Considering Going to Open Loop Only - TBI

[Daveo91]

I have run my mild mod 454 in open loop for years now. The motor is much smoother, allows me to have highway lean cruise and pull down with converter locked. I have intake air temp incorperated so I don't need to change my tune from summer to winter. I use a 7427 and if I go much change in altitude I have a closed loop chip I put in. One of the biggest things with a open loop tune is it has to be correct. No sloppy ve table.

Thanks for the info. This is exactly what I want to do. I’m actually kind of excited about the tuning part and totally get it about the dialed in VE table since I’m completely eliminating the feedback loop and resulting adjustments it causes. Thanks.

[1project2many]

Thanks for the input. So I’m actually talking about OL everything, not just idle.

Well, you need a way to make the vehicle driveable. There were export cal / leaded fuel vehicles sold in the US which were OL. Closed loop enable temp was set sky-high. Personally I think your combination is not extreme enough to warrant OL only but it's obvious you're frustrated with the situation. Setting up a carb manifold under tbi can be tough. There's a larger variation in intake air temperature and you need to deal with the effects of fuel separation when it occurs. I've always tried to deal with it by getting steady-state right before investing any serious time in the transitions. In some cases it takes far more time working out what's happening than figuring out what to change.

[steveo]

open loop can be great. it's nice to be able to forcibly give an engine 'what it wants' rather than giving up that control to a feedback sensor.

some engines just don't take kindly to closed loop, because things aren't linear enough in the exhaust and the oxygen content in a particular moment no longer reflects reality of current air fuel ratio. reversion, misfires, big loping cams, look just like fueling errors, but the corrections cause a feedback problem and make things worse. in reality, these things usually only happen way outside the operating range of the engine and its cam, and if you don't have a really big cam, usually just idling.

coincidentally that's exactly when the engine seems a bit more 'thirsty', because a bunch of it pools in the manifold and cylinder walls, some natural misfiring occurs and dumps it out the exhaust, and inefficient combustion just needs a bit more margin of error, right when your closed loop system starts pulling fuel. you cant really fix that with a narrowband system.

it's a real shame most ECMs we work with don't have a good way to enable closed loop just for specific operating ranges, so we can say 'closed loop when >2000RPM and >30KPA MAP', that way we can be open loop where closed loop just doesn't work properly.

there are lots of tricks to making closed loop work better though especially in 'driving' ranges

having a near-perfect VE table sure helps a lot to start with so the corrections are fairly small and consistent especially through transitions...

usually if things run poorly in closed loop its because the corrections being made, whether they're correct or not, are too extreme, then it has to swing back the other way.. and you get modulation and feedback.

you can go WAY lean or rich (like 10:1 to 16:1+) in most cruising situations as long as its consistent and BARELY notice the difference, but you do notice modulating between AFRs as it's actually modulating engine torque by a decent percentage, and that's what you have to avoid (and why open loop cruising is so dramatically better if your current closed loop config is flunky)

restricting the integrator (INT) among other things can be a good start to make sure things dont swing too far out of control if there is a fueling error. logs should be very revealing about what's happening in closed loop and why it runs like crap. slowing down the reaction times of the closed loop system usually helps a bit too. most calibrations have an integrator delay time that needs to correspond to how far down the pipe the o2 sensor is.

[myburb]

I know this is off the OP question but if anyone knows where you can find factory open loop export bins for a 7.4 I would be interested to see how the factory went about it.

[1project2many]

it's a real shame most ECMs we work with don't have a good way to enable closed loop just for specific operating ranges, so we can say 'closed loop when >2000RPM and >30KPA MAP', that way we can be open loop where closed loop just doesn't work properly.

If the code is accessible it can be modified.

I know this is off the OP question but if anyone knows where you can find factory open loop export bins for a 7.4 I would be interested to see how the factory went about it.

OEM open loop calibrations are known as "leaded fuel" or "export" calibrations. Look for CTS enable to be set high to prevent CL operation. I will locate an '86 5.7 cal for leaded fuel (7747) and post it for evaluation.

[steveo]

If the code is accessible it can be modified.

of course, but a bit complicated for the new tuner. it's not just a matter of setting a flag if some variable exceeds some constant. my attempts to do this with $EE were buggy at best. it's pretty easy to stop it going into closed loop until a certain RPM, but making sure it exits closed loop as well, and still respects other closed loop parameters so you don't get thrown into closed loop while the engine is still warming up... all adds an extra layer of complexity.

even after i thought i had it nailed in EE it'd still randomly kick into closed loop and I could never figure out why

... ended up running full time open loop on that car

[delcowizzid]

If you can find the commanded fuel ratio tables its easy just set the rpm and map area you want open loop to anything but 14.7 or whatever your stoic value is

[Daveo91]

of course, but a bit complicated for the new tuner. it's not just a matter of setting a flag if some variable exceeds some constant. my attempts to do this with $EE were buggy at best. it's pretty easy to stop it going into closed loop until a certain RPM, but making sure it exits closed loop as well, and still respects other closed loop parameters so you don't get thrown into closed loop while the engine is still warming up... all adds an extra layer of complexity.

even after i thought i had it nailed in EE it'd still randomly kick into closed loop and I could never figure out why

... ended up running full time open loop on that car

Yeah, being able to do this would solve my problem and still allow CL in most situations. My issues all center around decel, idle, and just off idle accel. I'm interested in learning how to get deeper into the code, but I'm not there yet for sure. Making those mods are out of my league for now. That being said, I'm not sold on running OL all the time yet. I'm still doing some more experimentation with CL and I'll have my Moates Autoprom soon so I can do realtime emulation which will make that experimentation a lot easier. It should also allow me to combine my WB output with the PCM datastream.

Now I have a new problem. Not sure if it is connected with running in OL only or if it's something else. Runs fine when cold, but after it gets up to a certain temp it has an idle glitch - drops down to less than 500 rpm for a moment then recovers. Cycles about every 5 seconds. When it's doing this, part-throttle, off-idle accel is poor and uneven - exhaust note sounds like it's hitting on four cylinders. That's only off-idle though and only at light throttle. Hard accel is fine, and accel at speed is fine. I have a steep driveway, getting it up there at very light throttle it's cutting out and will barely go up it unless I use some significant pedal. But when it's cold it'll do the same scenario fine. Weird?

Thanks for the input from all.