FAQ: How to tune the PCM? (Series II SC)

A couple things I see with that scan:
You are pig rich. AFR cmd is 11 -> 10 that's way too much fuel.

O2 is weird - a steady 12.6? No change at all? Do you have a rear O2 sim? You might be reading the wrong O2 sensor.

Shifting can absolutely cause KR and it's generally okay to ignore. Based on both KR spikes I see, you should shorten the recovery rate.

Your STFT is 0 the entire time. That's not right either.

What kind of scanner is this? Are you using a WB O2 sensor?

Also - PM sent.

First thing to realize about KR: it is not RPM dependent. It's easy to assume knock is created as RPM increases, but it's actually not the case (I think it's assumed because boost rises with RPM). Actually though, increasing RPM decreases load, helping the engine breath, which lowers knock.

Knock is more dependent on octane, boost, timing, A/F mix, and most of all, load.

Load, in simple terms, is when the engine/drivetrain is using the most torque. In our cars this occurs at low RPM. Our tall gearing has a lot to do with it. For given amounts of boost, A/F, and timing, you'll always see more KR at very low RPM under high loads. When you shift up a gear at WOT you increase load abruptly, lowering RPM. This is what I see happening in your 2nd graph.

Another example would be in OD climbing a hill. If you're not tuned for it, it'll be KR city. Drop to 3rd gear on the same hill and what happens? RPM increases, load decreases, and KR goes away.

One more example: pulling a trailer, the added weight is an extra "load". A car tuned for no KR without the trailer will see quite a bit more when pulling. RPM has nothing to do with it.

Knock is also timing dependent, so at any given instance if KR, retarding spark timing will help. This is why the PCM counters knock with KR - it's the quickest and surest way to make the adjustment electronically. Look at your graphs. After the KR steps, you can see timing is instantly retarded. This is partly why KR starts to roll off.

The other reason KR rolls off in the graph is because you are increasing RPM, which decreases load. The blips that don't follow upshifts are probably the result of A/F or timing issues. These can be tuned for correction.

Timing and A/F ratio can be tuned to a point, but there is an optimal value for each. Once you've gotten to that value it makes sense to stay there. Example, an A/F ratio of ~11.5:1 is ideal at WOT in my set-up. I can keep dumping fuel at it, but a 9:1 ratio isn't gaining anything power-wise, and it wastes fuel. Same with timing. There is a point where optimal timing produces the perfect amount of torque at a particular RPM. Beyond that point the efficiency is lost. No reason to push any further.

Your KR issue is the main reason I eventually started to support breathing mods and shift tuning (including minimizing TCC-lockup), rather than focusing on high torque and increasing boost with smaller pulleys. Why increase load and KR when you can maximize power through RPM and air flow, automatically addressing the knock issue. This is also why I don't have an IC - don't need one.

deekster_caddy wrote:

A couple things I see with that scan:
You are pig rich. AFR cmd is 11 -> 10 that's way too much fuel.

O2 is weird - a steady 12.6? No change at all? Do you have a rear O2 sim? You might be reading the wrong O2 sensor.

Shifting can absolutely cause KR and it's generally okay to ignore. Based on both KR spikes I see, you should shorten the recovery rate.

Your STFT is 0 the entire time. That's not right either.

What kind of scanner is this? Are you using a WB O2 sensor?

Also - PM sent.

Yeah,,,,, dont mind that WB O2. I dont have one installed and it just reads 12.6 the whole time.

I will take it off for the next scan I post.

charlieRobinson wrote:

deekster_caddy wrote:

A couple things I see with that scan:
You are pig rich. AFR cmd is 11 -> 10 that's way too much fuel.

O2 is weird - a steady 12.6? No change at all? Do you have a rear O2 sim? You might be reading the wrong O2 sensor.

Shifting can absolutely cause KR and it's generally okay to ignore. Based on both KR spikes I see, you should shorten the recovery rate.

Your STFT is 0 the entire time. That's not right either.

What kind of scanner is this? Are you using a WB O2 sensor?

Also - PM sent.

Yeah,,,,, dont mind that WB O2. I dont have one installed and it just reads 12.6 the whole time.

I will take it off for the next scan I post.

Oh - just saw the 'real' O2 sensor. That's not much better. Seems like you are way too rich. Kind of normal for a starter tune though, so no worries there.

Those KR blips are small and go away quickly, plus are right at the shift, so I'm not majorly concerned with them. Just the KR sensor doing it's job. Did you read through the tuning guides?

Not yet,but thanks for the reminder, Derek. I will try to review them tonight.

They should make more sense to me know that I have some EXP.

Also, do you guys use the 'scan thingy' tool?

I used it last night for my MAF tuning and timing tables and the results it spit out looked legit.

I uploaded the new tune before my commute to work. I am going to scan it as soon as I can then get back to the tweaking.


SO GLAD I MADE THE DIVE AND GOT THE HPT PRO. THIS IS SO REWARDING AND SO MUCH FUN.
INSTANT POWER AND MPG GAINS AND I DONT EVEN KNOW WHAT IM DOING.

I will be ordering a wideband O2 sensor this weekend.

Now that I have HPT I don't use the scan thingy anymore. Once the tune is close I prefer to make changes by hand. It's typically only 3 or 4 cells I want to make changes to at a time on the big tables, once the main tuning work is done.

I'm psyched because HPT just gave me beta access to work on the Volt!

deekster_caddy wrote:

Now that I have HPT I don't use the scan thingy anymore. Once the tune is close I prefer to make changes by hand. It's typically only 3 or 4 cells I want to make changes to at a time on the big tables, once the main tuning work is done.

I'm psyched because HPT just gave me beta access to work on the Volt!

That's awesome. Did you have to spend credits to unlock the Volt or was it a beta free trial deal?


Is this another example of transmission noise causing KR? Would increasing shift pressure prevent this?

I havent even started tuning the transmission yet.

No, it's not the transmission. It's the increased load after the shift, on top of a lean A/F mix (O2 value is low, and LTFT is +).

Does IFR determine LTFT? It's pulse width right? so lower values = more fueling?

IFR is one way, but setting IFR is better used for skewing rates to the middle cylinders, which run hotter. The right way to tune fueling is to modify MAF table. When you go WOT (open loop), your O2 stops sending feedback to the PCM, so it reads directly from the MAF table for fueling.

What you want to do is move your scan preview slider to the area where you have KR. Note the MAFF Hz values. Find those values in your MAF table and increase by 1%. Run again and scan. See if the knock goes away. If no improvement, you may consider cutting timing at the RPM where the KR is occurring.

I understand that the MAF table is used when we are at WOT but what determines LTFT when are at closed loop and just cruising?

That's what im trying to tune now.

I changed my IFR back to stock after INCREASING the values slightly and scanned this morning and now my LTFT is 0 to -5. So I think I am on the right track.

Sorry, I was thinking of injector skew perfect re: middle cylinders. 3 & 4 at 100, 1,2,5,6 at 96.

For IFR, mine start at 8% over stock (0 kPa), then 7% (10 kPa), 6% (20 kPa), and 4% for all the others above that. This is the way DHP programmed the 1.5 BIN, and I never changed it.

deekster can confirm, but my understanding is fuel trims start with the MAF table and then use input from O2 when at idle/cruise. So if your LTFT is off, it means O2 is telling the PCM to correct by x amount. Log the LTFT at specific MAFF values, then adjust the MAF table at those values according to how your rich or lean your LTFT indicates. For example, if LTFT is +2.3 at 4000 Hz MAFF, bump up the fueling in the table by 1% at 4000 Hz and do another run. You'll never get it perfect, but after about 50 tunes, my LTFTs are mostly between 0 and -2.3. Don't worry about decel numbers - they are more difficult to tune.

Be careful when you reset fuel trims, LFTF will set back to zero at start. Give it a few miles before you make any conclusions.

Fuel trims are based off your MAF Hz and O2 primarily. LTFT shows you the average for that particular point in the system (MAF table, VE table, MAP sensor, ECT, etc all taken into account and more). It's actually several tables working together, but the MAF and current O2 have the biggest impact. LTFT shows you the average correction needed. STFT shows you the immediate corrections (with LTFT already being accounted for).

There is always a little load spike during a shift. Don't fret over it. It's not "Noise" - it's actual KR from a massive change in load. Increasing shift pressure and shortening shift times will help improve your shift, but may not have any change on the shift KR. It may actually make that a little worse (it's a big load and shifting faster may make that load even bigger). Shorten the KR recovery time a little and don't fret over it. It's not a big deal.

deekster wrote:

There is always a little load spike during a shift.

Just to add, raising the upshift points will lessen the severity of the load change a little (make sure your upper valvetrain is up for it). Also programming downshifts to happen sooner will keep you out of low RPM/high load territory, also makes throttle more responsive.

Just optimizing the PCM's downshifting is a great way to tune out KR, seriously.

Is the AFR commanded a value that can be set in HPT or is it a compilation of different sets of data and dynamically generated?

If there is a table for it to set it, I cant find it.


How we lookin now? I am much happier with the LTFT. I think i will move on to transmission soon and see if i can eliminate the rest of the KR.

I went and did it...

My car won't start..


Everything was going fine. I mean perfect. Then on my last drive my fuel trims go wack. -20s for Short and LTFT.

I have no idea why. Everything was just perfect. Then the engine started to bog down and choke. Couldnt hold idle RPMs and the engine would shut off. I restarted it and gave it gas but as soon as I let off the gas RPMs would fall to 0 and shut off.

I started scanning and the MAF is going CRAZY. like 100-10,000 and all values in between. Just flipping out.

Nothing changed from my last tune and everything was PERFECT. What have I done!?


I tried reflashing back to stock OEM tune. turned car back on and same symptoms.

 

Ok. Here is the scan where it all went wrong. MAF is reading a lot higher than what it should be and then look where it starts going CRAZY. The MAF values are all over the place. At that time when the MAF starts freaking, the STFT goes -25 and just stays there!

Any codes stored?

No codes. I just pasted a shot of the scan above.

MAF just went nuts and looks like it affected fueling with its craziness.

After the MAF goes crazy the STFT never fully recovers. It comes back and normalizes around 0 but then dips back -10 to -25.

I tried resetting fuel trims but that didnt help.


Here's the scan file that captures where it all went wrong. I will greatly appreciate it if you guys would check it out and let me know what you think.

The action starts at about 9:40
http://www.datafilehost.com/get.php?file=62254e72


If the MAF sensor went bad all of a sudden??? I should have a code thrown for that right? I will check tomorrow. If I unlplug the MAF sensor, the VE should take over and it should run 'better' right?
I will try that too.


I will check under the hood for anything out of order too
     

Revert back to previous tune and see if problem persists. Don't tinker under the hood until you've ruled out the PCM programming - it's very possible since you've been changing things recently.

My MAF was bad. I ran a week with it unplugged on the VE table. Got a new MAF last weekend and I am back in action!

I was running my 3.6" with 0 KR and I noticed my timing table was stupid high. I dropped to the 3.4" pulley and I am seeing 2-3 degrees of KR at the peaks. About to tune that out right quick!

They dont call it digital horse power for nothin!

Here is what my current tune is giving me with current mods in sig.
3.4"
104s
180-T°
3.5" FWI
PLOG + Ceramic Crossover + 3" Downpipe + O2 emu. - Cat. + Aero Resonator



Still have ~2 degrees of KR to get rid of. I havent even touched fueling/PE yet.

How high is your timing in 3rd gear at WOT?

I want to tune the 3.4 and 3.6 as best I can then take them to the track and see how different the times are. The 3.4 just pulls so hard. I read that timing > boost but man. If there is a middle ground, I want to find it!

That doesn't seem like too much timing at all. I have a lot of different stuff of course but I'm pretty sure I'm in the upper teens at WOT when things are right. If you do the same thing with the 3.6 pulley, you have no KR? I'd try to confirm that. Curious - how is your injector duty%? 10.7 commanded is pretty darn rich, I'd expect to see 9xx coming from the o2 at that point if it's actually getting that AFR but I dunno...

Codith, how do I scan for injector duty cycle with HPT? I dont see it as a PID and when I try to scan injector pulse width (INJ B1 m/sec) I never see any values.

Thank you!

Interesting read:
http://tuning.wikispaces.com/SD+vs+MAF

Quote :

So why is everybody saying I have to tune my car in SD before I tune the MAF?
The PCM for most Gen III GM motors uses a blending function that makes the vehicle run on a sort of hybrid system which uses both Speed Density and MAF functionality. The amount of blending can be changed through tuning, but as the factory has it set, the VE tables (SD fueling values) are referenced all the way up to 4000 RPM, so as you can imagine, there’s quite a bit of driveability to be gained by tuning the VE at least up to 4000 RPM, as anything else is providing a band-aide solution.

The principal reason behind tuning in SD is the ability to arbitrarily quantify it's correctness, using oxygen sensors (Narrowband for stoich and Wideband for non-stoich ranges). If the oxygen sensor says we get ideal ratio to the fuel injected, we know it was because our estimation of all the airflow characteristics was also close to perfect. There are many other factors (intake and coolant temperatures, RPM, pressure in the manifold, displacement) but they're either constant or easily measured and continuously updated with sensors. Volumetric Efficiency is dependent on RPM and MAP, and as such, it must be obtained empirically, and that is the whole point of SD tuning. Once we obtain full, correct VE table, we have a solid base for tuning all derivative modes of operation (WOT being the most popular and satisfactory), and as such it will provide a base for ALL future airflow and fueling decisions. As such, it is absolutely crucial to have it perfectly calibrated.

SD tuning is important, but not as a goal as most people think about it, but as a base for sucessful tuning of every other aspect, as they all depend on airflow characteristics of an engine.