poormxdad's consolidated tuning shenanigans thread
 

Gentlemen,

I honestly believe Nigel's new injectors have fixed the intermittent issue from the previous thread. I thought I'd start a new thread so there's not a lot of jumping around.

Here's my latest tune and a short log from driving around town. I had to zip the log to upload it. The only difference between this tune and the last is I turned on Baro Correction, set the Default to 100, and left it on.
She seemed just a bit richer, but it could just have been the placebo effect. She ran around town fine. I was expecting to not even get out of the driveway. Does the log reveal anything about baro correction?

I still have a problem with a stumble. This time, I did not go to WOT, I just accelerated gently and when she passed through about 5700, she stumbled. I lifted off the gas and she was fine. Didn't even get the check engine light.

I'm indebted to youse guys for all your help. My plan is to start making the other suggested adjustments, Any advice would be greatly appreciated.

Thanks much,

Okay, I'm even more confused now. I turned on Baro Correction and really didn't notice anything. Is that because the Barometric Correction Table is blank--all zeroes?

Also, it appears that 100% correction on that table is no correction. Is that correct? Jeez.

Thanks,

My advance curve goes up to 7000 rpm. What happens above 7000?

Have you done a lot of auto tuning? Looks like a lot of it is just poor afrs. Also, when you hit 5700, you were only at 40% throttle, is the TPS calibrated? Trying to tune your partial throttle map before full throttle seems a little backwards to me.

Are you supercharged right now? That's stupid of me, yes, you hit 128kpa.

I might take the belt off if I were you and tune that naturally aspirated cells first. You're requesting similar fuel at full throttle and 5000rpms and at idle, but you are asking for boosted afrs. 100kpa row should be more like 12.5-13, not 11.3. You also have EGO correction turned on, leave it off for tuning. Your dead time is also very wrong. There's a few issues to say the least.

I'm learning there are a few things wrong. However, before having the intermittent running like crap issue, I had taken that tune to the track for two seasons with the supercharger and she was a beast. Two earlier seasons had the MegaSquirt, normally aspirated with the squaretop and MSM intake cam. All those values, like baro correction, have probably been in there from the first tune. My lack of knowledge is at fault here. Also, I don't know if any of these issues were caused when I updated the firmware a week or so ago. I had been running 1.2.0.

I've traded several emails with Nigel from Flow Force. He says these are the same injectors as when I bought them before, but he's using a new testing process, and new correction factors. 0.630 was the dead time recommended two years ago, and 621 was where my set flowed. Dead time now looks to be twice that, and my new ones flow at 667. I created a new graph using the chart on the Flow Force website. I have no explanation why my Curve 1 doesn't cross 100% at 13.2 volts. If I were to double the dead time, she would be rich as all hell. Correct? Is that why my VE table numbers are so high? Updating the injector size to 667 should take some of that out. Correct? Is there a general correction for the entire VE table that would get me in the ballpark? It seems like those are the first things I need to fix.

I used to autotune all the time before I got the Rotrex, especially if I knew the weather during the trip to the track would be the same as track day. Now that just scares me. I have not autotuned this tune at all in two years.

I've fixed the MAT Correction Table but there are so many things I just don't understand. Comparing my tune to DNM's also scares me.

Thanks much,

I think, personally, start from a clean map.
That injector deadtime should just be set to whatever Nigel said. Don't touch the graph.
Tune idle. Manually.
(Disable EGO correction, 100% out MAT correction)
Tune WOT. Auto
Tune cruising. Auto
(Enable EGO, change MAT if you want, I don't works a lot better for me.)

Use safe timing numbers until you get to a local tuner who knows his shit and pay him good money to do it right.

That all sounds reasonable.

Seems like the first thing I need to do is figure out why I get the stumble. It was not there before updating to 1.4.1 firmware. It could be a coincidence, but I'm not a big fan of coincidences.

From Nigel's webpage Injector Data ? Flow Force Injectors I get a deadtime of 1.263ms @ 13.2 volts using the 4Bar curve. Does that look right?

It sounds like I should be using an AFR table that isn't so rich. I'm running about 7 pounds of boost max right now. Is there a known good table that someone could share?

I directed that my advance table be conservative. I wasn't looking for every last bit of power. Does it look good enough to keep? It's been used for two seasons of forced induction track time.

Gents, please keep your explanations in noob terminology. I read and write for a living. But sometimes, when I'm told something like the winkiedoodle in my tune is curved too much, I don't necessarily know how to fix that.

Thanks much,

Spark table looks fine. Just take your 105 afr row and set it to more like 12, along with the 95 row that has some 11s. And no matter your boost, I wouldn't run 11.3 anywhere, 11.5 is fine. At 7psi you're probably fine running 11.5-11.7, that's not a lot of boost. And yes, use the 4 bar graph for any NB, 3 bar for any NA. After those changes, you'll probably have to take a lot out of your fuel table.

Remember this handy calculation, if you're running a certain percentage richer than desired, multiply the effected portion of the fuel table by that percentage to gain your desired afr.

For example, if you're idling at 11afr but want 14.7, 11/14.7=.74, multiply your idle cells by .74 to achieve 14.7. Works great for EGO too, if you're averaging 105% in cruise, select your cruise cells and multiply by 1.05, etc etc.

NB FPR is not intake manifold pressure referenced,
so 5bar graph for correct idle would be better

I thought FPR pressure was about 60 psi. I think mine read 61.5 with a cheap HF gauge. That's why I chose 4Bar. I also have the high pressure Walbro 190. The 4Bar graph seemed like the right choice for both idle and boost.

differential pressure across the injector = fuel rail pressure - intake manifold pressure = 61.5 -(-9psi@idle)=70.5psi

Hmm, you read a lot of books, don't you...

From Injector Dynamics Tech Support Ticket # 32051 :

Question: "Please confirm: for non-MAP referenced returnless fuel system with 61psi base pressure I have to choose offset for 70psi(Differential Fuel Pressure during idling is 70psi=61psi fuel rail -(-9)psi manifold vacuum) "

Answer: "Confirmed. And given an offset error is largest when the pulse width is smallest, it will be best to use the offset data at 70psi in your situation, it's the best compromise. Tony Palo, T1 Race Development,Injector Dynamics "

I have not been able to come up with any benefits of using baro in a speed density system.

I would love to be educated here if there are any.

If it was my car, I would turn baro off.

Gents,

I increased the injector dead time to 1.263 while idling in the driveway. AFRs went way rich. Is there a formula I can use to reduce the entire VE table based on increasing the dead time?

Thanks,

you could derive it. Would come down to something like newVE = oldVE + (oldDeadTime - newDeadTime)/(ReqFuel / squirts)



that is just an example, don't use it, derive it for yourself.

edit: there is absolutely an error in that equation! DO NOT USE IT!!!

I interpolated the 5Bar curve and get ~1.46ms dead time @13.2 volts. Is anyone else running something close to that number with the Flow Force injectors?

I understand the simple math relationships above. I was wondering if there were some kind of mathematical correlation between the dead time and the VE Table. For example, if I double the dead time, I need to cut the VE Table in half to get me in the ball park. Or, if I increase the dead time by 0.600ms, I should multiply the VE Table cells by 0.600 times the square root of pi.

Nigel passed me a dead time, a voltage curve and a small pulse width curve, attached.

He also said the injectors are flow tested at 43psi. Mine would flow about 788 corrected for 60 psi. Should I change the injector size to 788, since I'm making all these other changes?

Can anyone tell me why changing one curve in Tuner Studio messes up the remaining curves? It doesn't duplicate the new Curve 1 in the other curves, but rather just changes them, like it's adding the new values into the old curve rather than replacing them.

Thanks,

Yes, you should change your required fuel to include the 788 size.

This should help lean out your fuel table, but I don't think it'll be enough.

Dead time is directly added to your pulsewidth, so after MS calculates your PW off your VE table, warm up enrichments, etc etc, it adds dead time. It's a big chunk. I'd find the PW calculation and solve for X. X being your VE btw. Then you'll have your multiplication factor.

I decided to hang it all out and make a bunch of changes all at once.

I changed the AFR Table after reading through all the AFR table threads here. I changed the Required Fuel and Injector Size, Dead Time, Dead Time Voltage Curve, and Injector Small Pulsewidths Curve based on Nigel's suggestions. Nigel's dead time number comes from the 4Bar graph. I changed the MAT Air Density Table per the suggestions here, and turned off Baro Correction (after resetting the Default to 100 kpa), She started right up but was running really rich. I pulled about 40 units from the Idle VE Table--90s to 50s. Then I messed with the Idle Valve calibration. She idles nice and smooth, but the rest of the VE Table is obviously crap now. I'm sure I missed a few things.

I removed the Rotrex belt. If the weather is decent tomorrow, I'll take her out for some autotuning.

I still have to deal with the stumble.

Thanks for all the inputs so far.

I've done a few hours of searching and reading. I'm going to go ahead and scrap the current tune (which I suppose I kinda have by changing all the stuff above) and start using Include AFR Target. My proposed AFR Table is attached. Believe it or not, it is a few tenths leaner almost everywhere than before. I want to balance running a little rich to be safe on track with not running too rich that she breaks down the oil or destroys my cat. I've always thought she was too rich on track.

What do I do to the current VE Table since I'm switching over to including the AFR Target? Are there Startup/Idle changes that need to be made?

My research indicates you all are comfortable autotuning fuel and taking the car to the track. Can I replace the 90mm pulley with an 85mm, and use that AFR table and the Ignition table included in the tune in Post #1? Are there advantages to autotuning the non-boost parts without the supercharger belt attached if I'm including the AFR Target?

Thanks,

Congrats on switching to include afr target. Smart move ;)

You will need to retune VE everywhere, but it shouldn't take too much autotuning to get you close.

Equations for determining the injector flow rate with a pressure change are out on the internets. I think witchhunter might have one.

I would not tune with the belt off. I assume you are running speed density, so the fuel equation inputs don't care either way, but AIT frequently has error that is compensated for in the VE table.

Steps in the right direction. Tune with the belt on. Do not let VEAL tune idle or trailing throttle, and only fully warmed up. There are settings that will do this for you automatically. Highlight cells and block them and set a minimum CLT.

Including AFR will not affect start. It will affect afterstart and idle VERY LITTLE as you are nearly stoichiometric in those areas of your AFR table. See the picture? When desired AFR = stoichiometric AFR, there is no adjustment of fueling.

Seems like you are going use 4bar.
Correct injectors deadtime is a corner stone of any tune.
You have to use differential pressure to find your injectors deadtime.
Differential pressure on non MAP referenced returnless fuel systems is difference between rail pressure and manifold pressure.
So it will be different on different manifold pressures.
It is 70psi@idle on on your car with 60psi and 53psi@full boost(I assume your compressor makes 7psi)
Best compromise is using 5 bar deadtime because correct deadtime is much more important on low pulse width.

I already mentioned my question to Injector Dynamics tech support and their answer. I believe ID is reputable and reliable source of knowledge.
Also you can search mt.net forum with "differential pressure" keywords.

P.S.: Another question: Is provided deadtimes correct?

No congrats necessary. I wanted fuel tuned on a dyno. Apparently, it's better/easier to do it without including the AFR Target. I ran that tune with only minor tweaks for two seasons, I blame myself for anything non-optimum.

Yes. Flow Force provided recommended settings. My research indicates dead time is not life or death. Mine was very wrong, and I drove it on track, and to and from events, for two years without issue. I admit, however, the new settings feel better, but I have a ways to go to get the new tune complete.

Not to cross hairs...

I have actually had great success using veal for idle tuning, and I have a full 16x16 table dedicated to idle. The trick is to idle while moving. Slowly let clutch out to get moving, shift into 2nd and 3rd, still at idle. Also useful to do it on a grade and adjust target rpm when you get the lower loads at the target idle rpm dialed in.

Determine the optimal idle AFR by adjusting AFR for lowest manifold pressure (load). This requires tuning idle control and VE first. You can also do it with adaptive idle advance, and tune to the lowest timing applied to maintain idle. My 1.6 for instance prefers an idle AFR of around 13.5 in the current configuration. At other times (different setups) it has been as low as 12.5 or as high as 14.3.

Alrighty then... I took her out for some autotuning. I had two small, but noticeable issues.

First, it seemed there were times when VEAL stopped making changes. I had Max Cell Value Change at 40, Max Cell Percentage Change at 40, and Very Easy selected. There were times when I was holding a throttle position on some very lean AFRs, but they weren't changing.

After I had her good and warm (so to speak), the idle behavior changed. When I pulled to a stop, the idle would be up around 1400 and would slowly ramp down. Before she was fully warmed up, the idle seemed more normal. I've obviously got work to do.

Here's a log. Thanks for taking a look.

My Friend,

I doubt I'm smart enough to accomplish what you suggest, but I do appreciate your input.

Static fuel pressure will wonk the fuel equation via breaking ReqFuel. This is somewhat counter productive in terms of the advantages offered by incorporating AFR target. That being said, the error in ReqFuel will be absorbed into VE directly due to the product nature of the equation...

except of course for changes in dead time due to dynamic differential pressure.

ID 850s for instance are way more sensitive to voltage than pressure. By the time you are in boost, the error in dead time will tepresent a much smaller percentage of the pulse width.

ID850 Injectors - Injector Dynamics

not as hard as it sounds. When you get there, you will be ready.

tune file might be more helpful

Yes. Attachificated.

BTW, I also replaced the crank position sensor. I had no trouble revving past 6000, but I admit I did not push her into WOT. Is there anything in the log or tune that could be causing a stumble at WOT? I hope to take her out again this afternoon.

Thanks,

New Weirdness
 

It's supposed to be warm this afternoon, so I left work early to go autotune. She started and every few seconds would cut out momentarily. I have a log file. Also the check engine light gave one long blink, then eight short blinks.

The cut outs got spaced longer and longer apart, and stopped completely after she warmed up. She did not do this the last time I drove her around. I went for the last restore point, but TS said it was the same.

Any help would be greatly appreciated.

Thanks,

I'm going to need some serious help. Maybe I just don't understand the theory.

I didn't have any problem with cut outs I mentioned in the previous post after she warmed up. She ran really well. No problems going into WOT. She's a beast, even untuned. So what was happening?

I enabled AE, Accel Pump-AE. Now, the low load AFRs are much higher than before. I'm not sure how to fix that and VEAL did not. High 16s-low 17s. I never saw numbers that high before.

When I first started her up, the idle AFRs were really rich. I pulled some fuel. That might have led to issues later.

The big issue came when I parked her and let her heat soak with the hood closed. When I started her back up, the AFRs were really high. IAT was over 100. With the old tune, it did not take long for the cool air from my actual cold air intake to bring the IAT down and everything would settle out. This time, the AFRs remained high, both driving and idling. I would have thought they would be really rich since I have MAT Correction pretty much flat with 70 degrees at 100%. The attached log is with the high AFRs after the heat soak.

One other weirdness. I have the rev limiter ON and set at 7200. It does not appear to be working. It used to.

I plead severe noobness.

Thanks,

you turned EAE on ... did you mean to do that?

it is also helpful to use project backups instead of msq. Not sure we can upload them here due to size, but you can cut down the size by not including datalogs and restore points.

Without trying to be flip, I basically turned everything on. That's the first thing I'll turn off.

Your definition of structured approach differs greatly from mine. If I were your engine, I'd be fearing for my life.

I decided I was making so many changes, I might as well start over.

What's really interesting is comparing my current tune to last year's tune under 1.2.0. Accepting changes to the AFR table (and Including the AFR Table in the VE table), the VE Table itself, injector size and dead time, it's mostly idle stuff. She only seems to have problems with the new idle parameters and temperature corrections. Highway driving and acceleration were fantastic.

I cannot get my brain around why she's really lean throughout the lower rpm range when she gets heat soaked if the MAT correction is flat.

I really need help to get the idle right, and I haven't been able to figure out the cut out problem at idle when she was warming up (Post #34).

Thanks,

Just to clarify Post #34, the engine never actually quit. There was a lost of sync, she stuttered for an instant, but kept running. That happened over and over till she warmed up.

Any help will be greatly appreciated,

Gents,

I turned off EAE and the sync loss disappeared. I backed away from some other things as well.

I turned off Overrun Fuel Cut, and I went back to Open Loop Idle control. She runs great. However, I still have a couple of issues.

First, I tried to hit some of the low load, higher rpm cells with VEAL and TS pulled shitloads of fuel. I was basically trying to hold 40 mph in second gear. I briefly saw AFRs in the 18s. How can I stop that from happening? In the attached log, it's around 40 kpa at 4500 rpm.

When the coolant temp starts the fan going, the AFRs rise into the mid 15s, even though it appears she's still running under the Idle Control VE Table. What can I do?

The car ran fantastically at speed. Other than the above, I had no issues driving. However... I parked her and let her sit for 20 minutes or so. When I restarted her, she wouldn't idle without me manually fanning the throttle. The AFRs were in the high 16s. I'll never figure it out myself.

I still have work to do with respect to the MAT Correction and some other temp curves.

Thanks in advance,

Ah, the incredibly common hot-restart issues.

This problem doesn't exist, something must be wrong on your end.

Do you think the hot start lean issue is separate from the hot lean running issue?

My guess on the hot lean is your injector voltage dead time compensation. Although, I didn't open this log, I did open a different one and your voltages looked stellar. Check the viltage at the injector vs what is being reported in TS when the fan is on as well as when it is off.

thumbs down. what's the reasoning behind turning off overrun?

looks like youre idling 100RPM higher, so the fueling is slightly different, but youre fueling the exact same amount for both conditions.

this seems to plague NB cars, i think it's an issue with the fuel getting hot. Turning on EGO would help and other closed loop code would help.

Definitely a NB (no-return) heatsoak issue. You might have to tweak your ASE curves to send extra fuel through on hot-restart conditions. The only REAL solution is to convert to a full-return system so that the fuel rail doesn't heatsoak as badly.

Thing is I've seen this before, but with the previous tune it only lasted a few seconds and I never had the hot no-start issue.

I'll change back to closed loop idle, but I was pretty happy with the open loop idle behavior.

Is there a way to not completely lock a VE cell, but rather prevent VEAL from reducing it?

Thanks

if you were happy with open loop idle, then you should be able to populate a good initial duties table based on your open-loop table, then CL should be able to work from that point.


there's a filter for VEAL. I tend to never let it tune under 30-40kPa, its never tuned my bottom rows of cells very well.

Gents,

It's not a hot idle problem. The car sat for almost 24 hours in the cold garage. I pushed her out into the driveway after checking a few settings. Ambient temp is in the low 50s. She would not run unless I put a little throttle in. I turned on and off a bunch of things--went back and forth between open and closed loop, Idle VE Table on and off, etc. I recalibrated the TPS. She would not run on her own till the coolant got above about 160, and then she was idling at or below 700, at about 12 AFR. I did the Idle Valve test. No change in rpms at all.

Something must have happened on my last leg to the house yesterday. It might be in the last log I posted yesterday. I've also attached a log file and a composite log from today.

I do not remember changing any settings after I got back in the driveway that would cause this. Did the idle valve shit the bed?

I hit the spacebar when she was able to run on her own.

I also have an unhandled error on startup, and the first time I started her, I got one long blink and eight short blinks of the CEL. I did not get the CEL when I was able to keep her running.

Any help would be greatly appreciated.

Thanks,

PoorDad: The above quote is what I was referring to above. I try to look at your logs and mss's, but when I go to make a suggestion, you've already changed all kinds of things.

We need to kind of try to work on one thing at a time.

So much of the advice that is now being presented to you, has been given before in the other (2) threads. But you don't stay on one subject long enough to find an answer.

As a concrete example, you mentioned that you were running open loop, and yet idle VE. Well, when I opened the latest msg, it was closed loop. Kind of hard to address an issue when there is so much jumping around.

That said.
1) VEAL works correctly even when EGO is operating. So, I suggest you turn on EGO while you are trying to tune.
2) As has been said on multiple occasions, don't let VEAL tune idle, nor low kPa (by the way, please be consistent in your defining low kPa... are you speaking 45kPa, or trailing throttle below 25 kPa?)
3) You said you parked for 20 minutes and then had idle issues. When we said, "hot restart", you say, "no! it happens after 24 hours". Help us help you.
4) Here's the spoon:
CEL
2: MAP
3: MAT
4: CLT
5: TPS
6: Battery
7: EGO (AFR0)
8: Sync Fault
12: Knock Sensor
17: AFR Shutdown

5) I admit that I do not understand where you are running a pretty steady load and RPM that your AFR does not remain consistent.
6) I find it helpful with VEAL to occasionally Save the Tune As. This resets the tables and allows VEAL to start anew.
7) I don't know how to set up VEAL to only add fuel. I suggest you manually add some fuel, and then run VEAL to begin with a rich situation and remove fuel as needed.
8) Please elaborate on exactly how you ran the idle valve test.

I will check you log and msg and see if I can discern anything.

It would seem to me that your WUE is rather low. Before you tackle that, though, you need to get your fully warmed up VE correct in the idle area. I don't care if you use Idle VE or not, but tune whichever you are going to use... or set them both the same to avoid confusion.

THEN you can set up your WUE. That is likely the problem with running rich or not wanting to run until you are past 160*. As an example, my 120* WUE = 108%. Your's is 101%, on the last MSQ I have. So 7% leaner than mine, which would yield 15.7 AFR for my 14.7 AFR, if all other things were equal. However, you are running quite rich, not lean.

At your red line, that is when you finally went into closed loop. At that time, your Advance kept you running. I agree that it does not look like your IAC valve is working. I also don't know what kept you out of Closed Loop for so long. Well, likely your TPS minimum setting.

Also, until 295 seconds, your target idle RPM is "0". Oh, again, because you are holding the throttle open above the set-point.

Side question, what O2 sensor / controller are you using?

Sir,

I can't thank you enough for your time. You are suffering from Forum Confusion Temporal Disorder. I'm working the issues when I have time, but you're answering my questions when you have time, and those times aren't synching. My posts tell a sad, sad story, chapter by chapter. You can't read the story starting at the last chapter.

This is a perfect example. I stated in my newest, previous post that I went back and forth between some settings, trying to find out if I'd done something to cause the issue I outlined. I must have left it in Closed Loop when I started the log.

I've read several posts recommending turning EGO off when tuning.

I stated in Post #42, "In the attached log, it's around 40 kpa at 4500 rpm."

My later posts trump earlier posts. I thought I had a hot restart issue when I tried restarting her last night when I got back home after she sat for 20 minutes. I knew I was wrong when she wouldn't idle cold today without me fanning the throttle.

I had that data. I don't see one long blink and eight short blinks (18) in there.

Again, I can't thank you enough.

Sir,

I don't know if it went into closed loop. I pressed the spacebar (red line) when she would idle without me adding any pedal.

I thought I had the VE table basically conquered.

I believe something is wrong with the hardware. I changed back any tuning values I might have changed yesterday that could have caused the problem. I couldn't fix it. I have a complete throttle body assembly, with idle valve, somewhere. I plan to swap that in tomorrow, unless someone can point to an issue in the log file beforehand.

Thanks,

But isn't the table interpolated?

From previous post, the long CEL is just to signify the start, only the short pulses are counted.

From this post: You can tell when it goes into CL because Idle Advance Correction goes from ")" to about 14* (consequently, total Advance goes from 10 to about 20).

And I think much will go better with another IAC Valve.

That's really interesting. So, AFR Shutdown (whatever that means) is actually 17 short pulses. Why do they need the one long pulse as a spacer when they could just use a long blank? I'm fairly certain my '90 Miata uses the OBDI tens digit and ones digit convention, which is why I assumed the code was 18.

Is there a way to actually tell if the Idle Valve is bad rather than just swapping out the whole throttle body to see? I forgot about the coolant lines running to the TB, and I don't actually know if the TB is good. It came on the JDM motor I bought.

If you ran the idle valve test correctly, and RPM did not change, then
1) it needs cleaning
2) it needs replacing
3) it is not getting the signal from the ECU.

On #3, since you have 'Run Valve Before Start' set to 'On', you should be able to hear the 512Hz buzz KOEO.

There may be other possibilities, but I cannot think of any.

What is your cranking pulse like at full hot? My nb doesn't like an 80 there. More like 150. I couldn't start the car at all. And when it did start up it had trouble keeping idle.
Try that and report back.