help with KR in turbo l36 3800 series II tune.

[thebeewantsboost]

2001 3.8 series II camaro L36. stock motor, turbocharged, intercooled. 180 thermostat. ngk tr6 plugs, 91 octane

i'm running a 8# spring in my ebay wastegate. Through 1st and 2nd gear my autometer boost gauge shows 5psi. and in 3rd gear it creeps up to 10.
i'm not sure if the gauge is incorrect or if its the wastegate. so it could be 8psi and 13psi.
regardless, in 3rd gear i'm running single digit timing. and about .78 lambda. which is 11.5 afr i'm getting up to 6 counts of KR once i hit 3rd gear/higher boost.
any advice? is there anything in my scans i'm missing? i know i need to get my fuel trims adjusted for part throttle they used to be dead on. but its been sitting for a while. i plan on getting some race gas. and seeing if things change. i forgot to log maf hz in this scan so its missig the afr error histogram. but u can still see the lambda error in the table.

should i richen it up for higher boost?
use another step colder plugs?

[SultanHassanMasTuning]

there are many factors that cause boost creep. but its normal in lower gears that you do not build enough boost and engine load is lower and gearing is higher.

[thebeewantsboost]

I'm not concerned with the boost creep so much I plan on adding more boost and at that point I doubt it will be an issue. I'm just not sure why I'm getting KR. According to my Wideband I thought my Afr was good. I pulled the plugs last night and they had a yellow tint to the insulator and electrode. So it looks like I am running a bit lean. I'm going to add some more fuel and see how It goes.

[TurboGP33]

I read when GM pulls timing for torque management it shows up as KR reduction. Can't confirm 100% but it seems that way.

Try disabling torque management to see if that makes a difference. Though running lean is the first i would suspect.

I noticed my turbo can trigger false KR at a specific spool up speed, must be the frequency of noise triggering the knock sensors. Headers can also trigger KR, as well as any part of the exhaust resting on the frame or subframe.

[thebeewantsboost]

i believe all of my tq management is disabled, please let me know if you saw some that i missed. i bet you're right about it showing it as KR though. Something pretty important that i failed to mention. the night of the scan i did a 40-110 roll before i started scanning and at the end of third it stopped pulling and i got a check engine light. the check engine light code didn't show up on my scanner for some reason. however yesterday the same thing happened and it showed up as a p0300...(multiple misfire) this was right after i changed plugs so i'm pretty sure its not plugs, although i did put in some tr7's, which is another step colder after seeing the yellow on the previous. wires are new. coil packs are stock with 60k on them. plugs are gapped at .35 so i guess there's a slight possibility of the spark being blown out. with a higher booost level. but more than likely its probably a lean miss.... i've read different things about what to gap my plugs at. some people are gapping at .25 most seem to gap at .35 though

[TurboGP33]

All seems fine except for few things.

1- TPS Threshold to PE Mode is WAAAAAY too high for FI. Try this table below, as you are going into boost at stoich, not PE like you should.
Hot Power Enrich Enable TPS Threshold vs RPM.txt

2- Decel Enleanment tables are zeroed out but your Fuel Trim is showing excessively long decel conditions. Try setting that back to stock values extracted from an L67 SC 3800.

3- Extract Max Torque Timing tables from L67 file as well. You matched the HO/LO spark tables, now match the max torque table. But do not alter it, simply import one from a stock factory L67 SC 3800.

4- Fuel injector table Offset vs PW vs MAP is goofy looking. Import the proper table for your fuel injectors. Be sure to get the right table to match your injectors. This could possibly toss your tune out in the idle range. Just rescale MAF table again.

5- Maf Allowed Delta Airflow MAX is far too low just past 800 rpm. You can max that table though a rough curve above your max is better for transient fueling and MAF failure detection. For max decel you could set the max flow in the high end the same value as MAF Hz 11125 or 11250, you'll never blow off more air then what your ingesting at that RPM. Leave a little above to indicate MAF failure. This way a max possible reading will trigger MAF failure.

Try this table, works well for my fast spooling turbo. Should be sufficient for your SC.
MAF Allowed Positive Delta Airflow.txt

If you are curious at my finer tunings, please feel free to inspect my file. Even though it is for a 3100 turbo, some tables like IAT spark correction is helpful since we are both boosted.

turbopcm40.hpt

[thebeewantsboost]

i appreciate your response so much! last night i was thinking i needed to change my pe enable. most of the time when i see boost its because i'm WOT throttle but i gave it some throttle yesterday when i was more or less cruising and went into boost and saw a bit of KR. i'm at work right now so i don't have time to look over everything. I am actually turbo charged and not SC. does your advice for copying those tables still stand? i imagine its still a good idea.

i pulled the plugs after that run and it looks like i was running a bit lean up top. i retuned my part throttle and my PE. yesterday and richened it up. and it runs alot stronger it seems the only KR i got was when shifting into 4th. i guess i shifted to slow and it actually commanded a leaner afr for a second. any ideas on how to fix that? i'm going to go over everything you mentioned, i'm sure i will have more questions... i can't thank you enough for your response!

[TurboGP33]

Here is a spark map for turbo 3800 at about 8-10 psi. As well as an Open Loop AFR Adder table so you don't run way to lean if car is still warming up on a cold morning before going into closed loop feed back.

Open Loop AFR Adder vs Cylair vs RPM.txt
HO Spark Map for Turbo 3800.txt

The lowered TPS Table will REALLY help out a lot. It will also speed up spool which is good. Here is my latest TPS HOT PE which is slightly more fuel saving, not going into PE until about 75 kPa load and above. Which happens to be the curve I charted. If you need to bump it up or down, skew the whole table as it should be a universal curve (rpm vs kPa vs TPS)

Hot Power Enrich Enable TPS Threshold vs RPM II.txt

Good luck, post up the log if you need any further help.

Here is a video of me getting boost in rev at idle, bad a$$.




About the shifting, are you manual? If so, there are some tweeks you can use on your spark maps. And running lean for a brief moment between shifts....that can be fixed with DE (decel Enleanment values)....which is not tune until the rest is good, then you tune in AE and DE slowly over time until any lean and/or rich spots disappear during transitional fueling....especially when discharging pressurized air, which is a HUGE change for the PCM to make. The larger the change, the greater chances of AFR going out of whack.

Once you get the basics tuned up I can advise you how to further fine tune. But you need to be somewhat knowledgeable in Microsoft Excel or Kingsoft Spreadsheets (free version).

Excel tuning will take it to a whole new level.


Not a problem buddy, I don't mind at all.


EDIT: BTW, you may have False KR from the turbo. My turbo would trigger maximum KR when the turbine noise matched the frequency for KR detection. So I had to shut off KR protection from 4400 RPM and lower. Which sucks but isn't too bad. My spark tune is gentle then most.

[thebeewantsboost]

prety cool video, what car is that? looks pretty sick.

my car is a 5spd

1) PE TP is fixed

2) used data from a GTP to fill out the DE
pathetically i'm not sure what should actually happen here but if i decel in gear it goes lean. but if i put it in neutral it goes way rich.

3) i used all of the timing from the GTP and am able to run all of it.(@current boost levels) HO for both of my tables(for now)
and the max TQ timing table. The timing i had in my original tune was really low because i was battling false knock from the sound of compressor surge.

4) i can't find the correct fuel injector data for "offset vs pw vs map" i'm still on the prowl however. the stuff thats in there is what was there.

5 i haven't touched the maf delta airflow yet. i'm to tired at this point to completley comprehend it , and i've been trying to research everything as i do it.

car is running good though.

[TurboGP33]

If the engine is running good, chances are the MAF Delta tables are good enough. Just remember them in case you need to hunt down an issue.

I'll get the injector table data for you. 42.5 lbers aren't too common but it's out there.

The spark map will probably need some advance in the low throttle end. Automatics vs manuals have different driving needs on the spark map. I have driven my 5 speed on an automatic file for months, wasn't bad at all. But the tweeks for a manual really helped pick up the low end torque.

96 Grand Prix......an odd ball year/model. I had to modify SOOOO much just to get it to work right. 97 PCM, had to drill/weld stuff for the clutch. At least the subframe had provisions to bolt up the trans mount for the HM-282.

What I did was compared same year/model cars with auto and manual. I calculated the percentage difference between the two and then applied that to my auto spark map. Turned it to a manual spark map. So for what ever reasons the factory changed the spark map between trans types, the same changes are made to my map. It worked pretty well.

Glad everything is doing well for you. My turbo also triggered LOTS of false knock. As of right now, I turned it off. I suppose I'll have to tune the old school way, copper pipe to the block running into the cab for me to listen. Or I might rig a microphone to the block and use a spectra analyzer to view anomalies in the noise signature.

[TurboGP33]

Checked the log, good good. Nice long pull there.

Your DE and AE are out of tune, but that is expected.

Best way to tune that is log a long drive and try to maintain a steady throttle, no gentle transients. ie- 0 to 25% tps and hold it for atleast 1-2 seconds. Same with decel, drop to 0TPS with out too much delay and keep it 0 TPS for 1-2 seconds. Repeat at many different throttle ranges.

Then view the log and look at each event of throttle. 3-5 ticks after the TPS change the wideband will have a lean spike.

Go into the tune and adjust these two tables.....AE vs TPS Max Delta and DE vs TPS Max Delta.

If your throttle in of 25% TPS has a lean spike, add a little more at that throttle range in the DE table. Like wise, any throttle event IN and you have a lean or rich spot, adjust likewise. It is best to have the AE richen up and land at 12.4 AFR range for power, have it stay stoich for fuel economy (no fun...lol)

Futher, once you get the AE/DE under better control, you can view the tail end of DE and see if it is too lean/rich at different VAC levels from the MAP kPa sensor. If you notice it is too lean/rich at 20 kPa (only seen in decel) then alter the VAC table in the Accel Enrich tab, decel section.

Even yet further, the decay profile can also adjust the amounts of AE/DE, but is better tuned to shape the curve, incase either AE/DE is too much at first but is not enough part way in. ie- AE at 40% TPS has a rich spot then a lean spot. The decay profile needs reduced in the beginning and raised towards the end.


But all this AE/DE tuning is only done AFTER the MAF and everything else fueling and spark is finished or almost finished. The car will run good with a sloppy AE/DE tune, but it will run like a well oiled machine with proper AE/DE. A great AE tune will produce a very crisp throttle and lots of instant power.

[thebeewantsboost]

They are actually the ford green top 42lb injectors. i've read that the specs for them are in one of greg banish's books. and i found the rest of the information for them online just not anything for that table. i've gotta go mow grass ; / just wanted to update that so you didn't spend time looking for the wrong thing. i've got a post on 3800pro.com asking for the specs also.. thanx again

[TurboGP33]

They are actually the ford green top 42lb injectors. i've read that the specs for them are in one of greg banish's books. and i found the rest of the information for them online just not anything for that table. i've gotta go mow grass ; / just wanted to update that so you didn't spend time looking for the wrong thing. i've got a post on 3800pro.com asking for the specs also.. thanx again

Sure thing, also one more detail. Since you are using a stock map sensor, your tune will always be slightly off until after 1st boost. Or the other way around, in tune until 1st boost. To confirm if you have this issue, log the BARO sensor (which isn't a real sensor, just a max reading from the MAP sensor).

It will register atmospheric from start up. The PCM will register this as the atmospheric pressure. Once you hit boost, the max at 104 kPa becomes the new barometric value, thus tossing the tune out. If you didn't realize this your tune would be giving you mixed signals and probably confuse you some till you figured it out. Took me 2 months before I realized what was going on. I just ignore data before 1st boost, or don't even log until after 1 st boost since start up. Fuel trims will keep the idle happy for inspection.

[thebeewantsboost]

i've read something about the tune being off until you hit boost... that sucks. i wonder if anything can be done about it. have you heard of "tiny tuner" i'm not sure if it will support your pcm but its a free editor. it may have a solution but i doubt it. do you know what it effects? if it effects fuel trims, which i doubt it does, but if it does, it would make so much sense to me. as far as getting mine tuned correctly

if the Maf delta isnt setup correctly for the amount of air being put into the motor, and the computer freaks out. does it put you in reduced power mod/limp mode until you power off/on the car? i've had that happen once before in the past. not recently though.

i need to work on my fuel trims more.
i've also got another issue, dont know if its mechanical or tune. i'm getting some misfires. i haven't scanned them but i got a solid check engine light earlier (HPTUNERS WOULDN'T PULL THE CODE) but it did reset it. On the highway after giving it full throttle in 3rd from a 50. i did get Knock retard. and it looks ilke it was because it was running lean...i can only assume its because of my maf tune. it was showing low 12:1 afr. during the i didn't feel the misfires. when i got to my destination it seemed like i was still hearing slight misses tho. its at the beginning of the scan.. you can't see my wideband readings, can you?
so yeah i'll check that out and try to get my MAF more in order before carrying on. if you have any input, its always appreciated, and if you have a paypal, i will gladly float you a few bones for the help you're giving me! did you learn from the forums, books? class? trial/error?

[thebeewantsboost]

still investigating my misfire problem. seems i'm getting misfires on cylinders 1,3, and 6 while cruising. i also think i found out why i've been chasing my tail while trying to tune the maf. my wideband is reading leaner than my narrow bands. with negative fuel trims (0, -3) while my wideband is reading 15ish. i'm not sure if i have an exhaust leak. its possible where the bung is welded in it could be leaking. but regardless i think im going to try tuning the maf cruise areas by narrowband for now. and maybe smear some jb weld on the welds where the bung is.

Update: I do have exhaust leaks at both sides of the turbo. AND a boost leak. Boost leak will be easy enough turbo won't be so much fun. I'm just glad I know what's causing the difference in o2's. Also. I guess i hadn't set all of my "bank 1 sensor 2 o2 sensor"(post cat) codes to "no report" i was getting different codes for that and didn't even realize. so when i thought i was getting misfire codes that hptuners wasn't reading, i believe that's all it was.

[TurboGP33]

oh yeah, any leak is going to greatly reduce performance. Especially an intake leak, that can cause the turbo to spool up faster and faster trying to pressurize the intake. It is hard on the bearings and the turbo better be well balanced to push at a higher spool rpm.

When you get that fixed I can offer some advise tuning the MAF.

Specifically, turn off fuel trim learning and reset the trims in the scanner. This way there is zero fuel trim action messing with your log. Also, your wideband sensor, try adding a multiplier to the sensor in HPScan. Whatever gets the reading accurate. You can log both wideband and narrowband. Make a histogram to see the average wideband reading at 450mv. Then multiply as needed to get 14.4 at 450mv. I say 14.4 because most (if not 99%) of US gas stations sell E10, which is more like 14.4 stoichiometric. You should put that in the stoich variable under fuel. It will help the trims.

Also, with the MAF tuning, make a histogram that logs the AFR difference vs MAF Hz. You'll have to create a custom histogram.

Misfire could be too much spark gap (I like 0.045" gap versus the stock 0.060") and get a plug 1 range cooler then stock as well.

Could also be from the rough tune. Other options are the plug wires are not fully seated or the plug wires broke from removal and created a high resistance spot. Doesn't become obvious at low load/idle. Only rears it's ugly head under full load.

You might see a spark jump from the cable to a ground when you rev it up hard from TB. That indicated high resistance somewhere in the circuit.

[thebeewantsboost]

i'm going to order the new gaskets soon. and get a coupler maybe today. i'm running 100% gasoline. and fill up at the same place each time. i wanted to avoid the variables there.
i've got the afr error histogram and a ltft+stft vs maf hz histogram. it was frustrating because when i tuned with one, the other was off and vice versa. hopefully that will all be taken care of. and i have adjusted my wideband pid to read what my gauge is reading.
i'm really only feeling the miss at idle but it only records it under load.. plugs are gapped at .035, which is what most people recommend. i've thought of closing the gap a little but i wouldn't think the gap would cause a miss at idle. One of the plug wires seemed loose at the coil. i haven't noticed if it was missing or not since i fixed that.

" You can log both wideband and narrowband. Make a histogram to see the average wideband reading at 450mv. Then multiply as needed to get 14.4 at 450mv" <--- i don't quite understand this. i get the 14.4 part but not the histogram and multiplying stuff...

[TurboGP33]

i'm going to order the new gaskets soon. and get a coupler maybe today. i'm running 100% gasoline. and fill up at the same place each time. i wanted to avoid the variables there.
i've got the afr error histogram and a ltft+stft vs maf hz histogram. it was frustrating because when i tuned with one, the other was off and vice versa. hopefully that will all be taken care of. and i have adjusted my wideband pid to read what my gauge is reading.
i'm really only feeling the miss at idle but it only records it under load.. plugs are gapped at .035, which is what most people recommend. i've thought of closing the gap a little but i wouldn't think the gap would cause a miss at idle. One of the plug wires seemed loose at the coil. i haven't noticed if it was missing or not since i fixed that.

" You can log both wideband and narrowband. Make a histogram to see the average wideband reading at 450mv. Then multiply as needed to get 14.4 at 450mv" <--- i don't quite understand this. i get the 14.4 part but not the histogram and multiplying stuff...

What I meant, you already fixed. Calibrating the wideband data log to the sensor. A multiplier is a better way to skew the readings instead of addition. Doesn't matter too much, just a point of accuracy.

Make a histogram that averages wideband readings. The top axis is narrowband o2 mv readings. (0-50-100-150-200-250-etc....-950-1000) Then you will see what the average wideband reading is against the narrowband. This will give you a good calculation of your fuels stoich against the wideband.

Mine is 14.43:1 to get a solid 450mv NB. Stoich for the fuel I use is 14.35:1. Means my wideband is about 0.092 off. (good enough for me as the AFR fluctuates 0.2-0.3 at times in idle/low load)




Once you know exactly where stoich is on the wideband (which is important IMHO), you can be sure the WOT fueling readings is accurate.

[TurboGP33]

The fuel trims, I do believe you multiply the STFT by the LTFT to get the over all fuel trim.

LTFT: 1.1
STFT: 1.2

Adding gives 1.3% trim but the real number is this, 1.2% STFT from the LTFT at 1.1%... 1.1 * 1.2 = 1.32 means the real fuel trim is 1.32%

Not a huge difference, but accuracy is what you make of it.


Then you can multiply that number by 0.01 and then add 1. This number can be multiplied against the MAF table to correct the MAF.

(1.32*0.01)+1 = 1.0132

Negative numbers also work in that formula. (-0.98*0.01)+1 = 0.9902

Using those equations/process, you can make tuning the MAF really easy with excel. Log the run, pull the numbers off the scanner, plug into excel doc for tuning the maf and reflash.

This way I can drive around town, pull over and adjust the trims on the fly and be back on the road in a few minutes. Makes tuning the MAF much quicker.


Hope that isn't confusing.

[thebeewantsboost]

I've always read that you add the ltft and stft. if it is multiplication it would definitely make a big difference when it comes to larger larger fuel trims. makes me want to just log stft.
my cars been down for a while, a brake line melted and i've been fixing the exhaust and compressor leaks also. should be good to go by tonight/tomorrow.