Excessive Closed Loop Oscillation between 1%-9% TPS

[CGF20BOOSTED]

I have a F20B with about 10lbs of boost.

stock block
ram horn mani
garrett turbo
plx wideband
1000cc injectors
e85 flex
omni 4bar

While cruising at any speed in any gear, I get a radical fuel trim sweep between 1-9% throttle position. I can see it on my wide band and I can hear it from my exhaust. At 10% TPS or higher, I do not have that problem. Also, I do not recall having this problem before and can't pinpoint when it started. I disabled closed loop and noticed i was rich at those loads. So I corrected and retried. Same result. I tinkered with the fuel trim switch rate and still have the oscillation. So I've been cruising in open loop while data logging and cleaning up the part throttle fuel map. I suspect injector dead times could be an issue, but I believe they are dialed in correctly. I do notice a lean spike at throttle tip-in on the 1st-2nd gear part throttle shift (ex/ while driving in parking lot). I've been tuning my own car for about a 6 months or so. I have been successful for the most part, except for the part throttle fine tuning here. I have 18 years professional experience in automotive repair and engine performance diagnostics. However, I'm fairly new to tuning and this is a learning experience. Here are some data logs with open loop and closed loop and my calibration. Please let me know what you think and any pointers to correct this issue. Thanks for your time and efforts.

[Hondata]

The injector dead times are far too low. At 14V for 1000cc injectors they normally are 1.25ms, not 0.3ms. If you change the dead times then you will need to retune, especially at low load.
Other causes for excessive short term trim swing we've seen in the past are exhaust leaks, o2 sensor mounted too far away from the exhaust center line, o2 heater not working.

[CGF20BOOSTED]

So the dead times will cause the excessive oscillation? The other possibilities are not likely but I will recheck of course. Is 1.25ms a rule of thumb for 1000cc injectors? I had a dead time list that came with the injectors. However, I wasnt successful with those numbers at the time of installation. At 14v, it said 0.56ms. Then with 1200cc injectors, it states 1.62ms at 14v. 1600cc injectors with .92ms at 14v. So I didnt trust the dead times. There has to be a typo there. I'll retry with the provided dead times, and I'll try with your recommendation. YAY! More tuning! The never ending project. I need a driver.

[forevertrj]

what brand of 1000cc injectors are they, FIC?, ID? i would trust the datasheet, if they aren't FIC or ID injectors maybe the data sheet is wrong, my 1000cc FIC injectors at 14v and 1046CC calibration the deadtimes are 1.22ms, PM me if you want a picture of my datasheet for my Fuel Injector Clinic injectors, they are not going to be the same as yours but could provide some insight.

[forevertrj]

here is a datasheet of mine, sorry I couldn't private message the attachment, maybe I wasn't looking hard enough but I'm pretty sure you can't, admins why can't I private message attachments?

[CGF20BOOSTED]

Cool! Thanks. I tried with the 1.2ms and it was not working out for me. I have the map with the previous 700cc injectors. Looking at the dead times on those was like .26ms at 14v. And that was from my old tuner. So I went with the provided dead times (.56ms @ 14v) and got it working from the previous map after changing the settings. I can tell the difference in the part throttle and the N/A fuel tables are looking different from the previous tune I had made. When I get a chance at work, I'll alter the Battery voltage to dial in the injector dead times to insure proper change rate in voltage. I'll also try with longer dead times (like 1.0ms) later. I'll keep ya updated.

[forevertrj]

don't forget your fuel tables being at like 300 for -21.0 "of Hg will probably be reduced by 10 or more percent with the new dead times. I would try setting the dead times to what my datasheet says, and then dial in an overall compensation percentage, to get the AFR's back to normal once you see the overall % dial it back to 0% and then go into the fuel tables and apply that to all the NA fuel rates, select -25.9 Hg all the way to 0.4 psi from 0 - 9000 RPM and then hit ctrl+j then type in the percentage, in reality the lower Map fuel values will be the most affected by this change so you could apply for example -30% change to (-25.9Hg to -10Hg) then 15% to the rest and then you would just dial in the whole RPM range per Map pressure as a whole to dial them in correctly It's hard for me to write a great description of what to do but look at it like this


set dead times to my data sheet.
set the idle to around 1100 or 1300 rpm as a minimum to work with a more linear part of the injectors dead time, "since it's using more fuel"
set the overall fuel trim starting at maybe -10% and working your way up to -30% in 10% increments. until car starts and Idles at 14.7 as it did originally. then adjust the overall fuel trim by 2 and then 1 % until it's dialed in perfectly.
then go into your fuel maps select all the low range fuel values of all RPM's and map pressures starting from the column 1 and working you way up to column 6.
you can select fuel tables with your mouse by click+dragging or by selecting column 1 and rpm 0 position and holding shift and using the arrow keys to go right and down to select all fuels up to column 6 and down to 7000 RPM for non vtec lobe fuel table.
once all of these are highlighted pres alt+j i believe to bring up the fuel adjustment table, use the percentage negative or positive value that was in the overall fuel trim and apply this value to all the fuels that you have selected.
Now go back into the fuel trim tab and change the overall fuel trim back to 0% since you've change the actual map values.
apply the calibration and start the car up again, now you will be able to adjust each "columns" percentage individually from 0 rpm to 7000rpm to get you close to 14.7 if you've done all the previous steps correctly.

you've already tuned the fuel map so the whole map shouldn't change at individual rpm's in theory.
so when selecting columns select all the RPM's per Manifold pressure and adjust them as a whole by percentage.

once all of this is complete and the AFR's are where you want them turn on the AC and the blower motor, and watch the AFR's not move at all.
If they move at all weather it be richer or leaner, you'll need to adjust the deadtimes until they don't move at all or a little as possible.

AGAIN as i mentioned in a previous post, those dead times are for my particular injectors from Fuel Injector Clinic and are a ball tip high impedance (8-12 ohm) if your injectors are not the same type these dead times may be useless to you.

as injectors flow more fuel they hang because they are no longer producing linear results, so the larger the injector the higher your idle RPM should be 1000cc injectors probably should be idling around 950-1075 RPM, also if you have cams this idle value will increase, keep this in mind, the reason you bring up the idle is because of the linearity you want to keep the injectors are predictable and you can dial them in correctly.


hoped this helped (correct me anyone if i'm wrong ^.^)

[forevertrj]

So I took your map, modified the deadtimes, applied an overall fuel trim of 0%, modified the fuel tables to reflect the -40% to all the values and also removed an additional 20% (-20%) so when testing with this calibration use the overall fuel trim again to find a good idling AFR, i've also fixed your 100 rpm idle as i'm sure you meant 1000 but were missing a 0, also you've got some weird stuff happening with your fuel trim with IAT compensation, my car is tuned around 100 degrees fahrenheit intake air temp, but that doesn't mean you move 67 degrees to 0 and then 123 degrees to 2.4% it needs to be progressive and linear, if you find that your car starts running leaner as you move from 67 degrees to 123 degrees, add more fuel to the 123 degrees and the 159 and 212

Low MAP pressure
your tables
-40 -24 7 32 67 123 159 212
10.0 10.0 6.9 4.0 0.0 2.4 -0.5 -1.3

changes to be made
-40 -24 7 32 67 123 159 212
10.0 10.0 6.9 4.0 1.1 0.5 0.0 -0.9

your medium engine load and high engine load were changed slightly from their defaults and this is probably ok the only reason I suggested changes to your low engine load is because of how 67 degrees was a lesser value than 123 degrees and thats not how thermodynamic changes too fuel/air combinations work, needs to be mostly linear.

there are a few sites I've seen that you can put the temperature your tuning at and build the correct compensations for manifold pressure on the RSX forums, maybe hondata has some insight into this?

[CGF20BOOSTED]

I tried your modified map, and I had the over all fuel trim at -99% and it was still way too rich. It wouldnt lean up no matter what I did to the fuel table. I am at .93ms/14v and still tuning the fuel table at low load. I'll let you know how it turns out.

[forevertrj]

That's insane, and I saw the calibration so I know your in open loop, honestly I think o saw your 21.0 map column at 200 fuel? Try setting it all to 90 and the -25.9 inches to 40 and the -17 column to 130 it could be a good starting point those are my engines values

[CGF20BOOSTED]

This is what I ended up with so far. The oscillation has stopped at the part throttle between 1-9%. The throttle response is smoother and better. Here are a couple of data logs and my new calibration. Let me know what you think.

[forevertrj]

I'll take a look at it closer when I get home, I'll be home in 20 minutes

[forevertrj]

I smoothed out your lower fuel tables, yours were all jagged, try these and make the adjustments you need to but look at the linearity of them, and if you make a change to 1 position RPM and Map pressure, this fuel will also effect 1 quadrant on all sides of it, kind of like that PC game mines.

Fuel map
||||||||||Mani Pressure|||
RPM|||||||25.9|||||20.7||||||17.1|||||||10.1
1000||||||80|||||||90||||||||100||||||||110
2000||||||85|||||||90||||||||100||||||||110
3000||||||80|||||||90|||||||||100|||||||110
4000||||||85|||||||90|||||||||100||||||||110
5000||||||85|||||||95|||||||||105||||||||115

as you can see by manifold pressure 25.9 there is a dip in the #'s, basically making a jagged sort of line on the 2d graph to the right of your fuel tables, you would want it to look more like this because the system gives fuel (interpolated) between all the mainifold pressures and RPM's, so you basically want it to steadily climb into the fuel on a gradual slope and not going up and down the map should look more like this


Fuel map
||||||||||Mani Pressure|||
RPM|||||||25.9|||||20.7||||||17.1|||||||10.1
1000||||||80|||||||90||||||||100||||||||110
2000||||||82|||||||90||||||||100||||||||110
3000||||||83|||||||90|||||||||100|||||||110
4000||||||85|||||||90|||||||||100||||||||110
5000||||||85|||||||95|||||||||105||||||||115

Noticed how I changed that 85 to and 82 and the 80 to an 83, this looks more gradual and more linear, now also this is easy to see because th 20.7 map pressure is constantly 90, if there were rises and dips they would effect each manifold pressure above and below it. try the new map I have made you and adjust your fuels in this new manner of gradual incline or even just a straight line.

ALSO!!! most important I just noticed your running in CLOSED LOOP?!?!?! wtf, you can't tune fuel in closed loop!!! or did you change it back to closed loop after you had adjusted your fuels? also if you were having to add a correction of -99% to the overall trim this is because you were running in closed loop and can't do any effective tuning unless your in open loop.

[forevertrj]

After seeing this closed loop business, you will have to change your fuel tables again, and you will find that they are probably dangerously lean now in the columns from 1-6

[CGF20BOOSTED]

I enabled closed loop to see if the oscillation ceased. In which it did for the most part. I was driving around all evening through the night tuning in open loop. I'll try this map. Thank you for all you help sir. Highly appreciated.