ECMlink LinkV3: BoostEst and boost gauge not equal {attached log}

[razrman8]

At least I have an explanation for it now. It was really killing me that I didn't.

I will set them all to 0 then do as you stated. Had to put it back the way it was to get home.

After that is set, should I drive it to get some MAFcomp calibration going?

A lot of that is probably due to inaccurate deadtime values, which haven't been corrected yet.

Go into the InjBatAdj DA table, and change the following and then save it (with the car off, ignition on):

7v - 3189
9v - 2443
12v - 1457
14v - 1103
16v - 831
19v - 762

Go to the fuel tab, and set:

Global deadtime: 0 usecs

...and assuming your base fuel pressure is around 43psi and you are running pure gasoline, set:

Global fuel: -61.7%

Zero out the MAFComp sliders and fuel trims, and run through the "drive a while/run the CombinedFT adjust tool" process again and see if it's closer.

And still do this?

 

[Calan]

After that is set, should I drive it to get some MAFcomp calibration going?

Yep. Set those injector values, reset your fuel trims, zero MAFComp except for bumping up the 50hz slider, and see if it will start and idle. If it does, you're off and running as before.

Just an FYI -

As a shortcut, you can also just watch CombinedFT while manually changing the sliders; once CombinedFT is hovering around zero for any given slider, you're done. You just need to get the airflow to be steady at each breakpoint by holding the throttle when the airflow equals a value that a slider corresponds to (50hz, 100hz, 150hz, etc).

You can find this by logging the "MAFRaw" parameter. You will probably be idling at something like 35-50hz; if it's somewhere in between sliders, just interpolate between the two sliders so they "share" the correction. You can then hold the throttle down to raise the RPM and MAFRaw value, and do the same thing with higher sliders. Just don't go too high in RPMs (something like 3k max), don't rev it for long periods, and keep an eye on coolant temp. This should get you dialed in up to 150hz or so, and let the car idle pretty well. Then you can work on the higher sliders as usual with the MAF comp tool.

Or just let the tool do it all if you are more comfortable with that.

 

[razrman8]

I'll give it a shot. I like trying to learn as much as possible about this so I don't look like a complete noob.

 

[Calan]

I'll give it a shot. I like trying to learn as much as possible about this so I don't look like a complete noob.

Yeah I picked up on that. I like helping people (where I can) that want to learn and help themselves...which is why I'm here in this thread.

Check out some other tuning threads and you'll start seeing a pattern of when some of us actively participate and when we just stay out of it. Weird how that works huh?

 

[razrman8]

It is indeed

Okay, so I zero'd out the sliders. All but the 50hz one, got it idling. Let it idle for a bit, used MAFComp to adjust it, logged. Drove it a little, MAFComp. Drove, MAFComp.

I don't know that I really got anywhere. They are all in numerical order except the last two are flip flopped

. I notice though, before I went into the InjBatterAdj and changed that, the slider wasn't maxed out in low Hz.

Right now, my idle is really lumpy, and my CombinedFT are still really high.

 

[Calan]

Looking at the last log, you still haven't set the values in the InjBatteryAdj table, so your deadtime is still off (actually way off right now).

You need to set them to this:

7v - 3189
9v - 2443
12v - 1457
14v - 1103
16v - 831
19v - 762

...and then click "Copy to ECU".

 

[razrman8]

Wow, I am sorry. I swear this is all making me go crazy.

Here is a log with those adjustments in. CombinedFT is way better, below 0.

So do I take this and start MAFComp adjusting? Afprev seems high, but I am guessing it has to do with MAFcomp.

Figured I would note that I now get a AFR on my wideband vs it going pig lean and displaying ---

 

[Calan]

Ahhh... much better.

Zero your MAFComp sliders again, bump the 0 and 50hz sliders up to around 20% or so (since you know you will have to compensate for that PCV breather leak), and try again. You should be much closer now.

AirflowPerRev is a bit high, but don't get too wrapped up in it. Just watch what it does as you adjust MAFComp for now.

It's an iterative process...you're probably starting to understand my comment about it taking a lot of time to get things calibrated.

It's also about paying attention to details; a simple thing like not setting those injector values could have you running in circles and scratching your head for hours.

 

[razrman8]

I am indeed understanding that comment more and more as we go. I also appreciate you taking your time out to help me as much as you have. You are helping me understand this more than I thought that I would have. Thank you.

As for trying again, you saying I should let it idle, MAFComp adjust, then try driving it to get some MAFComp adjust in?

 

[Calan]

I also appreciate you taking your time out to help me as much as you have. You are helping me understand this more than I thought that I would have. Thank you.

NP. It's my "job" here, and you are holding up your end of the deal.

As for trying again, you saying I should let it idle, MAFComp adjust, then try driving it to get some MAFComp adjust in?

Sure.

Work on the idle, then some cruising around. Just make it routine to stream some data and run the MAFComp tool every time you go to the store or whatever, and eventually it will be pretty solid. (Remember to reset the fuel trims when you run the tool also). BTW - There is some excellent info on how the fuel trims work on this page, if I didn't mention it yet.

Once you get the hang of this closed loop MAFComp calibration, you can start manually adjusting the higher airflow/open loop sliders, by using WBFactor; which is the difference between the AFR you should be at (from the OpenLoopMaxOct DA table), and the actual AFR being logged by your wideband. It's the same process, except you do it manually with the WBFactor instead of using the automatic tool with CombinedFT.

I'm about to head out to the garage and dive head first into dropping my tranny and replacing a fried clutch, but I'll be in and out and checking this thread every now and then.

 

[razrman8]

Alrighty, I am on my way to get started. Thanks again and good luck!

Just got back from doing some cruising, man I was starving

These adjustments are really looking to be doing some good. It seems as if my AFRest and wideband gauge are closer (while cruising).BoostEst is matching the gauge when in vacuum. CombinedFT are starting to look real good as well.

Here's a partial driving log with idle at the end. Only thing I am seeing is AirFlowPerRev being off a little, other than that, looks way better than before.

 

[Calan]

Very cool. It will just keep getting better the more you work on it.

I'll look at the log here in a while. (I only have to drop the suspension, pull the axles, and remove the bell housing bolts and the tranny is OUT!

 

[razrman8]

Badabing, do work!

 

[boost97gst]

I recommend this thread be stickied. Damn, craig, u are the man. Ive learned more in 3 pages than in hours of reading on link.

 

[razrman8]

I agree, Craig's the man. This thread has been severely helpful in broadening my knowledge on DSMlink and calibration

 

[Calan]

Thanks, but I was just trying to find an excuse to not go drop my transmission.

Glad to help. There is a lot of info to digest, and it's difficult to connect all the dots sometimes without having a "big picture" view of what is going on.

 

[razrman8]

Precisely. Now I get to move onto the next part.

 

[MY1GDSM]

Are you 1000% sure you don't have an exhaust leak before the wideband?

 

[razrman8]

I can check again.

 

[razrman8]

Test pipe to cat back.

 

[gbell]

Test pipe to cat back.

The rear 02 sensor doesn't really matter for tuning as it is the front sensor in the 02 housing that is used for closed loop, so you need to check the exhaust manifold and 02 housing connections for leaks.

 

[Calan]

That 500hz spike in your log is probably an anomaly. That airflow point is close to where you would be switching into open loop, so the MAFComp tool probably just picked up one or two random values in that area. I would zero it back out until you get more data at that point.

I agree with MY1GDSM; something is up with your wideband. Either you have an exhaust leak, or the sensor definition in ECMLink isn't matching what your box is sending out. Since your fuel trims look decent on the stock O2 sensor, the leak would most likely be between the front O2 sensor and your wideband. That doesn't leave many leak points, which is why I think it may also be something going on with your wideband itself.

BTW - If you are wondering why I'm saying this, notice the difference between the CombinedFT value and your wideband error value ("WBFactor") at the end of the log while idling. CombinedFT says your pretty closely dialed in at idle, but your wideband says you are running over 20% lean. They should agree with each other during closed loop, within a few percentage points at least.

 

[MY1GDSM]

The rear 02 sensor doesn't really matter for tuning as it is the front sensor in the 02 housing that is used for closed loop, so you need to check the exhaust manifold and 02 housing connections for leaks.

FOR TUNING? The rear o2 is the wideband and the only thing that realy matters as far as any real tuning IMHO... and the reason I asked. Yes the front o2 is controlling the car. But it seems for the most part to be cycling hence why I asked again if there was a leak before the W/b

If you take a look at the log you can see where at cruise the W/b and N/b o2 are switching at the same point this is not happening at idle.

Calan The raw voltage looks about right so I would think that the Wideband is reading right...

OP there are a few more things I don't see in the log that would be good to keep track of

LTFT low
LTFT mid
LTFT High (yes I know 2g's don't have one But you can turn on it's function)
Battery
ISCPosition
Baro
LrnIdleadjust
TPS Delta

 

[Calan]

FOR TUNING? The rear o2 is the wideband and the only thing that realy matters as far as any real tuning IMHO

I think he was referring to a stock rear O2 sensor, which would be used for emissions rather than engine control.

Calan The raw voltage looks about right so I would think that the Wideband is reading right...

Good point.

That would point back to a leak between the two sensors, and it's only having a noticeable effect during long exhaust pulses. That's a pretty small area to search for one though, so it should be easy to find.

 

[razrman8]

I'm saying test pipe to cat back because my wb is in rear o2 location, right after this flange.