll have to run it REALLY fat in the lower rpm, high throttle sections to compensate.
q2418130103p
09-25-2011, 06:55 PM
Ok, reset the ecu, trying to adjust off the ft02 right now.
q2418130103p
09-26-2011, 01:51 AM
I believe I got the low points set correctly, so I went ahead and moved onto the fun part.
I went ahead and started doing 2nd gear WOT pulls. The first pull was pig rich at all rpms, no knock... not that it means anything in this case. I went ahead and started making adjustments, and doing more pulls. As I made adjustments, I started to get a bit of knock, but as I got the AFR closer and closer to the factory values the knock began to go away. The last run I did had a afr curve that looked decent, but the knock went back up, way up.
I have attached a picture showing two wot runs. The brighter lines are from the run with the 550cc injectors with boost set at 15psi. The darker lines are from an almost identical pull made yesterday with the stock injectors and the boost set to 9psi.
3257
q2418130103p
09-26-2011, 08:22 PM
Here is a better copy of the image, its a bit cleaner.
http://i540.photobucket.com/albums/gg354/q2418130103p/compare2.jpg
q2418130103p
09-26-2011, 09:11 PM
Here is an AFR plot for the run (550s, 15psi). I also plotted the 'factory afr' estimate, which only has 500rpm steps, then I plotted a best fit polynomial to try and smooth out the 'factory afr'. I knew I was running it rich, but I didnt realize I was THAT rich... The factory table is obviously already conservative.
http://i540.photobucket.com/albums/gg354/q2418130103p/afrplot.png
q2418130103p
09-26-2011, 09:25 PM
Knock for the same run:
http://i540.photobucket.com/albums/gg354/q2418130103p/knockplot.png
q2418130103p
09-26-2011, 09:26 PM
Timing for the same run:
http://i540.photobucket.com/albums/gg354/q2418130103p/timingplot.png
q2418130103p
09-26-2011, 09:28 PM
IDC Plot:
http://i540.photobucket.com/albums/gg354/q2418130103p/idcplot.png
RealMcCoy
09-26-2011, 10:57 PM
In my experience, once it starts knocking, it often doesn't stop until you get out of the throttle. Even pulling timing, once the fire is lit, it stays lit... It looks like you need to add a bunch of fuel at lower RPMs. I usually keep the low throttle numbers fairly flat, and add fuel to the high throttle points. That way it compensates for the hole you made in the load curve in a gradual manner, without drowning it in fuel at low load. But I would concentrate on the area that the knock starts to take off first, then work up from there to find your curve.
I wouldn't be the least bit surprised to see the area around 1500-3000 RPM, high throttle, as much as +10% or more from the rest of your settings.
q2418130103p
09-26-2011, 11:10 PM
I noticed that happening (the +10% or more from the rest of the points at low rpm), the lower rpms for the hi throttle settings are already much richer, I will go ahead and richen it up some more. I will probably lean everything else out though, since I am well below the 'factory' curve. I dont understand this though, the mixture is so rich when it starts to knock... am I missing a piece here or are you thinking that the very lean low rpm points might start the following chain of knocking?
I kept my low throttle points fairly flat across the band, no more than a 2% variance, but that is mostly because I dont know what else to do besides make it cycle o2s and the car does a decent job of figuring that out for itself. I did like you said and reset the ecu and tried to get my adjustments in before the long term trims reset, i only did this until 3k.
I think after I get it running clean that I will click the neo into 'pro' mode to smooth the curves out some more. I have never done this before, but it would seem that having the AFR in the logger data makes it a lot easier. Well... it at least makes for some perdy graphs.
RealMcCoy
09-26-2011, 11:21 PM
am I missing a piece here or are you thinking that the very lean low rpm points might start the following chain of knocking?
That's pretty much exactly what I'm saying...
I kept my low throttle points fairly flat across the band, no more than a 2% variance, but that is mostly because I dont know what else to do besides make it cycle o2s and the car does a decent job of figuring that out for itself. I did like you said and reset the ecu and tried to get my adjustments in before the long term trims reset, i only did this until 3k.
That's usually pretty much what you have to do to keep the fuel trims in line... Just remember if you have to sacrifice perfect fuel trims for proper open loop mixture, it's not that big of a deal. As long as it's still in it's range of adjustment you'll be OK. I like to get it as close as possible just so I don't have to deal with fuel trims being way off if I have to reset the ECU.
You can always see the point where it go into open loop on the logs. the FTO2 will fix at 100.
q2418130103p
09-26-2011, 11:34 PM
I didn't know that about the ft02, that should come in handy. I think I have the low good enough that it doesn't fall on its face after an ecu reset.
I should hopefully post updated logs Thursday or Saturday night.
I think I am happy with how it is going so far. It is starting to feel, and sound, cleaner.
q2418130103p
10-05-2011, 07:47 PM
I went ahead an dumped a bunch more fuel into the 1k bin, and took fuel out of every other band. I then did a run to 6000rpm @ 15psi with no knock. I then removed some more fuel from every bin besides 1k, I did a run with that to 6500rpm and got 6 counts of knock at 3k. The changes I made to the fuel were so slight that that it didnt really even affect the AFR curve, so I am not sure why that run had knock.
Here is a 'compare session'. The brighter, thicker line is the new run. The darker, thinner line is from before installing the injectors (and running 9psi as opposed to 15psi):
http://i540.photobucket.com/albums/gg354/q2418130103p/comparesession-3.png
I am going to put together some other graphs from the run right now.
q2418130103p
10-05-2011, 08:15 PM
Here is the AFR vs RPM graph:
http://i540.photobucket.com/albums/gg354/q2418130103p/afr-oct2.png
q2418130103p
10-05-2011, 08:25 PM
Timing and knock together:
http://i540.photobucket.com/albums/gg354/q2418130103p/timingandknockoct2.png
q2418130103p
10-05-2011, 08:37 PM
I thought this was interesting. This chart only works if the timestamps in HHH are accurate. This chart shows the amount of time (in ms) it takes to gain 500rpm. So the first point you see is the amount of time it took to get from 1000 rpm to 1500rpm:
http://i540.photobucket.com/albums/gg354/q2418130103p/acceloct2-1.png
UTRacerX9
10-06-2011, 12:37 PM
From what I remember, the stock ECU doesn't view knock counts below 7 or 8 as anything to worry about, so don't be too concerned with the occasional blip of knock. Best way to tune is to be methodical. I'd also recommend doing 3rd gear pulls instead of 2nd. (Faster speed and more room needed, yes, but also is easier to tune with a more gradual rpm increase.)
q2418130103p
10-07-2011, 07:35 AM
I have read that as well. I know 6 counts isnt a big deal, but I want to see zero. Plus when you look at the timing, it is a bit lower across the board than the baseline run was. I know that is partially because of the load map, but it shouldnt be terribly off...
I will do some 3rd gear runs this weekend. It takes around 20 minutes to get somewhere acceptable to do it, that is why I havent done any yet.
q2418130103p
10-07-2011, 07:36 AM
Scratch that, I just looked at the timing again... looks almost the same as the baseline run.
futurevr4man
10-07-2011, 09:33 AM
im not sure how familiar you are with tuning and how everything communicates to eachother so im going to just say some general things.
its going to be hard to hold off knock and still hit a good amount of boost with the amount of timing you are running. if you had turned up your 360's before you would see that timing usually hovers around 18 - 20*, but because your injectors are so big, your IDC is going to be really low, which is a major player in the ecu determining how much timing to add/pull.
13t's are just a baby upgrade in the turbo world, and you will not really be pushing that much more boost at 18psi with those compared to 18psi in 9b's. dont get me wrong, there IS a difference, its just not that big of one. anyhow, with 26* of timing, working your way up to 30, that is how people kill engines. that is just too much timing. im not trying to be a negative nancy, im just trying to warn you and hopefully help you.
anyhow, with the injectors you have... its very common to richen up the idle like you noticed you had to. our ECU's will work out the idle for you (subtract or add fuel) so you dont have to worry too much about it, as long as your o2's are cycling. dont get too crazy with the richening though. keep it about 10% more rich than everything else (from my experience). otherwise you seem to be on the right track. if you get prolonged double-digit counts of knock like you did on one of your graphs, your ecu has a "knock mode" or limp mode, or whatever you wanna call it. it will set you in a table with less timing, and you will need to reset your ecu.
i know i am a "nobody" to you, but i have tuned multiple times with the safc. i have seen a lot of people kill engines with knock... knock kills your rod bearings quick, so you want to keep it safe, but fun :)
futurevr4man
10-07-2011, 09:50 AM
so that you dont think im blowing it out my butt:
Stealth 316 - 3S Ignition System (DOHC) (http://www.stealth316.com/2-ignitionsystem.htm#j2c)
scroll half way down and check out the 2 charts that are posted. low 20's is the very highest i would go with your timing. no more than 23*, and even that is considerably high.
q2418130103p
10-07-2011, 06:46 PM
I appreciate your input, this is my first time tuning anything. I do believe that I have a reasonable understanding of the basics involved here.
I have no intention of hitting a higher boost. I wasnt planning on going much past 15psi, if I move it at all. On the first page I have an IDC chart, which you will see is hitting 90% at 6000rpm, which I thought was a good sign about the size of the injectors. I figured I MIGHT try and raise the boost to get some higher IDCs if I have to (while maintaining little to no knock), but since I am already at 90% IDC at only 6000rpm, I dont think I am going to have to. Now, I am assuming that HHH reports the IDC correctly, I dont know if that is the case.
I will get some 3rd gear plots up this weekend.
Thanks again.
q2418130103p
10-09-2011, 01:30 PM
Just did two 3rd gear pulls.
The first run had double digit knock, and I noticed the timing dropping from 2k to 4k, so I richened up that section by a couple points and did another run.
This second run had no knock, but it started a little bit later, like 1600rpm instead of 1200rpm. It also ended a bit sooner.
Here is a compare session. The darker line is the first run (with knock). The yellowish line is the good run:
http://i540.photobucket.com/albums/gg354/q2418130103p/3rdgearcompares.png
Here is the afr for the good run. It goes all wacky right at the beginning, just like the 2nd gear logs. Except that the 2nd gear logs started under 1k and these ones start well above 1k. The timing for this good run slowly rose, with no dips. The IDC maxed at 85% (but at 6200 rpm).
http://i540.photobucket.com/albums/gg354/q2418130103p/3rdgearafrnoknock.png
UTRacerX9
10-09-2011, 10:42 PM
That's more like it. I'd advise to get a few completed 3rd gear pulls to redline to make sure that one run wasn't a fluke, but from the data, looks like richening up in the midrange solved your knock problem. If it holds true for a couple more runs, then that's where you can start to make really fine adjustments to get your AFR set at a safe level.
Let us know how it goes :)
Oh, almost forgot... After you get a basic tune with no knock from a full 3rd gear pull, I'd recommend doing a 5th to 3rd downshift and floor it and log that. Assuming your mounts are good and not getting any phantom knock, you should be able to see whether your tune is really dialed in or not. I can't count how many times I've heard of someone who tuned their car, then went to do an impromptu race and something bad happens in the heat of the moment, and they have no idea why.
futurevr4man
10-10-2011, 09:08 AM
that second run looks pretty clean! im sure it pulls like a beast with that much timing too!
you mentioned how HHH reports IDC... you can go well over 100% i saw 120% on my stock injectors without hitting fuel cut still. im not saying to do that... im just saying that 100% on HHH is not really 100% of your injectors. you have to manually do the calulations with your CFM, engine size, injector size etc to get a real estimation on your IDC.
q2418130103p
10-10-2011, 10:21 AM
Yea, it is really starting to feel good. I have only done long wot pulls with it so far (1k-6.5k), so it only feels fast for a bit. I haven't tried going 1st-3rd yet. Although I have to say I am a bit concerned that I could be on the edge of it working or not working. I am not sure why I feel that way, but I feel like any day some environmental condition could change it and would push the car into a danger region. I dont know.
Anyway, I have heard differing things about loggers and IDC. Some people have said that loggers misreport the IDC, which is why it can go above 100%. Others have said that the IDC is reported correctly, but a number past 100% means the ECU is demanding the entire injection window be used for fuel delivery.
I always believed the 2nd thing since I am fairly certain that the IPW is reported correctly, and whenever I have manually done the calculation below, it matches the IDC reported by the logger. And whenever I see an IDC above 100% is corresponds to an IPW LONGER than the window which the injector can fire.
This is based on some stuff from stealth316, reworked for presentation purposes:
2 * 60/RPM = W(seconds) (time to complete two revolutions, total window that you could inject fuel)
IPW (Seconds) / W(seconds) * 100 = IDC%
For instance, I am looking back at my pre-injectors run right now. The IDC is greater than 100% for a little while. During this time the IPW (17.6ms) is greater than the W (17ms, the time to complete two revolutions).
By the way, this is totally unrelated, but I just wanted to put it in teh thread so that when I am looking back at it later I will have the data:
Air/Fuel Ratio Limits
***6.0:1 *Rich run limit*
***9.0:1 *Low power, black smoke*
*11.5:1 *Rich best torque at WOT*
*12.5:1 *Safe best power at WOT*
*13.2:1 *Lean best torque at WOT*
*14.7:1 *Chemically ideal*
*15.5:1 *Lean light load, part throttle*
*16.2:1 *Best economy, part throttle*
*18-22:1 *Lean run limit*
q2418130103p
10-10-2011, 10:36 AM
By the way, does anyone know conclusively whether long term fuel trims affect open loop operation. The stealth316 article says it doesn't, but it also doesn't seem too confident in that. The reason I ask is because I dont want my WOT tune to change as a result of the fuel trims changing (like after an ecu reset, or if I adjust the low throttle correction points).
q2418130103p
10-10-2011, 08:13 PM
Today I did a 1st - 4th WOT run (didnt finish 4th for obvious reasons). No knock for the whole run. I made no changes to the correction points from the last run, so all the plots are basically the same as before.
Including my 1.1s shift from 1st to 2nd, and my 1.4s shift from 2nd to 3rd, both of which were early shifts (~6.3k) the 0-100mph time was 11.9s. That being said, I cheated a bit. I didnt actually 'launch', I started at 15mph and extrapolated the speed vs time data for speeds before 15 mph. I am sure the slope of my extrapolated data is correct, but what is off is obviously the time for the launch. But I think with 100mph worth of time that the margin of error is acceptable. Anyway, this was just for fun and should be taken with a very large grain of salt.
http://i540.photobucket.com/albums/gg354/q2418130103p/mph.png
q2418130103p
10-10-2011, 08:50 PM
oh yeah, I forogt. I didnt have the latest version of HHH installed which compensates for the difference in timing between DSMs and 3Ss. All my timing graphs are 12 degrees too high. I just opened a couple of my old logs to verify this.
q2418130103p
10-11-2011, 06:59 PM
That's more like it. I'd advise to get a few completed 3rd gear pulls to redline to make sure that one run wasn't a fluke, but from the data, looks like richening up in the midrange solved your knock problem. If it holds true for a couple more runs, then that's where you can start to make really fine adjustments to get your AFR set at a safe level.
Let us know how it goes :)
Oh, almost forgot... After you get a basic tune with no knock from a full 3rd gear pull, I'd recommend doing a 5th to 3rd downshift and floor it and log that. Assuming your mounts are good and not getting any phantom knock, you should be able to see whether your tune is really dialed in or not. I can't count how many times I've heard of someone who tuned their car, then went to do an impromptu race and something bad happens in the heat of the moment, and they have no idea why.
I totally missed this post. I will try what you suggested when I test again this weekend.
Powered by vBulletin® Version 4.2.3 Copyright © 2026 vBulletin Solutions, Inc. All rights reserved.