Running Rich at Idle w/ DSMlink v3

[volkri25]

Mod list for what matters.

-8 blad HX35 w/ .55 AR housing.
-Built 6 bolt block and 7 bolt head
-Greentop Cas
-850cc injectors
-Fuelab AFPR
-Wally 255 rewired.
-Speed Density (AEM 3.5 bar boost and AEM AIT)
-BPR8es plugs
-COP setup
-FIAV partial blockoff (could affect ISC number?)
NO EGR or anything left on the car


So heres the deal. I followed the beginners guide to dsmlink tuning and nothing is working out. My airflow metering if a little off. Its at about .4 at ilde when it should be in the range of .24-.3. So my guess is that it is in the .24-.3 range but the ecu believes its around .4 therefore injecting 25% more fuel then needed, making me pretty rich at idle.

Im averaging 11.0:1 A/F at idle when it should be 14.7:1. Am I correct when assuming this? I have FIC 850cc injectors. Deadtime is set at 330 and Global fuel at -47% like it should be at.

Heres my log. Is there some core value i have to edit (F8) in datalog screen?

 

[Calan]

At a quick glance, it looks like your front O2 is dead and the ECU dropped into open loop. (STFT was pegged while it was in closed loop). It looks like your VE table hasn't been tweaked yet, so you are running rich.

I would start by figuring out what is up with the O2 sensor, or simulate it with your WB.

 

[volkri25]

Sounds like your probably right man. Thanks A lot

The SD VE table unaltered wouldnt kick it out of a closed loop would it?

Also since im using an AEM uego as the rear sensor, can i just cut and wire in my rear sensor to the front wiring? The sensors are the same just the wiring is different correct?

 

[snowborder714]

Did you even build your VE table using a MAS/MAF or did you just hard switch over? If you switched over, you'll have to build it by hand or find a MAS/MAF to install temporarily to build the table.

 

[Calan]

The SD VE table unaltered wouldnt kick it out of a closed loop would it?

No... a fubared front O2 will though.

Also since im using an AEM uego as the rear sensor, can i just cut and wire in my rear sensor to the front wiring? The sensors are the same just the wiring is different correct?

You don't have to cut anything; just check the box on the "NBO2 sim" tab to tell the ECU to use your wideband to simulate the front O2 sensor. You've already told the ECU where the WB is located on the "ECU Inputs" tab, so nothing else is required.

Although simulated NB works nicely and this should get you into closed loop so you can get fuel trims in line... I would still replace the front O2 sensor and use that while dialing the car in. For one thing, it gives you another independent source to compare against. It will also prevent errors in initial calibration if your WB is off a little.

Brian brings up a good point. It looks like your VE table hasn't been touched, and without a MAF calibration to start with, it's going to be more work to get it dialed in. Not impossible, but more involved.

Get the car to idle in closed loop first, then go from there.

 

[volkri25]

Yea I kinda know someone who was trying to borrow a GM maf to dial it in before tuning. I don't really have that option. And Id rather pay a shop 100+ an hour to do it without as it would probably be cheaper.

Yes unfortunately its hard wired over. The car was built with the sensors being used and that is all.

BTW, I used my rear o2 by cutting the wires from the front harness and rear sensor and will see if that works tomorrow.

(Rear and Front o2 both use a blue, white, and 2 black wires, so IMO the sensors are identical minus the plug) A simple solution till this one dies (that is if its in working order)

Thanks a lot for the tips guys!!!

 

[JayRolla]

If you try to simulate the wbo2 that far from the o2 housing it will be confused. Ecmlink has a feature to slow down the wbo2's pulse to help it read correclty but it wont work for the rear o2 location. The afr at cruise will fluctuate like crazy because the ecu sees a broken signal when the o2 is located so far from the o2 housing.

 

[Calan]

BTW, I used my rear o2 by cutting the wires from the front harness and rear sensor and will see if that works tomorrow.

I'm confused...If you already have a wideband installed, you can use it to simulate the front (narrowband) O2 sensor without cutting or rewiring anything. You just have to tell the ECU where to find it.

If you try to simulate the wbo2 that far from the o2 housing it will be confused. Ecmlink has a feature to slow down the wbo2's pulse to help it read correclty but it wont work for the rear o2 location. The afr at cruise will fluctuate like crazy because the ecu sees a broken signal when the o2 is located so far from the o2 housing.

If it causes problems with the FT's because of the delay, he can play with the O2 feedback adjustment rate in the "STFTAndO2Feedback" DA table. I agree it's not the best solution, but better than nothing.

OP - You could always just lock the ECU in open loop and start calibrating (I actually prefer to work in open loop).

 

[volkri25]

yea Craig I know I can just simulate it but I kind of want a working front o2. The rear o2 sensors on our cars are the exact same as the front, just the plug is different. If you cut the plug off you can wire a rear o2 into the front o2 location, as I just had it laying around due to the fact it is replaced by the uego in the rear. I just want to see if this will keep my car in a closed loop, without checking the nbo2sim box.

I will keep you guys updated.

 

[Calan]

Ahhh... I see. I thought you were saying that you were cutting wires so that you could simulate the front O2.

 

[snowborder714]

Last I knew is that the front o2 and rear o2 were 2 different sensors. The rear was just to check if the catalytic converter was there and working, while the front dealt with fuel trims and such.

 

[volkri25]

So i tested it out. And it works but the car is still running rich. About 12.0:1 now. The sensor i wired in was reading between .8-.9 which indicates rich, yet my ecu isnt slowing down the fuel to get it to .45 and it wont fix my idle. Something is kicking me out of closed loop and I have no idea what it is......

I tried using a narrowbando2sim but it didnt do anything. Maybe it because i dont have the white? wire from the wideband connected to anything?

 

[Calan]

So i tested it out. And it works but the car is still running rich. About 12.0:1 now. The sensor i wired in was reading between .8-.9 which indicates rich, yet my ecu isnt slowing down the fuel to get it to .45 and it wont fix my idle. Something is kicking me out of closed loop and I have no idea what it is......

If you are dropping into open loop, the ECU won't use the trims to do anything with the fuel. The AFR will be result of your airflow and fuel calibration. So if your VE table is off and you're in open loop, your AFR could be anywhere.

Either lock the ECU in open loop and start dialing in your VE table via your wideband readings, or get the front O2 sensor working and use the fuel trims in closed loop.

I tried using a narrowbando2sim but it didnt do anything. Maybe it because i dont have the white? wire from the wideband connected to anything?

The NB simulation doesn't require any special wiring. Assuming your wideband is hooked up properly (to display wideband), ECMLink will use it to simulate the NB switching point.

 

[volkri25]

Kind of getting frustrated. Tried to raise all parameters in the open loop tab to keep it in closed loop and it never enter a closed loop. What I found suprising is that my MDP tab is reading 12 in/hg? that would be equal to 5 psi? Which would automatically put me in an open loop. Heres my log. Front o2 is locked until the end and as you can see my wired in o2 works just fine.

At this point there is no reason the car should not be in a closed loop...... unless the computer is seeing MDP as positive pressure when its not. (I have my AEM3.5boost wired into the MDP and its all setup correctly in the fields. The AEM3.5boost shows a level close to that of my boost gauge.)

Not sure if the computer is being fooled by the MDP as boost instead of vaccum (should be -12 in/hg.) Regardless something else is going on. Even when front o2 is locked and narrowband sim is on, its still in open loop....

Any more help would be appreciated .

Heres my log.

 

[Calan]

You are dropping into closed loop now as expected, and for the most part things look pretty normal (other than no airflow calibration yet)...except your front O2 still doesn't appear to be cycling correctly. The voltage seems about right, but the waveform doesn't look like anything I have seen (could be due to the rich/maxed STFT situation though).

You are in closed loop at around 350 secs in, and you are rich because your STFT is maxed and the ECU can't pull anymore fuel. Pull some airflow at that area of your VE table and you will see the trims come up and your AFR's begin to stabilize.

Your deadtime is probably off also, considering how far off the airflow is.

 

[volkri25]

Yea i just saw that. So my car should be in closed loop after about 30 seconds but its not till about 8 minutes in. Whats the deal. I have a mishimoto 143 degree thermostat, is that to cold? Should still be in closed loop at this temp correct. Seems that at about 170 is when the car enters a closed loop.... So confused on why it took so long to enter the closed loop.


On a second thought, a richer car will take much longer to warm up because less heat is generated during combustion. So if I adjust my airflow table and the car begins to lean out a little it should take just a minute or two to warm up to operating temp?

I tried adjusting the mafcomp sliders (with the hz) and it didnt affect my airflowperrev number which is way off the correct value. I suppose as I dont have a MAS this wont do anything? I have to adjust the numbers in the SD table. The problem is I have no idea what the number in the SD table mean :/. like for example 55?

(The reasoning the the thermostat is that im moving to arizona next summer.)

Deadtime and Global fuel is set per DSMlinks info.

 

[Calan]

It's most likely due to the O2 sensor, but I'm not sure.

If I were you, I would take a break from it and do some reading on the ECMLink wiki. Start over with the basics (base timing, BLT, exhaust leaks, plug gap, etc), and then go through the base ECMLink setup.

It may be something simple and you are just too close to it at this point.

start [ECMTuning - wiki]
fueltrimupdatepoints [ECMTuning - wiki]

 

[volkri25]

Alright well I got it to enter closed loop in about 10 seconds after startup

It was just that I had to lower the 1k rpm band on my VE table and it began to stabalize like you said. The thing is why was there default setting way to high? There default around the area I changed was 55.0 where as now its 41.0 Is my engine that inefficient? Then again would 272 cams and higher compression pistons alter this number?

The AFR for some reason doesnt want to bounce around 14.7 but rather bounce around the mid 15's I dont suspect this to be a bad thing persay, but is there any harm in haveing a 15 afr for cruise and idle?

Also im having a tough time making both STFT and CombFT equal zero. The issue is what affects these values. I slowly adjusted Maf Comp Adjust, SD VE, and Injector deadtime but couldnt get these values to hit zero. Is +-3 an any direction really an issue?

Just curious though, What affects the STFT value and what affects the CombFT value?

Also Is the whole point to get the AirFlowPerRev to match SDAirFlowPerRev?

Heres my log.

Thanks!

NM log is to big, will take another one tomorrow for some final idleing critiquing

 

[Calan]

MAFComp doesn't do anything unless you are running on a MAF.

With a MAF, you use MAFComp to calibrate airflow. With SD, you use the VE table to calibrate airflow. MAF and speed density are just two different ways of accomplishing the same thing; letting the ECU know how much air is entering the engine.

Think of CombinedFT as the amount of error there is in the ECU's calculated fuel (or air) while operating in closed loop. The ECU thinks there is a certain amount of air entering the engine, and injects a calculated amount of fuel to get a target AFR of 14.7:1. The front O2 sensor then lets the ECU know if it hit that AFR or not. CombinedFT is the percentage it missed it by. (It's actually a combination of the long-term fuel trims and the short term measurement that the ECU sees from the O2 sensor).

So anything that affects fuel or airflow calibration can affect the fuel trims. Assuming your injector data is set up properly (global fuel and deadtime), the most likely culprit is uncalibrated airflow; the engine isn't really getting the amount of air it thinks it is. When you adjust the VE table or MAFComp sliders, you are calibrating the airflow so that the ECU sees the actual amount of air entering the engine and can calculate the correct amount of fuel.